CA2109258C

CA2109258C - Electrical circuit interrupting device

Info

Publication number: CA2109258C
Application number: CA002109258A
Authority: CA
Inventors: James L. Lagree; Joseph J. Matsko
Original assignee: Eaton Corp
Current assignee: Eaton Corp
Priority date: 1992-10-27
Filing date: 1993-10-26
Publication date: 2003-03-11
Anticipated expiration: 2013-10-26
Also published as: JPH0775243A; EP0596643A1; CA2109258A1; JP3811194B2; EP0596643B1; DE69311059D1; US6055145A; DE69311059T2

Abstract

A microprocessor based overcurrent trip unit which generates trip signals as an adjustable function of current and time, has a visual representation of the trip function on a front panel with 2-color LEDs associated with the trip function serving as indicators of trip conditions when red, and of a selected programmable parameter when green. The gren LEDs flash to indicate a parameter selected for modi-fication in a program mode and illuminate steady in a view mode.

Description

54,2=~-:-2 ELECTRICAL CIRCUIT INTERRUPTING DEVICE
BACKGROUND OF THE INVENTION
Field of the Invention The present invention relates to wn electrical circuit current interrupting device including an overcurrent trip unit for an electrical circuit interrupting device preferably for a metal clad switch gear, molded case circuit breaker and the like, protecting electrical conductors from the damage due to excessive electrical currents and, more particularly, to a microprocessor based overcurrent trip unit with adjustable tripping characteristics which is adapted to continuously monitor the electrical current flowing through the circuit interrupting device and initiate a trip as a function of a selectable tripping characteristic and with visual - 2 7*.~_a"_-indicators on a front panel which provide visual indications for programming the adjustable tripping characteristics and a visual indication of the condition causing the trip.
Hackgr~und r.~.for:nat:~r.
Various overcvrre~. devices are k::own :.~. _=e a:.
for protecting electrical co::ducto:s i.~. an electr_ca_ dis-tribut'_on system f:om damage due to an excessive e:ect::._a_ current. Suez overc;:rre~: cevices are typically c::a:
acterized oy their time-cu::ent characteristics or p:otec-tion curve. Such protecticn curves a:e normally uti:ized _..
limit the temper~atu:e :ise c' a:~ elect:'_cal cc~ducto: d;~e =
an excessive electrical csrrent in order to prevent damage. For example, the =e:rperature rise oL the electrical conductors during certain excessive current conditions can be approximated by the produce o~ the square o. the electrical current and the ci:ne period that such elect: ica current is applied to the electrical conductors (e. g., I2t). Thus, for an electrical motor sated for a predece:-mined temperatsre rise, for example, SS°C, such overcur:e~~_ devices are used to limit the temperature rise of the electrical conductors within the motor to the :aced temperature rise.

3 3~,Z_3-_ In order to facilitate selection of an overcurrent device with a suitable characteristic for use with an electrical motor, motor operating curves (for example, as shown in FIG. 2) are normally provided by a motor manu-facturer. Such motor operating curves graphically illus-trate the normal time and current characteristics of a particular electrical motor at its rated temperature rise.
Accordingly, in order to protect the motor from damage and at the same time prevent spurious tripping of the motor during start-up, it is necessary to "coordinate" the motor operatinq curve with ~~e ~iTe-cir=e..~.t c~arac~e=istics .._ 3-:
e'~ect: ical overcur rent .:evice utilized ~n ..::e elects ~~a:.
circiit brea<er _eedi::g _~e :~o_or.
T_t is a'_so '<~own to coordinate _::e overc::==e-:t device provided on the electrical circiit b:ea~cer 'eedi~g ci_c;:its and loads protected by war°~c~:s cverc~~:=e.~t devices ;:ti=iZed in the elect:ical distribution syste:r, in ore:
;prevent unnecessary trypping o° such circuit brea:ce;s.
'.hus. the time-current characte:istics for all c: .~e various overcurrent devices in the electrical distributicl ;system are coordinated to provide for "selective"
_:ipping. Se=ective trippi~g refers so trippi::7 of c::_y whose portions of an electrical distribution system necessary to isolate an excessive electrical csrre::t.
selective tripping provides for several advantages in an electrical distribution system.

4 ~~.2=3-_-2 First, selective tripping greatly improves ~_~e reliability of tae electrical distribution system. ,or example. various electrical interrupting devices, for example. motor control centers. unit substations and ~~e like. include a plurality of c~~c~sit breakers and ~he 1<<e for. providing electrical power to various electrical loads. Hy utilizing selective tripping, a fault at o:
adjacent one of the electrical loads would result in or.~y that goad being isolated from the electrical distribution system. The balance of the electrical loads fed from the motor control center or the like would be undisturbed. As such, the reliability of the electrical distribution system is greatly improved.
Second, selective tripping facilitates the main-tenance cost for locating and repairing the source of an excessive electrical current. More specifically. by uti-lizing selective tripping, only the circuit breaker or other ~~,2_9-_-2 prCteCt7.Ve device i;IL'Tled=ate'=l '.:pS~=?dm C~ .::e Sol::Ce C_ excessive electrical cur=ent is t=ipped. ~ccorci_-.g:..~, _-a source or t!~e excessive overc::::e~t cap ;e::eral:l :,e :cca=?c _elat_~.e'_y c::ick:.y thereby decreasi.~,g _::e :nai.~.,e.~.a::ce __,e and also decreasing the down time for t:~e e=ec~r_cal _cad t'.~.at was tripped. Moreover, suc.~. se_ec:ive t:'_poi-:g ai_..
prevents unnecessary tripping o~ interrupting devices, s.:~.-.
as fuses. which could req::ire replacement taereby cecreasi-7 tae maintenance cost and down t_:.-.e cz tae system.
Selective tripping fur~::er optimizes tae cyc=a life of the circuit breakers in t::e e1 ectr ical distr ibLt:o.~.
system. !fore speci=ica:.:.y, t::e var~.ous molded case circ::_t breakers and metal clad switc'.~.gear breakers ~ith:n a:.
electrical distribution system are generally adapted to overate a predetermined number of times before taey eit~er need to be replaced or serviced. ':h.is predetermined number is known as the cycle life. Hy preventing unnecessar~r ~~perations of the various elects is circuit breakers w:.::..~.
the distribution system, the cycle life of the var:.ous breakers is thus improved.

6 a~.2~a-_-z U.S. Patent Specification No. 4,827,369 discloses an electrical circuit interrupting device including an overcurrent trip unit comprising current sensing means for sensing an electrical current flowing through said electrical current interrupting device, trip means responsive to said current sensing means for generating a trip signal as an adjustable predetermined function of the time and the magnitude of the electrical current flowing through said electrical circuit interrupting device, a user interface panel provided to present a visual representation of said adjustable predetermined function, input means for selectively adjusting parameters of said adjustable predetermined function, and a plurality of indicator means associated on said user interface panel with said visual representation, each indicator means indicating a parameter of said adjustable predetermined function_ The object of the invention is to facilitate the adjustment of the various parameters of said adjustable predetermined function by an operator.

~.~, 2.~-.
According to the invention this object is achieved in that each said indicator means can take a first state in response to a trip signal generated by the trip means and a second, different state in response to adjusting of the corresponding parameter of said adjustable predetermined function through said input means.
An alpha-numeric. display can be used for presenting the value of a parameter of said adjustable predetermined function which is being adjusted and t:he indicator means can be used to relate the parameter being adjusted with the visual representation of the adjustable predetermined function.
Preferably, the indicator means are lights and, v_a particular two color LEDs displaying a first color to indicate the trip condition and a second color during programming. The overcurrent trip unit can Further have a viewing mode in which the values of the parameters are merely presented with the appropriate LED illumi.~.ated continuously, and an adjusting :node in which the proper LED
flashes.
Conveniently, switches on the user interface for programming the parameters of the adjustable predetermined function of 2~~~~~~ ;~,z18-_-z .~ - 8 current acid time are color coded to differentiate them 'rc:r test buttons on the tront panel used to test t!:e overcurre.~.c trip unit. This reduces contusion, and airs t:~e user in quickly and accurately selecting the correct s~aitches For performing the desired function.
DESCRI?TION OF THE DI~Av~I~lG
These and other objects of the present inve.~.~~oz will be readily apparent upon consideration of the following description and attached drawings, wherein:
FIG. 1 is a single line diagram of an exemplary 4.16 kV radial distribution system;
FIG. 2 is an exemplary graphical representation of the coordination of the various overcurrent devices of the system illustrated in FIG. 1;
FIG. 3 is an exemplary overcurrent protection curve;
FIG. 4 is a graphical representation of the long time delay portion o~ a protection curve for an overcurrent tripping device in accordance with the present invention illustrating an It characteristic for the long time delay portion;
FIG. 5 is similar to FIG. 4 illustrating an I2t characteristic;
FIG. 6 is similar to FIG. 4 illustrating an I4t characteristic:
FIG. 7 is a graphical representation of the long time delay and short time delay characteristic of the overcurrent tripping device in accordance with the present invention having relatively wide long time delay and short time delay adjustment ranges uncorrected for overlap;
FIG. 8 is similar to FIG. 7 illustrating the long time delay and short time delay portions corrected For overlap in accordance with the present invention;

- 9 2.~~~?~~ s4,21e-_-z FIG. 9 is an exemplary graphical iliustrac:on ~°
the output current wavezorm or a saturated current tra:,s-former;
FIGS. 10-16 represent a schematic reoresentaticn of the overcurrent tripping device in accords.nce wi=:~. ~~e present invention;
FIGS. 17-31 represent a flow chart for the over-current trippi~.g device in accordance with the present invention;
FIG. 32 is a functional representation of an ap-plication of t'.~.e IC in accordance with the present inven-tion;
FIG. 33 is a functional block diagram of the IC in accordance with the present invention;
FIG. 34 is a functional block diagram of a portion of the analog portion of the IC in accordance with the present invention;
FIGS. 35(a)-35(d) are diagrams of alternate clock generator connections for the IC in accordance with the present invention;
FIG. 36 is a memory address map for the IC in ac-cordance with the present invention;
FIG. 37 is a format diagram for configuration registers CFR and ACFR which form a portion of the IC in accordance with the present invention;
FIG. 38 is a format diagram for an EEPROM control register NVCR which forms a portion of the IC in accordance with the present invention;
FIG. 39 is.a format diagram for a dead-man control register DMC which forms a portion of the IC in accordance with the present invention;
FIG. 40 is a format diagram for A/D conversion interface registers ADZ, AMZ, ADCR, AMUX, ACSF, AVSF and ADC

~~a~~~~
-- - to s4,zls-_-z which foam a portion of the IC in accordance with ..::e present i~vention;
FIG. 41 is a format diagram fo= comparator mode control registers CMPI and CMPST and the pulse width modu-lated outaut control register PWM which form a oo~tion of the IC in accordance with the present izve.~.t~on;
FIG. 42 is a block diagram of a programmable ..imer which forms a portion of the IC in accordance with t;.e present invention;
FIGS. 43(a)-43(d) are timing diagrams for t:~.e timer of FIG. 42;
FIG. 44 is a for,aat diagram for program.~~able ti:r.er registers TCRH, TCRL, TARH, TARL, TICH, TICL, TOCH, TOCL, TCR and TSR which form a portion of the IC in accorda.~.ce with the present invention;
FIG. 45 is a connection diagram of a serial peripheral interface (SPI), single master, which for:r.s a portion of the present invention;
FIG. 46 is a connection diagram of a typical S2I
with multiple masters;
FIG. 47 are clock and data timing diagrams for the SPI illustrated in FIG. 46;
FIG. 48 is a format diagram for the SPI interface registers SPD, SPSR and SPCR which form a portion of the IC
in accordance with the present invention;
FIG. 49 is a format diagram for the PORT D in-terface registers PDC and PDD which form a portion of the IC
in accordance with the present invention;
FIG. 50 is a block diagram of a parallel port which forms a portion of the IC in accordance with the present invention;

11 ~~~~~ )~ X4.218-.-2 FIG. S1 is a format diagram for PORT 3 i.~.ter'ace registers PHC and PHD which form a portion cf the IC _., accordance faith the present invention;
F IG. 52 is a format diagram for PORT C inter 'ace registers PCC and PCD which form a portion of t'.~.e ~C i~.
accordance with the present inventio.~.;
FIG. 53 is a format diagram for PORT D interface registers PDC and PDD which form a portion of t!:e IC ..~.
accordance with the present invention;
FIG. 54 is a format diagram for the communicatic~s controller interface registers ICAH, ICAL, ICM3, ICM2, IC.'~1, IC."~0, ICSR and ICCR which form a portion of the IC i.~, accordance with the present invention;
FIG. 55 illustrates the control message and data . message format diagrams for the communication controller which forms a portion of the IC in accordance with the present invention;
FIG. 56 is an overall block diagram of the ccm-parator subsystem and A/D input subsystems of the IC in accordance with the present invention;
FIG. 57 is a schematic diagram of the quadcom-parator subsystem in accordance with the present invention;
FIG. 58 is a schematic diagram of the address decode logic for the comparator control registers CMPST and CMPI and the configuration register CFR in accordance with the present invention;
FIG. 59 is a block diagram of the analog subsystem in accordance with the present invention;
FIG. 60 is a schematic diagram of the micropro-cessor bus interface logic in accordance with the present invention;

21~~~a~
12 54,213-_-2 ..
FIG. 61 is a sc'.~.ematic diagram of t:~e add:ess decode logic for the ;nic:oprccessor bus i:~terface registers in accordance with the prese:~t invention;
FIG. 62 is a sc::ematic diagram of contro'_ a~d status registers in accordance with the present i.~.ver.tion;
FIG. 63 is a bloc:t diagram of the analog digital control logic portions of the IC in accordance ~aitn ..~:e present invention;
FIG. 64 is a block diagram of the analog co.~.tro=
logic in accordance with the present invention;
FIG. 65 is a schematic diagram of the curie.~.~
multiplexes (ML'Xy control logic in accordance :pith t::e present invention;
FIG. 66 is a schematic diagram of the voltage MUX
control logic in accordance with the present invention;
FIG. 67is a schematic diagram of the auto-zero registers in accordance with the present invention;
FIG. 68 is a schematic diagram of a pine microsecond timer in accordance with the present invention;
FIG. 69 is a schematic diagram of the auto-zero control logic in accordance with the present invention;
FIG. 70 is a schematic diagram of the auto-range control logic in accordance with the present invention;
FIG. 71 is a schematic diagram of the auto-range state machine in accordance with the present invention;
FIG. 72 is an overall block diagram of the analog circuitry in accordance with the present invention;
FIG. 73 is a schematic diagram of the input MUX
system in accordance with the present invention;
FIG. 74 is a block diagram of the quad comparator system in accordance with the present invention;
FIG. 75 is a schematic diagram of a band gap regulator in accordance with the present invention;

13 ~~~r~~~ 54.218-I-2 f FIG. 76A is a schematic diagram of a shunt regu-lator, B+ comparator and a power monitor in accordance the present invention;
FIG. 76H is a schematic diagram of exemplary ex-ternal conditioning circuitry and power supply circui=_y for use with the IC in accordance with t..".e present invention;
FiG. 76C is a schematic diagram of an exemplary external regulator circuit for use with the IC in accordance with the present invention;
FIG. 77 is a schematic diagram of a biasing cir-cuit in accordance with the present invention;
FIG. 78 is a schematic diagram of another biasi.~.g circuit in accordance with the present invention;
FIG. 79 is a schematic diagram of an analog tem-perature sensing circuit in accordance with the present invention;
FIG. 80 is a schematic diagram of the ranging circuitry for the voltage amplifier in accordance with the present invention;
FIG. 81 is a schematic diagram of the current mirror and amplifier in accordance with the present inven-tion;
FIG. 82 is a schematic diagram of the current mirror in accordance with the present invention;
FIG. 83 is a schematic diagram of the offset correction circuitry in accordance with the present inven-tion;
FIG. 84 is a schematic diagram of the auto-zeroable voltage and current amplifiers in accordance with the present invention;
FIG. 85 is a block diagram of the communication controller which forms a portion of the IC in accordance With the present invention;

_ 1,~ z~~~~~s 54,218-~-2 'FIG. 86 is a block diagram of a digital demcc~-lator which forms a portion of the communication cor.t:ol~e:
in accordance with the present invention;
FIG. 87 is a schematic diagram of a :raster c_oc<
generator which forms a portion of the communication con-troller in accordance with the present invention;
FIG. 88 is a schematic diagram of a bit phase timing generator which forms a portion of the commur.icato.~.
controller in accordance with the present invention;
FIG. 89 is a schematic diagram of a receive correlator which forms a portion of the communication con-troller in accordance with the present invention;
FIG. 90 is a partial schematic diagram of a cor-relator counter which forms a portion of the communication controller in accordance with the present invention;
FIG. 91 is the remaining portion, of the schematic diagram of a correlator counter which forms a portion of t.e communication controller in accordance with the present invention;
FIG. 92 is a schematic diagram of the demodulator control logic which forms a portion of the communication controller in accordance with the present invention;
FIG. 93 is a schematic diagram of a bit counter which forms a portion of the communication controller in accordance with the present invention;
FIG. 94 is a schematic diagram of bus interface logic circuit for the communications controller which forms a portion of the communications controller in accordance with the present invention;
FIG. 95 is a schematic diagram of an address de-code logic circuit for the communications controller's bus interface register in accordance with the present invention;

15 ~~~~~~~ 5:~~218-_-2 ~fFIG. 96 is a schematic diagram of the address registers which forms a portion of the communication con-troller in accordance with the present invention;
FIG. 97 is a schematic diagram of a messace register ICM1 which forms a portion of the commun'_cat=on controller in accordance with the present invention;
FIG. 98 is a schematic diagram of a message register ICM2 which forms a portion cf the communicat:.o.~.
controller in accordance with the present invention;
FIG. 99 is a schematic diagram of a message register ICM3 which forms a portion of the communication controller in accordance with the prese.~.t invention;
FIG. 100 is a schematic diagram of a messag a register ICMO which forms a portion of the communication controller in accordance with the present invention;
FIG. 101 is a schematic diagram of control and status registers which forms a portion of the communication controller in accordance with the present invention;
FIG. 102 is a schematic diagram of bits 26-19 of a shift register which forms a portion of the communication controller in accordance with the present invention;
FIG. 103 is a schematic diagram of bits 18-2 of the shift register of FIG. 102;
FIG. 104 is a block diagram of the communication controller control logic which forms a portion of the IC in accordance with the present invention;
FIG. 105 is a schematic diagram of the transmitter control logic circuit which forms a portion of the communication controller in accordance with the present invention;
FIG. 106 is a schematic diagram of a BCH computer Which forms a portion of the communication controller in accordance with the present invention;

X4,2_?-__~
FIG. 107 is a schematic diagram of an acd:=ss ccmparator circuit which forms a portion of the ccmm~,:~:ca-_ion controller in acco:dance wi_:~ t~e present ..~.ve.~.;ic.~.;
rFIG. i08 is a scaemat_c diagram of an i~str::ctic::
decoder circuit which forms a portion of t'.~.e commur.icat_=-:
controller in accordance with the present inve.~.tion;
FIG. 109 is a schematic diagram of a control a::d status logic circ;:it which forms a portion of t::e comp ;:::_ cation controller in accordance with the present inventi~::;
FIG. 110 is a continuation of FIG. 107;
FIG. 111 is a timing diagram of various str=oe signals utilized in the communication controller i.~. acc:,:-dance with the present invention;
FIG. 112 is a timing diagram of a reset signal utilized in the communication controller in accordance with the present invention;
FIG. 113 is a pin out diagram of the IC in accor-dance with the present invention;
FIG. 114 is a plan view of a front panel of t::e trip unit of the present invention; and FIGS. 115-124 are flow charts for controli:~g features of the front panel illustrated in FIG. 114.
DESCRIPTION OF T~iE PREFERRED E.~t80DIMENTS
General The present invention relates to an overcurrent trip unit for an electrical interrupting device. such as a molded case circuit breaker or a metal clad switchgear breaker of the type for example, disclosed in U.S. Patent Nos. 4,351,013 and 4,827,369.
as the assignee of the present invention, hereby Such overcurrent trip units are microprocessor based and include various input/output devices, such as membrane switches, light emitting diodes .' 17 '~~~~~~ e'~~218-~-2 (LED's) and displays, which form a user interface which provide for various functions including allowing the varicus setpoints to be selected and initiating a trip of the attendant electrical circuit interrupting device. Although such known overcurrent trip units are adapted to provide reasonably good coordination in an electrical distribution system, the overcurrent trip unit in accordance with the present invention provides for better coordination cf various overcurrent devices utilized in an electrical distribution system in order to provide relatively better selectivity and consequently improve the overall reliability of the system.
More specifically, in one embodiment of the invention. the function of the long time delay portion of the protection curve is selectable from a plurality of programmed functions, for example FLAT, It, I2t and I4t. By providing for such a selectable function, the overcurrent device incorporating the principles of the invention is able to be more readily coordinated with other overcurrent devices utilized in an electrical distribution system. In an alternate embodiment of the invention, the overcurrent trip unit in accordance with the present invention includes relatively wide long time delay and short time delay adjustment ranges. In order to prevent overlapping of the long time delay and short time delay portions, the trip unit in accordance with the present invention prevents any overlapping zones which could result in a loss of selectivity. In another alternate embodiment of the invention, the trip unit in accordance with the present invention solves the problem associated with instantaneous tripping during a condition when the current transformers are saturated.

2~~1~~~~
18 54,218-_-2 aDescription Of Radial Distribution System The various embodiments of the invention are best understood with reference to the single line .._.~gram illustrated in FIG. 1 which represents an exemplary 4.16 ~c'J
radial electrical distribution system. In the exemplary system, 4.16 kV medium voltage metal clad switchgear, shown within the dashed box identified with the reference numeral A20, forms the source. The 4.16 kV switchgear A20 includes a 4.16 kV bus, identified with the reference numeral A22, and a plurality of feeder breakers A24 for providing electrical power to various 4.16 kV electrical loads in the system. As shown, the 4.16 kV switchgear includes a feeder breaker A26 and a serially connected fuse A28 feeding a unit substation, shown within the dashed box identified with the reference numeral A30. The unit substation A30 includes an integral 4.16 kV to 480 V transformer A32 to form a 480 V
bus A34. The transformer A32 secondary winding is connected to the 480 V bus A34 by way of a 480 V transformer breaker A36. The 480 V bus A34 includes a plurality of feeder breakers A38 for typically feeding various non-cyclic electrical loads. such as 480 V motor control centers (MCC). In order to improve the reliability of the system the 480 V unit substation A30 includes a tie breaker A40 to allow electrical power to be supplied to the 480 V bus A34 in the event that the primary source (e.g., 4.16 kV bus A22 or transformer A32) becomes unavailable. As shown, a 480 V
feeder breaker A42 is used to feed a 480 V MCC, shown within the dashed box identified with the reference numeral A44.
Such MCC's A44 normally include a 480 V bus A45, a plurality of circuit breakers, starters and contactors and the like for supplying electrical power to various cyclic electrical loads. For example, a 480 V contactor A46 with an integral overload relay A48 is shown feeding a motor A50. As is 19 54,218-.-2 mown in' the art, such overload relays A48 include a bimetallic element, serially connected with the line conductors which interrupts the electrical power to the motor ASO during overload conditions, for example, due to fluctuations in the source voltage or mechanical problems, such as faulty motor bearings.
Moreover, each of the circuit interrupting. devices (e. g.. A26. A28~ A36. A40 and A42) in the illustrated radial distribution system include overcurrent protection. For example, the 4.16 kV feeder breaker A26 includes one or more current transformers A52 for monitoring the electrical cuzrent on the load side. These current transformers A52 are used to drive an instantaneous/overcurrent device A54, such as a Westinghouse type CO relay. As previously mentioned, the primary winding of the unit substation transformer A32 is additionally protected with the fuse A28. The circuit breakers A36, A40 and A42 in the a nit substation A30 are provided with solid state tripping units, for example, as generally described and illustrated in U.S.
Patent No. 4,827,369.
In such an application, it is necessary to coor-dinate all of the overcurrent devices in the distribution system to selectively isolate excessive electrical currents while at the same time leaving the unaffected electrical loads in the system undisturbed. For example, a faulty bearing on the motor A50 could result in a condition where the motor A50 is stuck in a locked .rotor condition - a condition where the .motor A50 draws between approximately four to six times its rated current. During such a condition, it is desirable that an overcurrent protection device isolate the motor A50 without disturbing any of the other electrical loads in the electrical distribution 20 54,23-_-2 including' other loads on the 480 volt MCC bus A45. S~~
selectively isolating the source of excessive electr:ca:
current and tripping only the motor ASO the system reliability is improved since the other electrical loads or.
the 480 volt MCC bus A45 would be virtually unaf'ec=ed.
Otherwise. without selective coordination, a fau_t at =~e motor ASO could cause tripping of the MCC feeder breaker A42 which, in turn, would cause a loss of the entire MCC A44.
At the same time, it is also desirable to avoid tripping the motor ASO during starting. Accordingly, t:.e overcurrent protective device used to protect the motor ( viz. overload relay A48) is coordinated with the normal time-current characteristics of the motor during a normal starting condition to prevent spurious tripping of the electrical motor during starting.
Coordination Of Interru ting Devices Such coordination is normally done graphically.
More specifically, the time-current characteristics of the overcurrent protective devices, and operating characteris-tics of various devices, such as electrical motors, .-e normally plotted on a logarithmic scale. In particular, tae setpoints and time-current characteristics are selected to provide coordination of all of the overcurrent devices in the distribution system to provide selectivity and avoid spurious tripping, thereby increasing the reliability of the system.
An example of such is illustrated in FIG. 2. More specifically, FIG. 2 is a graphical representation of the time-current characteristics of the various overcurrent devices in the electrical distribution system illustrated in FIG. 1 on a logarithmic scale. The vertical axis relates to time in seconds while the horizontal axis relates to current in amperes on a scale basis.

21 ~~~~~~ J 54,218-~-2 ~~Referring to FIG. 2, the curve, identified wit:
the reference numeral A56, illustrates the time-current characteristics of the motor ASO during normal operati.~.g conditions. Initially, when electrical Dowe r is first applied to the motor A50, the :rotor A50 is in what is '.known as a lock rotor condition. During this condition, as illustrated, the motor ASO draws anywhere from four to six times its normal full load rated current until the rotor reaches its rated speed. As illustrated in FIG. 2. this locked rotor condition is shown to last for about ten seconds. After the motor reaches its rated speed, the electrical current drawn by the motor ASO drops down to its rated full load current as illustrated by the portion of the curve A56 identified with the reference numeral A58.
During normal starting conditions, it is unde-sirable to trip the electrical motor ASO. Accordingly, the time-current characteristic selected for the overcurrent protective device (e.g., the overload relay A48) feeding the motor A50 is illustrated by the curve A60. As shown, the time-current characteristics A60 are selected to allow the motor A50 to start normally without tripping. However, should there be a mechanical problem or a fluctuation of the source voltage which causes the motor A50 to draw the locked rotor current for longer than normal (e.g., longer than ten seconds) the overload relay A48 would cause a trip the motor A50, thereby isolating the condition. The other electrical loads on the MCC bus A45 would thus be unaffected.
As previously mentioned, it is desirable to coordinate the time-current characteristics of the overload relay A48 used to protect the electric motor A50 with the other overcurrent protective devices described above in the electrical distribution system. Thus, the time-current characteristics of the overcurrent protection devices for 2~.~D~?~8 - 22 54,218-_-2 the 480 ~ circuit breakers A36, A.~O and A42 as well as t::e 4.16 kV feeder breaker A26 and fuse A28 (FIG. 1) are selected to coordinate with the time-current characteristic A60 for the overload relay A48, as shown. More specifically, as illustrated in FIG. 2, the curve identified with the reference numeral A62 illustrates the time-current characteristics of the overcurrent device utilized for t:~e 480 V feeder breaker A42. As shown in FIG. 2, t',~.e characteristics and setpoints are selected to protect t::e 480 V MCC bus A45 and the electrical conductor A63 feeding it. The 480 V feeder breaker A42 is used to protect the 480 V MCC bus A45~as well as the electrical conductors A47 feeding the bus A45 and would not normally interrupt for a fault downstream of the bus A45 unless there was a failure of a downstream interrupting device. For example, a fault on the feeder to the electrical motor A50 would normally be cleared by the overload relay A48 and would be "transparent"
to the overcurrent protection device associated with the MCC
feeder breaker A42.
In addition to coordinating the various over-current devices in the electrical distribution, considera-tion must also be given to the normal full load current on the 480 V MCC bus A45. Accordingly, the continuous current portion of the time-current characteristic of the overcurrent device associated with the 480 V feeder breaker A40 is normally selected to avoid tripping the MCC feeder breaker A42 during normal loading conditions. Accordingly.
the continuous current portion A64 of the curve A62, which represents the time-current characteristics of the overcurrent device associated with the MCC feeder breaker A42, is selected to be about 110% of the normal full load current of the 480 V MCC bus A45. Similarly, the normal full load current of the 480 V unit substation bus A34 must 23 ~~~~~~~ 54,218-~-2 also be taken into consideration in selecting the setpoint~
for a time-current characteristic for the overcurrent devices associated with the transformer breaker A36 and t::e tie breaker A40.
Additionally, the unit substation transformer A32 full load current and inrush current must be taken into account in selecting the setpoints for the overcurrent device associated with the transformer breaker A36. The transformer inrush current is indicated in FIG. 2 by the points designated with the reference numeral A68, while the full load current is indicated by the :...ne segment identified with the reference numeral A69. Thus, the setpoints are selected such that the overcurrent device associated with the transformer breaker A36 as well as the tie breaker A40, so as to prevent spurious tripping of the unit substation A30 during normal operating conditions and additionally avoid tripping for a fault downstream of the feeder breaker A42 unless there is a failure of the circuit breaker A42 or its associated overcurrent device. Thus, the time-current characteristics for the transformer breaker A36 as well as the tie breaker A42 are illustrated in FIG. 2 and identified with the reference numeral A70. As shown. the time-current characteristics for the overcurrent devices for these circuit breakers, generally allow for selectivity relative to the downstream circuit breaker protective devices.
The considerations for the overcurrent charac-teristics for the overcurrent device A54 for the 4.16 kV
feeder breaker A26 must take into account the normal full load current of the unit substation A30 including as well as any loads connected to the tap A71 and additionally be coordinated with the time-current characteristics of the fuse A28. In such an application, as illustrated in FIG. 2, 24 ~4,2i8-_-2 the fuse' A28 acts as backup protection for the °eeder breaker A26 as well as provide protection for the priTary winding of the unit substation transformer A32. Addition-ally, it is desirable for the feeder breaker A26 to trio prior to the fuse A28. Thus, the curve illustrated with t~e reference numeral A72, illustrates the time-current characteristics of the overcurrent device A54 associated with the feeder breaker A26. The curve A74 illustrated the time-current characteristics of the fuse A28.
As shown in the lower portion of FIG. 2, there is slight overlapping of the time-current characteristics of the fuse curve A74 and the feeder breaker curve A72.
Accordingly, some selectivity between the feeder breaker A26 and the fuse A28 is lost for relatively large magnitude faults. In the upper area of the curve, there is a significant disparity of the characteristics of the fuse curve A74 relative to a portion of the time-current curve A72 for the feeder breaker A26 known as the long time delay portion. As such it is relatively difficult to coordinate time-current characteristics of various known overcurrent trip units with devices, such as fuses.
~~~~u ataze ur1 1nQ UeylCe The time-current characteristics for an exemplary adjustable solid state tripping device, such as disclosed in U.S. Patent No. 4,827,369, are illustrated in FIG. 3. Such time-current characteristics are normally illustrated on a logarithmic scale whereby the horizontal axis, identified with the reference numeral A76, relates to electrical current in amperes and the vertical axis, identified with the reference numeral A78, relates to time in seconds.

_ 25 ~~~~~~~ S~,ZIa-.-2 ~~The bottom-most portion of the curve identiL~ed with the reference numeral A80 illustrates what is known as the instantaneous portion of the curve. The instantaneous portion of the curve is used during conditions where the electrical current magnitude is relatively large, for example, during a short circuit condition. During such a short circuit condition, the instantaneous portion A80 causes the interrupting device to be tripped, for example, in one cycle or less. The electrical current magnitude at which the instantaneous portion A80 becomes active is adjustable. Thus. the instantaneous portion can be adjusted relative to the horizontal axis A76 as indicated by the reference arrow A77.
The central portion of the curve illustrated by the bracket, identified with the reference numeral A82 relates to the short time delay portion of the curve. Both .inverse time characteristics and fixed time characteristics are known for the short time delay portion A82. More specifically, with reference to FIG. 3, fixed time characteristic A84 is illustrated in solid line while an inverse time portion curve A86 is illustrated in phantom.
The short time delay portion A82 is adjustable.
More specifically, the minimum electrical current magnitude at which the short time portion A82 becomes active is known as the short time delay pickup (SDPU) indicated by the reference numeral A88. The SDPU is adjustable relative to the horizontal axis A76 as indicated by the arrow A90. The time at which the short time delay portion is also adjustable relative to the vertical axis A78 as indicated by the arrow A92.
Although the SDPU and time settings are adjustable the characteristics (e. g., function) of a short time delay portion A82 have heretofore been known to be 26 » ,2_9-.-2 fixed. Such characteristics are generally inverse.
such inverse characteristics, t!:e tripping time is gene=al:.y inversely .elated to the magnitude of the elect==ca:
current. Thus, relatively larger electrical currents are tripped in relatively short time periods and reiat_vely smaller magnitude electrical currents in relatively longe:
time periods.
The portion of the curve identified with t:~e reference numeral A94 relates to the long time delay por-tion. This portion A94 is also inverse and general_y follows an I2t characteristic. Similar to the short time delay portion A82, the setpoints can be varied to adjust t::e time and electrical current magnitude at which this portion of the time-current characteristics become active. More specifically, the minimum electrical current at which the long time delay portion A94 becomes active is known as the long time delay pickup (LDPU) A96. This LOPU can thus be adjusted relative to the horizontal axis A76 as indicated by the arrow A98. The time at which the long time delay portion A94 becomes active can be varied to allow the long time delay characteristic to be moved upwardly and downwardly relative to the vertical axis A78 as indicated by the arrow A100.
Although the LDPU A96 and time setpoints can be varied as indicated by the arrows A98 and A100, the char-acteristic (e. g.. I2t function) has not heretofore been known to be adjustable which as described above, can cause difficulty in attaining good coordination with very inverse characteristics such as the fuse A28 Variable Long Time Delay Characteristics An aspect of the invention relates to the ability to vary the long time delay characteristics of the time-current curve for a solid state tripping device.

27 Z ~ O ~ ;~ ~ 54, 2i8-~-2 More specifically, as heretofore stated, known overcurrent devices including solid state tripping devices, for example, as described and illustrated in detail in U.S. Patent No.
4,827,369, are provided with long time delay characteristics which are fixed and generally follow an I2t function.
Although the magnitude and time at which such a characteristic becomes active can be adjusted as discussed above, the characteristic itself (e.g.. I2t function) is fixed and thus not adjustable.
The solid state tripping device in accordance with the present invention, provides for adjustability of the characteristic of the long time delay portion A94 of a solid state tripping device in order to provide better coordination and therefore better selectivity with other overcurrent devices in an electrical distribution system.
More specifically, referring back to FIG. 2, the charac-teristics of the curve of the fuse curve A74 are very inverse. Thus, as heretofore stated, it is rather difficult to coordinate an overcurrent device with the fuse A28 because of the disparity in the characteristics of the fuse curve A74 relative to the characteristics of various overcurrent devices as discussed above. Accordingly, in accordance with the present invention, a solid state trip-ping device is provided which includes a long time delay portion A94 having adjustable characteristics. More specifically, the long time delay portion A94 of the time-current characteristics not only allows the time and current at which the long time delay portion of the characteristic becomes active, as in U.S. Patent No. 4,827,369, but also allows the actual characteristics of this portion of the curve to be varied. For example. as described and illustrated hereinbelow, the long time delay characteristics A94 are adjustable to enable an operator to select between a 2e z~~~~~~ '4-2'-x-=-2 plurality~of characteristics, for example. a fixed tine (FLAT) characteristic, It, Izt and I4t. However, it should be understood by those of ordinary skill in the art that the principles of the invention are not related to any particular characteristics or any number of available characteristic selections.
Graphical illustrations of the adjustable char-acteristics in accordance with the present invention are illustrated in FIGS. 4, S and 6 with exemplary FLAT
characteristics shown in brackets in FIG. 4 and identified with the reference numeral A103. Each of the figures illustrate the time-current characteristics of the long time delay portion on a logarithmic scale whereby the horizontal axis relates to the current in amperes and the vertical axis relates to the time in seconds. More specifically, FIG. 4 illustrates a long time delay portion A102 which has a characteristic representative of the product of the current and the time; It. As mentioned above, the FLAT
characteristic is also illustrated in FIG. 4 and identified with the reference numeral A103. FIG. S illustrates a long time delay characteristic A104 having a characteristic which represents the product of the square of the current and the time; I2t. FIG. 6 represents a long time delay characteristic A106 which represents the product of the fourth power of the current and the time; I4t.
The characteristics of the long time portions A92, A94 and A96 all relate to the product of the time with the current to a predetermined factor. For example. the factors described and illustrated above are 1, 2 and 4. On a logarithmic scale for the example above, varying the power of the current simply relates to varying the slope of the characteristics of a straight line function and illustrated graphically in FIGS. 4-6.

29 54,218-I-2 ~ By providing adustment of2~t~~~.~~ of the long time characteristics of an overcurrent device, better coordination is possible for providing better selectivity and better electrical overcurrent protection relative to various devices including the fuse A28 illustrated in FIG.
1. More specifically, with reference to FIG. 2, the upper portion of the fuse characteristic A74 is very inverse and thus has a relatively steep slope. With known overcurrent devices having fixed characteristics which follow a general I2t characteristic, it is rather difficult to provide optimum coordination with a device having such an inverse characteristic 'such as the fuse A28. The solid state tripping device in accordance with the present invention, solves this problem by providing for adjustable tripping characteristics. Thus, in an application such as the application illustrated in FIG. 2, an I4t characteristic could be selected to provide better coordination between the 4.16 kV circuit breaker and the fuse A26.
Long Delay And Short Delay Coordination In Circuit Breaker Trip Units As mentioned above, solid state tripping units for use in molded case circuit breakers, metal clad switchgear and the like, include tripping curves having fixed characteristics wherein the time and the magnitude of electrical current are adjustable. More specifically, as described above and further described in detail in U.S.
Patent No. 4,827,369, such overcurrent characteristics normally include an instantaneous portion A80, a short time delay portion A82 and a long time delay portion A94 (FIG.
2). In order to prevent any overlapping in the short time delay portion A82 and the long time delay portion A94 of the time-current curves, the adjustment ranges for these portions are selected such that the short time delay portion A82 and the long time delay portion A94 of the curves will - 30 ~~~~?~~ 54,2_8-_-2 not overlap throug.out the entire adjustment range oL t?:e trip unit. As such, the coordination of such devices with other overcurrent devices in electrical distribution systems is rather limited.
In accordance with another aspect of the present invention, the adjustment ranges for the long time and short time delay portions of the trip unit are provided with relatively larger adjustment ranges than known solid state trip units in order to provide the capability to provide better coordination in an electrical distribution system.
However, by providing such relatively wider adjustment ranges. coupled with the selectable long time characteristics described above, overlapping of the short time and long time delay characteristics can result. More specifically, FIG. 7 illustrates a situation where the long time delay portion A94 and the short time delay portion A82 of a time-current curve or overcurrent device having relatively wider adjustment ranges have been adjusted to cause overlapping. In particular. with reference to FIG. 7, an electrical current having a magnitude I1 designated in FIG. 7 with the reference numeral A108 would pause a short time delay portion A82 of the trip unit to trip the interrupting device at a predetermined time period identified as Tl. However, a relatively lower magnitude current I2, identified as A109, would cause the long time delay pickup to become active to trip the interrupting device in a relatively shorter time T2, thus resulting in an overlap of the short time delay portion A82 and the long time delay portion A94 of the time-current curve perhaps causing a loss of selectivity in an electrical distribution system.
Typically, the long time delay portion A94 of a time-current curve is adapted to become active after the 31 2~~~~~~~ 54.218-.-2 short tide delay portion A82 to allow relatively lower magnitude electrical currents to flow in the electrical distribution system for relatively longer times without causing a trip of the interrupting device. This is done to prevent tripping of any interrupting device due to transier.~
overcurrent conditions, for example, bus loading, voltage fluctuations in the system or the like. Accordingly, t:~e long time delay portion A94 normally allows for relatively lower magnitude electrical currents for relatively larger amounts of time.
The short time delay portion A82 of the curve is normally intended to become active prior to the long time delay portion A94. The short time delay portion A82 normally becomes active at relatively larger magnitude electrical currents in relatively shorter times than the long time delay portion A94.
Accordingly, it is necessary that the long time and short time delay portions A94 and A82, respectively, be coordinated. As mentioned above, one aspect of the present invention relates to providing relatively wider adjustment ranges of the short time delay portion A82 and the long time delay portion A94 of the time-current curves. Accordingly, in order to provide such relatively wide adjustment ranges for the short time delay portion A82 and the long time delay portion A94 of the time-current curve, the solid state trip unit in accordance with the present invention includes provisions (described below) for monitoring the short time delay timers as well as the long time delay timers to prevent the long time delay portion A94 of the time-current curve from tripping the interrupting device prior to the short time delay portion A82.
By preventing the long time delay portion A94 from tripping the interrupting device prior to the short time 32 . ~~~~~ '3 54~218-.-2 delay portion A82, the overlap is thus eliminated. More specifically, with reference to FIG. 8, the short time delay portion is indicated with the reference numeral A112 while the long time delay portion is identified with the reference numeral A114. By preventing the short time delay portion from tripping the interrupting device prior to the long time delay portion, the characteristic illustrated in FIG. 7 is modified to that shown in FIG. 8, thus eliminating overlapping, while at the same time. providing relative:.y wider adjustment ranges for the short time delay portion A112 and the long time delay portion All4 of the time-current curve. .
Instantaneous Fault Detection Method As previously indicated, the instantaneous portion A80 of a time-current curve is normally utilized to protect an electrical distribution system subject to a relatively large overcurrent, such as a short circuit current. As such, it is necessary to trip the immediately upstream circuit breaker virtually instantaneously to prevent damage to the electrical distribution system. In particular, it is necessary to trip an interrupting device before its withstand capability is exceeded. The withstand capability relates to the mechanical forces that an interrupting device can withstand without failure due to the excessive short circuit current.
The line current flowing through an interrupting device is normally sensed by one or more current trans-formers, such as the current transformers A52 illustrated in FIG. 1. During relatively large overcurrent conditions, such as a short circuit condition, such-current transformers A52 are known to become saturated. During saturation, the output waveform on the secondary of the current transformer is similar to that illustrated in FIG. 9, which, as shown, 33 ~~~~~~~ '4,218-~-2 includes.a plurality of relatively steep and narrow spi:{e portions A116. As such, such relatively steep and na-row spike portions A116 can cause problems and even result in the solid state tripping device not tripping in time to prevent damage to the system.
More specifically, known solid state tripping devices which utilize a microprocessor, normally sample the line current a specific number of times during each cycle of the waveform. Problems arise when the sampling ~i:~e interval is relatively longer than the time period of the relatively narrow and steep spike portions All6 of a saturated current transformer waveform. Consequently, in such an application, the instantaneous fault level may not be undetected in time to prevent damage to the electrical distribution system.
Some known devices utilize a modified sampling technique in order to cover different points of the wave-form. However, in such an application where the current transformers are saturated, such a sampling technique is relatively complex and also may still not be able to allow detection of the instantaneous saturated current transformer peak current in time to prevent damage to the electrical distribution system.
Analog circuits for monitoring the secondary electrical current of the current transformers have not been known to be used because of the interface problems with microprocessor-based systems. More specifically, such microprocessor-based systems normally include a adjustable instantaneous trip level with a range between 1 to 28 times a nominal 5 ampere current. The instantaneous trip level is generally programmed by the end user, thus making it difficult, if not impossible, to interface with an analog detection circuit.

~~~~~~3 3~ 54,218-_-2 An important aspect of the present invention relates to the use of an analog circuit for determining the highest negative peak level of the instantaneous phase current and summing it with a positive current which is proportional to the setpoint of the instantaneous trip level. The positive current set by the microprocessor is a pulse width modulated output wherein the pulse width is proportional to the desired instantaneous setpoint. The summed current is then compared in a comparator to determine if the instantaneous phase current is greater than the instantaneous setpoint. If so, the comparator produces an interrupt to the microprocessor. In response to the interrupt. the microprocessor samples the current waveform to measure the final value and initiate a trip command to the interrupting device.
The analog circuitry consists of a plurality of OR
diodes connected diodes to select the highest negative current available from the current transformers A52. For a three phase system, the analog circuitry would include three diodes connected in an OR configuration to select the highest peak negative electrical current value. This negative value is then summed with the positive electrical current generated by the microprocessor.
By utilizing an analog circuit for monitoring the secondary currents of the current transformers, the relatively steep spike portions A116 will be appropriately sensed and thus compared with the reference setpoint in order to initiate a trip command to the interrupting device if the setpoint is exceeded. In this way, the danger of utilizing a sampling scheme which could miss the relatively steep and narrow spike portion A116 of the current transformer secondary current waveform are virtually eliminated.

'- - 35 X4,218-~-2 Detailed Description Of The Embodiment The electrical circuitry for incorporating the solid state tripping device as described above is illus-trated in FIGS. 10-16. Such electrical circuitry includes a custom Westinghouse SURE CHIP PLUS microcontrolier illustrated in FIGS. 32-113 and described below. The microcontroller includes a microprocessor, an on-board A/D
converter, on-board comparators, as well as a plurality oL
input/output devices. The software control for the micro-processor is illustrated in FIGS. 17-31.
With reference to FIGS. 10-16, the alphanumeric characters within the circles are used to denote a contin-uation of the schematic on another figure. For example, a circle containing "P11" is used to designate a connection on another figure having a circle containing "P110" Also.
multiple bits of a bus or port are designated, for example, as PC[7...0] which denotes bits 0 through 7 of port C.
Similarly, individual bits are designated, for example, as PC[7,5,1] which denotes bits 7, 5 and 1 of port C.
Power supplies are well within the ordinary skill in the art. As such, the power supply inputs for the components illustrated in FIGS. 10-16 do not form a part of the present invention. Accordingly, power supply inputs to the various components are merely illustrated and not described. Similarly, ground inputs for the various components illustrated in FIGS. 10-16 are merely illustrated.
Referring first to FIG. 16, a Westinghouse custom SURE CHIP PLUS microcontroller is illustrated and identified with the reference numeral D20. As previously mentioned, the microcontroller D20 is illustrated in FIGS. 32-113 and described below.

!~
-- _ 36 ~ ~ ~ ~ ~ ~ ~ S4, 218-I-2 ~' The microcontroller D20 includes four parallel 8 bit input/output ports: PA, PH, PC and PD. The port PD(7...0] is used to form an eight bit multiplexed address/data bus, identified with the reference numeral D22. The port PB[6...0] is used to form a high address bus, identified with the reference numeral D24. Port PC is used for trip and alarm functions as well as for address decoding. More specifically, bits PC(1,0] are utilized °or instantaneous trip and overcurrent trip commands.
respectively. Bits PC [ 2, 3 ] are used for instantaneous ar.d overcurrent alarm functions, respectively. Bits PC(7...4;
are used for input/output address decoding defining the address lines IOAO, IOA1, IOA2 and IOA3.
Port D forms an input/output bus, identified with the reference numeral D26. As will be described in more detail below, the input/output bus D26 is utilized for the user interface as well as for a supervisory interface as will be described below.
The solid state trip unit also includes a pre-selected amount of read only memory (ROM), identified with the reference numeral D30, for program instructions and a preselected amount of random access memory (RAM), identified with the reference numeral D32, for data. A nominal 32K ROM
is selected for program instructions while a nominal 8K RAM
is selected for data. However, it will be appreciated by those of ordinary skill in the art that the principles of the invention are not limited by the sizes selected for the ROM D30 and the RAM D32.
The 32K ROM D30 requires 14 address lines.
Accordingly, a low address bus, identified with the refer-ence numeral D34, and the high address bus D24, (forming a total of 14 address lines) are applied to the address inputs AO through A14 of the ROM D30.

_ 3, 2~~~~~a s4,21a-_-2 -The high address bus D24 is provided by porn. D on the microcontroller D20 as discussed above. The low address bus D34 is derived from the multiplexed address/data bus D22. More specifically, the multiplexed address/data bus AD[7...0] D22 is applied to the data inputs DI[7...0] of an 8 bit address latch D36. The data output bits DO[7...0] of the address latch D36 form the low address bus D34.
The address latch D36 is under the control of an address latch enable signal ALE, available at the micro-controller D20. Thus, once an address is placed on the multiplexed address/data bus D22 by the microcontroller D20, this address is latched by the address latch D36 under the control of the address latch enable signal ALE, which in turn, is applied to the latch enable (LE) input on the address latch D36.
The address latch D36 is continuously selected by tying the chip select input (O~~ to ground by way of a resistor D38. Similarly, the ROM D30 is continuously selected by tying the select input; ~E to ground by way of the resistor D40.
The output pins DO[7...0] of the address latch D36 form the low address bus A[7...0] D34. The low address bus A[7...0] D34 is applied to the address inputs A[7...0] of the ROM D30 along with the high address bus A[14...8] D24 which is applied to the ROM address inputs A[14...8) to address the 32 kilobyte ROM D30 which, as indicated previously, contains the program instructions which will be described below.
The output pins 0[7...0] of the ROM D30 are applied to the multiplexed address/data bus D22 in order to allow the microcontroller D20 and, in particular. the microprocessor within the microcontroller D20 to access program instructions. The operation of the ROM D30 is under 38 2 ~ ~ ~ ~ ~ ~ 54, 218-.-2 the contr'bl of a program select enable signal (PSy~) .
available at the output of the microcontroller D20.
The 8 kilobyte RAM D32 only requires 12 address lines. Accordingly. the low address bus A(7...0] D34 is applied to the input terminals A[0...7] of the RAM D32. A
portion of the high address bus D24 (e.g., A[12...7 0 is applied to the address inputs A(12...7] of the RAM D32. The output of the 8 kilobyte RAM, 0[7...0], is tied to the multiplexed address/data bus D22 in order to allow the microcontroller D20 to access the RAM D30.
Analog values, such as electrical current values and in particular the phase and ground currents are digi-tized by the microcontroller D20 and stored in the RAM
D32. Read and write functions of the RAM D32 are under the control of the microcontroller D20. More particularly, the RAM D32 may be read when the microcontroller pulls its read output (R~ low which is tied to the operate enable input ~OE~ of the RAM D32. The microcontroller D20 can write to the RAM D32 by pulling it write output (W~ low which is tied to the write enable input of the RAM D32.
Input/output address bits IOA[3...0], available at the microcontroller D20, facilitate addressing of a plurality of user and supervisory control interfaces. More specifically, the input/output address bits IOA[3...0], are applied to two 3 x 8 address decoders D44 and D46. In particular, the input/output address bits IOA[2...0] are applied to the A. B, C inputs of each of the 3 x 8 address decoders D44 and 046. The address bit IOA[3] is applied to chip select inputs CSl and CS2 , respectively of the address decoders D44 and D46 by way of a resistor D48. The pSgN
signal, available from a microcontroller, is applied to the chip select inputs CS3 of the address decoders D44 and D46. The chip select input ~S2 of the address decoder D44 - 39 ~~~~~~J 5'~~218-I-2 is grounded while the chip select input CS1 of the address decoder D46 is tied to a S volt source by way of a resistor D50.
The address decoders D44 amd D46 are thus used to decode the input/output address bits IOA[3...OJ to provide twelve different address decode signals. (Four of t::ese select signals are unused.) These address decode signals are identified as C~,O through CE11 used for various use:
and supervisory control interfaces as discussed below. More specifically, referring to FIG. 10, the address decode signals CEO, CE1, C~2, and CE3 are applied to a clock input (CLK) of four 8 bit registers D52, D54, D56 and D58, respectively, while the input/output bus D26 is applied to the data input pins DI[7...0]. Data on the input/output bus D26 is clocked into the registers D52, D54, D56 and D58 under the control of the chip enable address decode signals CEO, CE1, CE2, and CE3 . The registers D52, D54 and D56 and D58 are reset by the microcontroller reset signal. The output pins DO[7...0] of the registers D52, D54 and D56 and D58 are used to drive various light emitting diodes (LED's), generally identified with the reference numeral D60 by way of various current limiting resistors generally identified with the reference numeral D62.
In addition to controlling the LED's D60, the registers D52 and D58 are also used for interlocking func-tions. More particularly, the signals, GND-ZONE-OUT and PHASE-ZONE-OUT, available at pin D07 on the registers D52 and D58, respectively, are used for interlocking with an upstream circuit breaker in the electrical distribution system to allow for selective tripping. In particular, the GND-ZONE-OUT and PHASE-ZONE-OUT signals are used to indicate whether the ground and phase overcurrent options have been selected. In particular, these signals are used to prevent -- _ ~o ~~.~~2~3 ~.~,21a-.-z tripping df an upstream breaker during the long time delay portion of the time-current current curve if these options were selected on the downstream trip unit in order to prevent the upstream circuit breaker for tripping on a long time delay. If the options are not selected, then the upstream circuit breaker will provide the long time delay protection not selected for the downstream trip unit.
These signals are applied to the registers D52 and D58 by way of the input/output bus D26 and are available at the' D07 outputs of the registers D52 and D58, respectively. These signals, in turn, are applied to two optocoupler networks generally identified with the reference numerals D69 and D71 (FIG. 11). The output of the optocoupler networks D69 and D71, in turn. are applied to terminals 8 and 10 of the user interface terminal block D68 for interlocking with an upstream circuit breaker.
Similarly. signals from downstream circuit breakers, identified as PHASE-ZONE-IN and GND-ZONE-IN from a downstream circuit breaker are applied to terminals 9 and 11 of the user interface terminal block D68. These signals are applied tb the signal conditioning and switching network generally identified with the reference numeral D69 and, in turn, to optocoupler networks D73 and D75, respectively.
The output of the optocouplers D73 and D75 are then applied to the input data bus D26 to inform the system that the downstream circuit breaker has been configured for both phase and ground current over protection in order to coordinate long time delay tripping functions therewith.
The outputs of the optocouplers D73 and D75 are applied to an 8 bit register D106 (FIG. 10). selectable by the address decode signal G~ , and applied to the input/output bus D26.
The register D58 is also used as an interface for a PONI network. The PONI network relates to a serial port 41 54W=8-_-2 interface for use with ~Jestir.ghouse's INCOM communicatio:.s network. The PONI interface is described in U.S. Patent No.

5,007,013.
An example of the IVCOM network is described in U.S. Patent No.
4,653,073_ In particular, the inputs to the PONI interface, identified as PONI-CLK and PONI-R/W, available at the DOS
and D06 pins of the register D58 are applied to a PONI
interface terminal block D76 (FIG. 11). The output signals from the PONI interface, PONI-INT and PONI-DATA, adapted to be connected to the PONI interface terminal block D76 are applied to the input/output bus D26 by way of the register D106.
The INCOM network, referred to above, is a communications network that allows the electrical inter-rupting device to communicate with a remote communications device as described in the aforementioned '073 patent. The SURE CHIP PLUS microcontroller D20 as discussed in detail enables the electrical interrupting device to communicate with the INCOM network. Hy enabling the electrical communications device to communicate with the INCOM network, the function (e. g., slope) and setpoints can be varied remotely.
The interface with the INCOM network is by way of a terminal block D77 (FIG. 14a). In particular, the INCOM
send and receive signals TX and RX are connected to terminals 1 and 2 of the terminal block 077 to connect the electrical interrupting device to an INCOM communications network as described in detail in the '073 patent.
Hoth the RX and TX signals are conditioned by signal conditioning circuitry shown within the dashed box 42 . ~~~~~~~ ~~,,213-_-2 identified with the reference numeral D79. The RX and signals as well as a common line are connected to the SURE
CHIP PLUS microcontroller D20 as shown in FIG. 16.
A 5 bit seven segment display (FIG. 12) is also locally mounted relative to the solid state trip device.
The seven segment display is adapted to display values of various parameters including the instantaneous phase and ground currents. The display generally includes five sever.
segment digits generally identified with the reference numeral D80. Each of the seven segment display elements are driven by a seven segment display driver, generally identified with the reference numeral D82, by way of a plurality of resistors generally identified with the reference numerals D84. The input/output bus D26 is applied to the input terminals A, B, C, D of the display drivers D82. The chip and decoded address chip enable signals CE7, CE9 and CE9 are applied to the enable input rte, oL
V
the display drivers D82.
In addition to the display, the trip unit also includes an alphanumeric display generally identified with the reference numeral D86, utilized during a calibration mode. The alphanumeric display D86 is selected by the decoded address inputs CEIO and CE11 ' More specifically, the decoded address input CE10 and CE11 are used to select the display D86. The input/output bus D26 is applied to a register D107.
The user interface also includes a plurality of membrane switches for selecting setpoints. viewing the setpoints on the display and testing the trip. unit as illustrated in FIG. 15. These membrane switches are generally identified with the reference numeral D88 and are applied to a register D90 (FIG. 10) by way of a plurality of resistors D92 for interfacing with the microcontroller - 43 ~ ~, ~ ~ ~ ~ ~ 84. 2i8-:-2 D20. More specifically, the membrane switches D88 together with the resistors D92 which form a voltage divider with the S volt source, function to change the voltage level and consequently the logic level applied to the register D90.
The output of the register D90 is applied to the IO bus D26. The register D90 is selected by the address decode select signal CES.
A user interface register D94 (FIG. 10) is also provided for programming the current transformer ratio as well as select options, such as a discriminator option or an auto reset option. Such data is programmed by the user by way of a gang switch D96 (FIG. 11). The gang switch D96 includes a plurality, for example 10, single pole single throw switches which are used to interface with the system. The switches D96 are applied to the register D94 by way of a plurality of pull up resistors D98. The output of the register D94 is applied to the IO bus D26. The register D94 is selected by the address decode chip enable signal CE4 ' In addition, the trip unit is adapted to provide a plurality of relay contact outputs for interfacing with a circuit breaker trip coil and customer remote annunciator.
More specifically, a CLOSE signal, available at register D54 and an instantaneous trip and overcurrent trip signals as well as alarm signals, available at the microcontroller D2 are applied to power transistors, generally identified with the reference numeral D110, which, in turn, are applied to a plurality of relays identified with the reference numeral D112. Each relay D112. Each relay D112 includes a single pole double throw contact D114. The relay contact outputs D114 are then applied to the customer interface terminal block D115 for use by the customer.

,... _ 44 2~~~?~~ ;,x,2,8_:_2 The system includes a plurality of auxil=ary current transformers for reducing the CT output currents to an acceptable level. More specifically, the primary current transformers provided on the particular interrupting device upon which the trip unit is used, are wired to a ter:ninai block D116 (FIG. 13) which, in turn, is wired to the auxiliary transformers D118. The secondary of the auxiliary transformers is wired to a terminal block D120. The terminal block D120 are wired to a CT module interface terminal block D122 (FIG. 14).
The outputs of the CT's are rectified by a plurality of rectifiers generally identified with the reference numeral D124. The rectified output is then applied to a monitoring circuit, generally identified with the reference numeral D126 (FIG. 16). More specifically.
the rectified phase current outputs IA, IB and IC are applied to a diode OR circuit, generally identified with the' reference numeral D126. The diode OR circuit D126 contains three diodes whose anodes are connected in common and whose cathodes are connected to the rectified phase currents IA, IH and IC.
Such a configuration can easily and continuously select the largest negative phase current even when the current transformers are saturated. This negative phase current is summed with a positive current developed by the microcontroller D20 which is related to the customer selected setpoint for the instantaneous/overcurrent setting as discussed below. This combined value is then applied to a comparator within the microcontroller D20 which will interrupt the microprocessor within the microcontroller D20 whenever the phase currents exceed the preselected setpoint.
The ground current is likewise monitored by the system. The ground current IG causes a voltage drop across 45 ~ ~ ~ ~ ~ ~ ~ 54, 2=8-~-2 a resistor D127. As long as this voltage is lower than tre voltage formed by a voltage divider network, a diode will conduct and indicate the instantaneous ground current which, in turn, is applied to the microcontroller D20.
System Configuration The trip unit in accordance with the present invention is configured by way of the membrane switches D88 (F~G. 15) and the gang switch D96 (FIG. 11). In particular, the gang switch D96 is a user interface to indicate to the system the particular current transformer ratio selected on the circuit breaker associated with the overcurrent trip device. The five switch contacts are utilized for the CT
ratio selection. This allows up to 32 various CT ratios to be programmed into the system.
Similarly, two switch contacts are used to program the ground current transformer ratio. These two switch contacts allow up to four different ratios to be selected.
In addition, the gang switch contacts D96 are also used for various options. For example, one gang switch contact D96 may be used to program an auto reset option.
The other gang switch D96 can be used to allow or block downloading of setpoints via INCOM.
The membrane switches generally, identified with the reference numeral D88, provide for a user interface with the trip unit. In particular, setpoints are programmed into the trip unit by depressing the membrane switch identified as ACTIVATE PROGRAM MODE. The setpoints for the various phase and ground overcurrent and instantaneous tripping functions including the slopes for the long time delay portion may be selected by depressing the membrane switch identified as SELECT. By depressing the SELECT setpoints switch various LED's D60 (FIG. 10) relative to various setpoints including long time delay pick-up, short time 46 ~~,2;8-_-2 delay pick-up, instantaneous delay tines and s'_ope wi sequentially be illuminated. Once the LED D60 on the ~--io unit adjacent the particular the desired setDOint is illuminated, the operator by way of the LOWERIRAISE'membrane switches can page through a number of preprogrammed setpoints until the desired setpoint is visible on t:e alphanumeric display D86. After the desired setpoint is reached the membrane switch identified as SAVE SETPOINTS is depressed which, in turn, stores the selected setpoint. Iz the event of an error, the operator can simply depress t::e membrane switch identified as RESET to exit the programming mode. The system also has capability for viewing all of the setpoints once they have been saved. In order to view the setpoints, the membrane switch identified VIEW SETPOINTS is depressed.
The system also has capability for testing. In order to test the unit, a membrane switch labeled ACTIVATE
TEST MODE is depressed. The word "TEST" then appears on the alphanumeric display D86. The operator can then press the SELECT SETPOINTS switch to select the desired test. Once the test is selected, the operator can press the RAISE LOWER
switch to increase or clear the level of current to run the test. Subsequently, the membrane switch labeled TEST is depressed. This will cause the unit to initiate a trip which is indicated by the LED's D60.
There is also a metering function, where phase currents IA, IH, IC are displayed along with IG ground current IG. To step between these displays the operator presses the SELECT switch. The currents are displayed on the display D80.
The user interface includes a front panel F20 illustrated in Figure 114. This front panel includes a visual representation F22 of the adjustable time versus .- ~ 47 ~~~~~~~ 54~218'1-2 current 'tripping characteristic representative of the protection curve shown in Figure 3. Associated with the protection curve F22 are the LcDs D60 illustrated in Figure 108. As indicated, lettering adjacent each LED identifies the portion of the curve for which the LED serves as an indicator. For instance, in tre center of the protection curve F22, the LED D60 adjacent the wording SHORT DELAY and SETTING is an indicator for the pickup for the SHORT
DELAY. Each of these LEDs is a two color device which is controlled by one of the registers D52-D58 shown in Figure 10B. The red portion of the LED is energized in the trip mode, and the green LED indicates that the unit is in the programming or view mode. In the trip mode. the appropriate LED such as the LONG DELAY SETTING LED flashes when the LONG
DELAY pickup current is exceeded. When the long delay times out, this LED emits a steady red light providing an indication of the condition which caused the trip. In the program mode, the appropriate LED is illuminated continuously in a viewing mode when the various parameter settings are only being viewed, and flashes in the adjust mode in which the parameters settings can be modified. The front panel also includes the OPERATIONAL LED, which flashes green at one second intervals to indicate that the unit is functioning properly and flashes red when there is a problem. The HIGH LOAD LED flashes red When the current exceeds 85~ of the long delay pickup current, and then comes on continuously if this condition persists for a predetermined time period. The green HIGH LOAD LED is on continuously in the view mode and flashes for programming of the high load time interval. The COMMUNICATIONS TRIP LED
indicates when a trip was ordered remotely.
The AMP DEMAND and phase~current IA, I8, I~, IG, LEDs are also located on the front panel F20 along with the .... - 48 ~ ~ ~ ~ ~ ~ ~ 54, 218-i-2 alpha-numeric display D80 (see Fig. 12A) which displays the current at trip.
Also mounted on the front panel are the membrane switches D88 shown in Figure 15. Associated with the program and test membrane switches is the alpha-numeric display D86 (see Fig. 12B).
The display panel F20 is designed to provide a clear unambiguous interface for the user. The visual representation F22 of the protection curve assists the user in programming the parameters. The LEDs associated with this representation of the protection curve provide a readily understood association of the parameter which appears on the display D86 with the function of the curve to which it pertains.
Another feature of the front panel F 20 is that the membrane buttons associated with the programming mode namely: PROGRAM MODE ON/OFF, SELECT SETTINGS. AND SAVE
SETTINGS are all color coded in one color, such as blue in the exemplary trip unit, while the membrane switches associated with the test mode, namely: TEST MODE ON/OFF, SELECT TESTS. AND TEST are all color coded in another color, which is yellow in the example. This color coding reduces confusion and mistakes in operating in these two modes. The common switches used in both of these modes namely: the RAISE and LOWER switches are a neutral color such as white.
The front panel F20 can also be used for inserting the address of the trip unit for identification of the unit on the communications network. This change is effected by simultaneously depressing the SELECT TESTS and TEST membrane switches which will bring up a four digit display on the alpha-numeric display 086. The first digit will be an Fi or L indicating a high or low baud rate for the communications. In the exemplary system, the high and low .~. - 49 ~ ~ ~ ~ ~ 64. 218-r-2 rates are~ 9600 and 1200 baud, respectively. The remaining three digits are the selectable unit address. The address and baud rate can be adjusted by cycling through the digits and utilizing the RAISE and LOWER push-buttons.
Program Control The program control for the trip unit in accor-dance with the present invention is illustrated by way of a flow chart in FIGS. 17-31. As previously mentioned, the program instructions are contained in the 32K ROM D30. In addition to the program instructions. the ROM D30 may also contain look-up tables for the various time-current char-acteristics for phase and ground overcurrents.
As previously mentioned, these phase currents IA, IH, IC. as .well as the ground current IG are applied to analog inputs MUXO. MUX1, MUX2 and MUX3 of the custom microcontroller D20. As discussed below, these analog in-puts are arranged and applied to an on-board A/D
converter. The digitized values are then stored in the RAM
D32.
The program control consists of a main program illustrated in FIGS. 17-23; a phase current long time delay subroutine illustrated in FIGS. 24 and 25, a ground current long time delay subroutine illustrated in FIGS. 26 and 27; a short time delay subroutine illustrated in FIG. 28, an instantaneous subroutine illustrated in FIG. 29 and a turbo instantaneous subroutine illustrated in FIGS. 30 and 31.
Program control. further consists of a front panel routine illustrated in Figures 115-118, a trip mode subroutine illustrated in Figures 119-120, a view mode subroutine illustrated in 121, an LED lighting subroutine illustrated in Figure 122, and a communication address mode routine shown in Figure 123.

2~~~?~o --- ~ so s4, zla-~-z ~'In general, the system operates on a continuous cycle of 6sK samples. More specifically, the phase currents (IA, IB and IC) and ground current (IG) are sampled every one-third cycle or 4.7 milliseconds. Two samples are used for instantaneous protection, while eight samples are used for short-time delay protection; sixteen samples are used for long time delay protection. In addition to overcurrent protection, the 256 samples are used for metering and 65K
samples to peak demand current in five minute intervals.
Initially, a sample timer, identified as PTIMER, is loaded in step P20 to cause the phase and ground currents. IA, IB, IC and IG to be sampled approximately every one-third cycle or 4.7 milliseconds. After the sample timer PTIMER is loaded, the first sample is taken in step P22 and the sample counter PTIMER is incremented in step P24. After the sample counter PTIMER has been incremented, the system checks in step P26 to determine if a time flag has been set. If so, the alphanumeric display D86 displays the elapsed time from which the fault occurred. If not, the system bypasses step P28. Subsequently, a trip time offset counter is incremented in step P30. The trip time offset counter is incremented every time a FAST STATUS command is received from a master controller in the network. FAST
STATUS commands are described below in relationship with the microcontroller. By incrementing the trip time offset counter, a time stamp of the trip event is obtainable.
After the trip time offset counter is incremented, the system determines in step P32 whether two samples have been taken. If not, during the first sample loop the system proceeds with various housekeeping chores in steps P34-P42, such as servicing the deadman and the non-volatile RAM.
reading the membrane switches and strobing the LED's D60.
The system then returns to step P20 to take additional - 51 ~ ~ ~ ~ ~ ~ ~ 54, 218-I-2 samples... Subsequently, the steps P22-P32 are repeated.
After two samples have been completed, the samples are scaled by a predetermined number, for example. four, in step P44. The two samples are added in a SUM2 register for use for instantaneous tripping and also saved for later use in a cumulative register for the short and long time delay functions in step P46. Subsequently, the sampled currents are auctioned in step P48. The auctioning relates to selecting of the highest of the input phase currents. These input phase currents are then checked by way of the instantaneous subroutine in step P50 in FIGS. 30 and 31 to be discussed below.
The system next checks in step P52 whether or not the trip flag has been set as a result of the instantaneous subroutine in step PSO. If the trip flag has been set at this stage, this would be indicative of an instantaneous trip. Accordingly, the system proceeds to step P56. If not, the SUM2 register is cleared in step P54 since this sum was previously stored above in a cumulative register discussed above. If the trip flag was set as a result of step P50 then the SUM2 register is saved for the display to indicate the electrical current at tripping and the system proceeds to step P56.
The system checks in step P56 whether eight samples have been taken. If not, the system returns to step P20 and additional samples are taken and stored. If eight samples have been taken, the system again checks in step P58 if the trip flag has been set. If not, the eight samples are stored in a register and saved for later use with the long time delay protection subroutine in step P60. The highest phase current of the eight sample currents are then auctioned in step P62. The SUM2 register is then cleared and step P64 and the eight auctioned samples are then 21~v~~J
-- - 51p, 54, 2i8-I-2 checked b'~r the short time delay subroutine in step P66 discussed below. Next, the system determines in step P68 whether the trip flag was set.
If there was a short time delay trip, as a result of step P66, then the register containing the sum of tire eight samples is used to indicate the value of the current at tripping and the system control proceeds to step P72. If not, this sum is cleared in step P70. The system next checks whether 16 samples have been taken in step P72. If not, the system then proceeds and does several housekeeping chores identified by steps P74, P76 and P78 while such 16 samples are in progress. If 16 samples have been taken, the system then checks to see whether the trip flag has been set in step P80. If not, the 16 samples are summed and also saved for later use in step P82. The 16 samples are then auctioned in step P84 and the SUM2 register is cleared in step P86. Subsequently, the auctioned samples are checked by the long time delay protection subroutine identified with the step P88 and the system checks to see if the trip flag was set in step P94. If the trip flag has been set, the value at which the trip occurred is displayed and the system proceeds directly to step P94. If the trip flag was not set, the register containing the 16 samples is cleared in step P92, since these samples were previously saved in the 64 sample register. The system then continues taking samples and checks to determine when 64 samples have been taken. Once 64. samples have been taken, the LED's D60 are then strobed for a quarter second in step P96. The system again checks to determine if the trip flag has been set in step P98. If so, the block P99 indicates the number of samples used for the various functions. If not, the 64 samples are loaded into a temporary buffer for metering in step P100. The 64 samples are then summed in a 256 sample S1B 54, 2_8-_-2 register and the 64 sample register is cleared in step P102. The system then checks to determine if 256 samples have been taken in step P104. If not, the samples are continued to be taken every 4.7 milliseconds. Once 256 samples are taken, a time out counter is incremented in step P106.
The time out counter is a timer which is initiated once the program mode is initiated. The time out counter is used to prevent users from initiating a program mode and delaying programming in setpoints.
The system subsequently checks to see if the trip flag has been set in step P108. If not, the 256 samples are summed in a register and also stored in a register used for adding 65K samples in step 110. The 256 sample register is then cleared. Subsequently the one second flag is complemented in step P112. This one second flag is used to flash an LED D60 to indicate that the system is functional. If the trip flag was set in step P108, various housekeeping chores are performed in steps P114-P118. For example, the control registers are tested to see if they are in working order in step P114. Also amplifiers in the SURE
CHIP PLUS microcontroller are zeroed in step P116 as discussed below. This time is also used to read the CT
ratio programmed in by the gang switch D90 in step P118.
Additionally, the Z count is sampled to indicate whether the system is operated at 50 or 60 Hz in step P122.
The Z count relates to circuitry illustrated in FIG. 16 for determining the frequency of the system. More specifically, the circuitry includes a voltage divider network, generally identified With the reference numeral D129 connected to the phase current IA, for example. Such circuitry is used to monitor the time period between zero crossings of the phase current IA. More specifically, the S1C X4.2'_8-_-2 output of~the voltage divider D129, identified as ZCOUNT, :s applied to the microcontroller D20. The signal ZCOUNT
produces pulses which represent the zero crossings. These pulses are timed by the microcontroller D20 to determine the frequency of the system.
The system is also able to store peak demand currents. for example, for five minutes (e.g., 65~c samples). Thus, the samples are summed and loaded into a peak demand buffer and compared with displayed currents in step P124. Next, the system checks to see if 65K samples were taken in step P126. If so, after the peak demand buffer is loaded, the 65K sum is then erased in step P128.
Subsequently. the system does housekeeping chores, for example, service the EPROM as indicated by steps P130 and P132. After the 65K samples have been completed, the system recycles.
Long Time Delay The long time delay protection subroutine P88 for use with phase overcurrent protection is illustrated in FIGS. 24 and 25. The long time delay protection subroutine for use for ground overcurrent protection is illustrated in FIGS. 26 and 27. Since the phase and ground long time subroutines are virtually identical, only the phase overcurrent subroutine is described and illustrated.
Initially, the system determines in step P134 if the electrical current flowing through the electrical interrupting device is above the long delay pickup (LDPU) setting. If not, the long time delay timer and trip tally is cleared in step P136. Additionally, the LED's D60 are cleared as well as the phase zone out signals described above. Subsequently, a HIGH LOAD is indicated by Way of the LED's D60 in step P138. The HIGH LOAD function is described in detail in U.S. Patent No. 4,827,369.

51D 54.218-i-2 Next, in step P140 the ground long time protection subroutine, illustrated in FIGS. 26 and 27 is tested.
IF the electrical current through the electrical interrupting device is above the LDPU, the HIGH LOAD i~di-cator is cleared in step 142. Subsequently, the system determines if the trip flag has been set in step P144. More specifically, once the electrical current is above the LDPU
a long time delay timer is initiated. Once the timer times out a trip flag is set. If the trip flag has been set, a trip command is initiated in step P146. If not, the LED's D60 as well as the PHASE-ZONE-IN interlock is set in step P148 as discussed above. Additionally, the PHASE-ZONE-OUT
interlock is read.
The system next determines in step P150 whether -the PHASE-ZONE-IN interlock has been set. If not, this indicates t!~at the long time delay protection will be pro-vided by another overcurrent device in the electrical dis-tribution system as discussed below. In this case, the system proceeds to step P152 to determine if the system is in the second loop. If so, a long time delay trip of the circuit breaker is initiated to clear the fault. If not, the system proceeds to step P156 and sets a second pass flag and subsequently tests the ground current long time delay protection in step P158.
If the interlock in step P150 is set indicating that long time delay protection is not being provided by another overcurrent device in the electrical distribution system, the system must then provide the long delay time protection feature. The system then checks the slope of the selected long time delay portion in steps P160 through P168 with the default setting set at the I2t slope.

2~.~~?~~
-- S1E 54, 21~-.-2 ;After the selected slope has been determined, the system proceeds to step P170 to determine whether the system is being utilized on a SO Hz or 60 Hz system. As previously discussed, zero crossings of the phase current are sensed and read by the microcontroller D20 and identified as ZCOGNT. Once a determination has been made in step P170, the system then obtains the selected trip level setting for the selected slope as generally indicated by the bracket identified with the reference numeral P172. Subsequently, in step PI74, the selected trip setting is compared with the maximum phase currents obtained in step P84. Next, in step P176, the system determines whether the long time delay trip tally timer has timed out. If not, the system tests the long time delay protection subroutine for the ground currents in step P178.
In order to prevent any overlap in the short time delay and long time delay protection curves, the system checks in step P180 whether the short time delay pickup value has been exceeded. If so, the system proceeds to step P184 and loads the short time delay trip time value in step P184 and determines if this time has been exceeded in step P186. If the short time delay pickup time value has been exceeded in step P186, the system then sets the trip flags in step P188 and initiates a trip in step P190. If the short time delay time value has not been exceeded, then the long time delay protection for the ground current is tested in step P192.
Short Time Delay Protection The subroutine for the short time delay protection is illustrated in FIG. 28. Initially the system determines in step P196 whether the short time delay phase protection has been enabled. If not, the system proceeds to step P198 and tests the short time delay function for the ground z~~~~~a - 51F 54, 218-I-2 currents:.. If the short time delay :rotection function has been enabled, a long/short delay tally timer is incremented in step P200. This timer is used to prevent overlap between the long time and short time delay functions. Once the long/short delay tally timer is incremented in step P200, the system checks in step P202 to determine whether the electrical current flowing through the interrupting device is greater than the LDPU in step P202. If not, the long/short delay tally timer is cleared in step P204. If so, the system proceeds to step P206 to determine if the electrical current is greater than the SDPU. If not, the long/short tally is cleared as well as the LED's D60 in step P208 and the system then proceeds to step P210 to test the ground current short time protection function. If the electrical current is greater than the SDPU, the system then determines whether the SDPU was picked up before in step P212. If not, the LED's D60 are set and a pickup flag is set as well in step P214. If so, the phase short delay tally is incremented in step P216. After the phase short delay tally is incremented. the system next determines in step P218 whether the short time delay timer has timed out. If not, the systems returns back to the main program. If so, the system sets the trip flags in step P220 and initiates a trip function in step P222.
Instantaneous Protection Instantaneous protection is illustrated in FIGS.
29 and 30. FIG. 29 relates to a discriminator protection routine which is used in systems where there is a report back of the status of the circuit breaker (e.g., open or closed) back to the trip unit.
Initially, the system determines whether the instantaneous phase protection is enabled in step P224. If not, the system then proceeds to step P226 and tests the 2Zfl~~~
,,_ 51G 54, 218-:-2 ground current instantaneous protection. If so. the system proceeds to step P226 and loads the pickup setting. Next, in step P228 the pickup setting is then compared with the maximum phase currents in step P228. Next, the system determines in step P230 whether the electrical currents flowing through the electrical interrupting device are greater than the pickup setting in step P230. If so, the trip flags are set in step P232 and a trip is initiated in step P234. If not, the system then proceeds and determines whether or not the discriminator option is enabled in step P236. If the discriminator option was not selected, the system proceeds to step P226 and tests the instantaneous ground current protection. If the discriminator option has been selected, the system proceeds to step P238 to determine the status of the interrupting device. If there is current flowing through the electrical circuit interrupting device.
the system then determines in step P241 how long the circuit breaker has been closed. This is done by determining how many times the discriminator routine was entered after the circuit breaker was closed. Each time the discriminator routine is entered, a discriminator counter DCOUNT is incremented. If, for example, DCOUNT is greater than 20 which indicates that the breaker has been closed for about ten cycles. the system proceeds to step P226 and tests the ground current instantaneous protection. If the discriminator counter DCOUNT is less than 20, the counter is incremented in step P242 and the current flowing through the circuit breaker is compared with an instantaneous setpoint. If the current through the electrical interrupting device is greater than the setpoint as determined in step P244, the discriminator flags are set in step P246 and a trip is initiated in step P234. If not, the system proceeds to step P226.

S1H ~~~~~~~ 54,218-:-2 ''The turbo instantaneous protection is shown in FIGS. 30 and 31. As previously discussed, this routine cooperates with the analog circuit D129 (FIG. 16) described above in order to provide for instantaneous tripping during conditions when the current transformers piloting the overcurrent trip unit are saturated. More specifically, the system initially samples the phase currents in step P250.
After sampling the phase currents in step P250, the system determines in step P252 whether the turbo phase protection has been enabled. If not, the system determines whether the current transformers are saturated in step P254. If so, a saturated CT flag is set in step P256 and the system returns to the main program. If the turbo phase protection has been enabled as indicated by step P252, the system then proceeds to step P258 to sample the highest phase current again.
Subsequently, the system determines in step P260 whether the last sample was greater than the previous sample. If so, the system returns to step P258 and takes another sample.
If not, the system compares the last three samples to determine if they were true samples or noise. If it is determined that the samples are true, these samples are then compared with the trip level values in step P262 to determine if they are greater than the selected trip settings in step P264. If not, the system returns to the main program. If so, the instantaneous turbo trip flags are set in step P266 and a trip is initiated in step P268.
FIG. 31 illustrates the set up for the turbo mode instantaneous program. This set up is generally enabled at power up after the.setpoint programming. Initially the system determines in step P270 whether the turbo phase protection has been enabled. If not, the pulse width modulator output is set to a preselected value, for example, 28 times the minimum pickup level of nominal pickup level of 2~Q~2~
- S1I 54, 218-I-2 S amperes' in step P272. If the turbo phase protection has been enabled, the system then ascertains the program trip level to determine if the trip level was set greater than 20 per unit. If not. the system proceeds to step P274 and sets the pulse width modulator output to 28 times the pickup value. If so, the pulse width modulator output is set to the trip level in step P276.
DISPLAYS
The routines for generating the various displays on the front panel F20 are disclosed in Figures 115 through 123. In the FRONT PANEL routine, P300 shown in Figures 115-118, if the reset push-button is depressed as determined at P302, the push-button byte is reset at P304 and the RESET
MODE subroutine is called at P306 after which the routine is exited through tag P308. The RESET MODE subroutine clears .the RESET VALID flag so that on the next execution of the FRONT PANEL routine TRIP MODE is called at P312, and then the routine jumps to Figure 118 at P314.
If the unit is not in the trip mode at P310, and is not in the program mode as determined at P316 and P318, the routine jumps to Figure 116 at P320.
On the first time through the routine after the program push-button has been selected as determined at P318, the push-button bytes are reset at P322. The program mode can only be entered when the circuit breaker is open. If it is open as determined at P324, the present values of the adjustable parameters are loaded at P326. Then, and on subsequent executions of the routine, if the circuit breaker remains open at P328. the PROGRAM MODE subroutine is called at P330 and then upon completion of this subroutine the FRONT PANEL routine is exited through tag P332. On the other hand, if the circuit breaker is not open, the flag for the program mode is cleared, the display D86 is made blank .- 54,218-I-2 and the program LEDs D60 are turned off at P334 and the routine is exited through tag P336.
The FRONT PANEL routine continues in Figure 116.
If the trip unit is in the trip mode at P338, the routine jumps to Figure 118 through tag P340. If not in the trip mode at P338, but the current is above one of the protection pickup levels at P342, the test flags are cleared at P344 and the routine branches to Figure 118 through tag P346.
On the first execution of the routine after the test push-button has been depressed, the unit will not yet be in the test mode at P348, but the test flag will be set by the push-button as detected at P350, and the push-button byte will be reset at P352. Then, and on subsequent executions of the routine, the TEST MODE routine is called at P354. If this results in the setting of the test flag as detected at P356, the routine branches to Figure 118 through tag P357, otherwise, the routine is exited through tag P358. Until the test push-button is depressed, the routine branches at P350 to Figure 117 through tag P360.
Figure 117 continues the description of the FRONT
PANEL routine. The first portion of Figure 117 calls the routine shown in Figure 123 for changing the communications parameters if, as previously described, the select test and the test push-buttons are pressed. On the first time through the INCOME ADD flag will not be set at P362. If the test push-button and the select tests push-buttons are both depressed at P364 and P366 the INCOME ADD MODE is called at P368. On subsequent executions, the INCOME ADD flag will be set at P362 and the INCOME ADD MODE routine will be called directly. If either of the TEST or SELECT TEST push buttons are not pressed, the flags for these push buttons are cleared at P370. Then, if the VIEW VALID flag is set at P374 indicating that the VIEW SETTINGS membrane switch is 21a~2~~
S1K S4, 218-I-2 depressed, a VIEW MODE flag is set at P376 and the VIEW MODE
subroutine is called at P380. On subsequent executions, when the VIEW SETTINGS membrane switch has been released, the VIEW MODE subroutine will be called as long as the VIEW
MODE flag remains set at P378.
The final portion of the FRONT PANEL routine is shown in Figure 118. If a time-out flag is set at P384 indicating that 2-1/2 minutes have elapsed since a push-button has been depressed, no RAM error flag has been set at P386 and there is no EEPROM error at P388, the time-out flag, the display D86, and the program and test LEDS D60 are cleared at P398. Then, or in the event there was an error or the timer had not timed out, if the unit is in the view mode as determined at P392, the next item to be displayed is determined at P394. In any event, the DO CURRENT display subroutine is then called at P396 before the program is exited at P398.
The TRIP MODE routine P312 is shown in Figures 119 and 120. When called, the program, test and view mode flags are cleared at P400. If the auto reset feature has not been selected at P402, the data is maintained constant until the operator depresses the reset button. As mentioned previously, the trip unit includes output contacts indicating the type of trip. The first time the routine is called, these relays will not be set as determined at P404. However, if the instantaneous, discriminator or override flag is set at P406, the instantaneous relay output will be energized at P408, otherwise the overcurrent relay output will be set at P410. Then, and on subsequent executions of the routine. the program will transfer to Figure 120 through tag P412.
If the. auto reset feature has been selected and the circuit breaker is open at P414 an OPEN CAN CLOSE flag 2~~~~~~
S1L 54, 218-_-2 is set and the output relays flags are cleared at P416. I
~s the trip was generated in the test mode at P418, the routine jumps to Figure 120 through tag P419. If not in the test mode, the DO BLINK flag is set at P420. This flag is toggled on subsequent execution. If the flag is set at P422, the routine transfers to Figure 120 through tag 424, and if not set through tag P419.
If the breaker is closed at P414, but it has been open on a previous pass as indicated by the fact that the OPEN CAN CLOSE flag is set at P426, then the display is reset by clearing this flag, the trip display, the ampere sums, the elapsed time since pickup timer and the trip flag s at P428. If the nonvolatile RAM is not busy at P430, the byte storing the clear trip data P434 is erased at P432 before the trip mode is exited through tag P434.
The TRIP MODE routine continues in Figure 120. If the system is not in the test mode at P438, the cause of the trip is loaded for the alpha message a P440, and the alpha-numeric display routine is called at P442. Whether in the test mode or not, the cause of the TRIP is determined at P444 and the appropriate red LED is turned on. The program then returns to the point where the TRIP MODE subroutine was called in Figure 115.
The flow chart for the VIEW MODE subroutine P381, which is called in Figure 17, is shown in Figure 121. This routine uses two flags: a VIEW VALID flag which is set when the VIEW SETTINGS push-button on the front panel is depressed, and a VIEW flag which is set on the first execution of the subroutine. Thus, when the subroutine is first entered, the VIEW VALID flag will be set at P448 and this flag, the inactivity timer and its flag are cleared at P450. Since the VIEW flag is not set at P452, all of the LEDs are cleared, the VIEW flag is set and a set point -2~O~~~a ''-~ S1M 54, 218-I-2 counter Which counts the parameters viewed is set at P454.
As this is the first time through the routine, the range of parameters which can be viewed will be initialized at P456 and a routine will be called at P458 which displays the selected set point on the alpha-numeric display and lights the green LED for the parameter being displayed. The next time the VIEW SETTINGS push-button is depressed, the set point counter will be incremented at P460 and the new set point value will be displayed and the appropriate green LED
will be illuminated at P458. When all of the parameters have been viewed, the display is cleared at P460. Each time the routine is completed, the program returns to Figure 117 where the VIEW MODE was called as indicated at P462.
Figure 122 illustrates the LITE EM UP subroutine P38 which is called in the protection routine shown in Figure 17. This subroutine turns on the selected LED in either the steady or flashing state through the appropriate register D52-D58 shown in Figure 10B. The first part of the routine controls the LEDs operated by the register D52. If the BLINKING flag is set at P448, then the data for the associated LEDs is complemented at P450, and in either case, the byte LEDO is output to the register D52 at P452.
Similarly, the LEDs controlled by the register D54 are turned on, if appropriate, and in the steady or flashing state as shown at P454-P458. The OPERATIONAL LED, which is one of the LEDs controlled by the register D56. flashes green at one second intervals to indicate that the unit is operating properly, or red to indicate a malfunction. If the one second flag (which is complemented at P112 in Figure 22), is set at P460, the bit for this LED is complemented at P462. If one of the other LEDs set by D56 is selected for blinking, the output is complemented when the BLINKING flag has been set as indicated at P464 and P466. This LED, if "~ 51N 54, 218-I-2 any, and the OPERATIONAL LED are then energized by the L~D2 byte which is outputted at P468 to the D56 register.
Finally, the LEDs controlled by the register D58 are energized if selected, and caused to flash if appropriate at P470-P474. The program then returns to the protection routine of Figure 17.
Figures 123 and 124 illustrate the flow chart for the INCOM ADD MODE P368 which is called by the FRONT PANEL
routine at Figure 117. This routine is used to change the communications address and the baud rate using the alpha-numeric display D86 -and the membrane switches. As discussed in connection with Figure 117. this mode is entered by simultaneously pressing the TEST and SELECT TEST switches.
When the subroutine is first entered, the ADD MODE flag will not be set at P476 and the current address and baud rate will be displayed on the alpha-numeric display D86 and the ADD MODE flag will be set at P478. The program then returns to the FRONT PANEL routine through the tag P480. The next time the routine is executed, if two and half minutes have elapsed since any of the membrane push-buttons have been depressed as determined at P482, the previous income address and baud rate are reloaded at P484 and the program jumps at P486. If the raise pushbutton has been pressed at P488 and the baud rate has been selected at P490, the baud rate is changed at P492 and the subroutine is exited through the tag P480. The baud rate and the digits of the address are selected through the SELECT SETTINGS membrane switch. If one of the address digits has been selected, the value of that digit is increased at P494. Similarly, if the LOWER
push-button has been selected at P496 the baud rate is changed or the address digit is decremented in a like manner as indicated at P498-P502.

510 ~~~~~~V 54,218-I-2 ~:If instead of the RAISE or LOWER push-button, the SELECT push-button has been selected at P504, the next digit of the address or the baud rate is selected at P506 for adjustment. Selection is indicated to the user by the flashing of the appropriate digit of the alpha numeric display.
After the operator has made the desired changes to the communications address and/or the baud rate, the SAVE
push button is depressed at P508 to make the indicated changes. If the newly selected address is valid as determined at P510 and P512 the interrupts are disabled while the changes are stored in nonvolatile memory at P514 P518, and the display is cleared at P520. On the other hand, if the address as changed is not valid an error message is displayed on the alpha-numeric display as indicated at P522. In either case, the ADD MODE flag is cleared at P524. This flag is also cleared when the inactivity timer has timed out as indicated by the tag P486.
INTEGRATED CIRCUIT
DEFINITIONS
The following definitions are to be used throughout:
Bit designations: Bits within a register will be designated by placing the bit number within square brackets. For example, bit S of register AHC is designated as: ABC[5]. Bits 5 through 0 of register AHC are designated: ABC[5...0]. Bits 4 and 5 of register AHC are designated as ABC[5,4], etc.
Hexadecimal: Base 16 numbers written with a dollar sign prefix. For example. $0100 = 256 decimal.
High-true: These signals are designated with the suffix "h" and are defined to be asserted (true or logical 1) when their electrical level is at or near the +VDD supply and are defined to be negated false or a logical zero) when -- s4,21a-.-2 their electrical level is at or near zero volts direct current (Vdc).
Low-true: These signals are designated by the suffix "b"~ and are defined to be asserted (true, or a logical 1) when their electrical level is at or near zero Vdc .and are defined to be negated (false or a logical zero) when their electrical level is at or near +vDp sup-ply.
Input: An input signal is received by the IC
10.
output: An output signal is driven by the I~
10.
Referring to the drawings, an Ic is generally identified with the reference nuaeral 10. The circuitry of the IC 10 has been standardized to enable it to be utilized with various types of electrical equipment including circuit breakers, motor controllers, and the like. For illustration and discussion purposes only, the IC 10 is shown utilized in a circuit breaker 1Z in FIG. 32. The circuit breaker 12 is shown as a three phase circuit breaker having phases iden-tified as "A~, ~H" and "C". It is to be understood by those of ordinary skill fn the art that the IC i0 is cap-able of being utilized with various other types of elec-trical equipment, such as motor controllers, contactors and the like.
The circuit breaker 1Z doss not form a portion of the present invention. As shown, the circuit breaker 1Z is provided with three current transformers 14, 16 and 18. These current transformers 14, 16 and 18 are disposed on a load aide Z0 of the circuit breaker 1Z to allow the circuit breaker 13 to be'monitored and controlled. The line side ZZ of the circuit breaker 1Z is generally con-nected to a three phase source of electrical energy (not shown). The load side 20 is generally connected to a three. phase load (not shown), such as an electrical motor.
aspect of the invention relates to the tact that the IC 10 is current driven as will be dis Hore specifically, electrical current from the current transformers 14, 16 and 18 is applied to the IC 10 by way of conditioning circuitry 19 (FIGS. 3Z and 76H). The conditioning circuitry 19 is used to provide an elsctrical current of about IO microamperes (~A) to the I
lo.
l~nother aspect of the invention re-lates to an on-board communication controller 29. This controller allows the IC 10 to communicate with devices, such as a panel meter 31 as illustrated in FIG. 32 by way of a communications network link, such as a twisted pair 33. l~rnother IC 10 or an INCOH chip, as disclosed in U. S.
Patent No. 4,644,566, can be disposed in the remote meter 31 to allow the resote panel meter 31 to be connected in another network, by way of another communications network link 35 which includes its own master controller (not shown). A general description of a similar communication controller connected in a network with a master controller is described in detail in U.S. Patent No. 4,644,566, A block diagram of the digital portion of an exemplary embodiment of the IC 10 is shown in FIG. 33.
Z0 Various configuration options are selectable by software Programming and mask options for specific applications as will be discussed below. In order to provide an Ic 10 with universal application for controlling and monitoring various types of electrical equipment, such as circuit breakers, actor controllers, contactors and the like, various peripherals say be provided. These peripheral de-vice say include a microprocessor 30 with a multiply in-struction, for example, a Motorola type IiC68H05. The micsoprocessor 3o communicates with the various other peripherals and external pins on the chip 10 by way of an internal address, data and control bus 34 and an external bus controller 31. A clock generator 36 provides timing for the microprocessor 30. Jan on-board memory subsystem is provided which may include read only memory (ROH) 38, electrically erasable read only memory (EEPROH) 40 and random access memory (R~) 4Z. The EEPROH may be provided with an internal charge pump 44 for obviating the head to prov~ida an external high voltage source for erasure in programming of the EEPROM 40.
Exemplary memory sizes are provided for illus tration and discussion purposes only. For example, 256 bytes of EEPROM 40 and 208 bytes of RAM 42 may be pro vided. The ROM 38 may include 4,096 bytes of mask pro-grammable user instruction memory and 240 bytes of self-teat memory.
The IC 10 has separate digital and analog power ~ supply systems. These power supply systems are segregated to prevent digital noise from affecting the analog cir cuitry. The digital power supply is directed to a VDD pin on the IC 10 formed from an external voltage regulator (FIG. 76C). The analog power supply is directed to a AVDD
pin. In many applications, an external transistor, diode and resistor are adequate.
Power monitor circuitry 47 is provided, which includes an internal voltage level detector for monitoring the gate drive to the shunt regulator associated with the 2o AVDD pin. The power monitor circuitry 47 deactivates a SHUNT output pin which, wham connected to the RESN input pin, provides a rest when the analog supply voltage AVDD
begins to fall below a preset threshold. Dead-man cir-cuitry 46 is provided which monitors operation of the aicroprocessor 30 and activates a reset upon detection of spurious operation.
An analog power supply subsystem 48 is provided !or the analog circuitry. This subsystem 48 includes a +i.ZS Vdc band gap regulator and a buffer amplifier to generate a +=,S Vdc reference. An external current source is used to power the analog power supply subsystem 48.
The current source is directed to an external pin AVDD.
An adjustaent pin VAD,T is provided to allow the voltage reference to be trig»d to exactly +Z,5 Vdc. Trimming may be provided by a voltage divider circuit, for example, two seriu connected resistors connected between the VREF pin and an analog ground pin AvSS, as illustrated in FIG. 75.
The interlace between the series connected resistors is 55 ~~~~~u .;
connected to the VADJ pin. A shunt regulator provides a nominal +5.0 Vdc source at the AVDD pin based on the ref-erence voltage at the VREF pin. The buffer amplifier is provided with an open drain output such that it can only source current. This will permit multiple devices to be paralleled. The regulator can also be slaved to another IC. This is accomplished by connecting the VADJ pin to the VREF pin on the slave IC and connecting the VREF pin on the slave IC to the VREF pin on the master IC.
A comparator subsystem is provided which in-cludes an A comparator 49, a 8+ comparator 50 and quadcom-parators 58. The A comparator 49 is !or use with the com-munication controller 29. A B+ comparator 50 is provided !or external power supply generation as illustrated in FIG. 76H and discussed below. The inverting input of the comparator 50 is referenced to the VREF pin (nominally, +2.5 Vdc). The input signal is applied to a non-inverting input pin BSENSE o! the comparator 50. The 8+ comparator 50 output is connected to an external pin BORIVE. The quadcomparators 58 include four comparator~ referenced to a fixed voltage, !or example +1.25 Vdc.
Various special functions may also be provided, such as a tiger 60 and a pulse width modulated output 61.
The tiger 60 gay be used !or time base or wavalor' genera-tion periodic measuruant or other periodic functions.
The P'w~i output is a periodic signal whose period may be controlled by the microprocessor's phase 2 clock divided by 4.
Various other peripheral devices may also be provided on the IC 10 to allow it to be adapted for various applications, such as circuit breakers, motor con trollers and the like. For example, these peripheral de vices may include lour gsneral purpose 8-bit bidirectional ports; port ~ (S?), Port 8 (53), Port C (56) and Port D
(57). A serial peripheral interlace 54 (SPI) may also be provided to allow !or etlicisnt connection o! peripheral devices that conunicate over a serial bug. The SPI 54 gay also be used !or interprocessor comsunication in a multiprocessor system. The SPI 54 supports several operating modes that permit connection of devices that communicate usinc various protocols.
7,n aspect of the invention relates to an analog subsystu, identified by the function blocks 62 and 64 in FIG. 33. A block diagram for this subsystem is illustrated in FIG. 34. The analog subsystem includes, for example, eight analog input channels for receiving analog voltage and current signals and converting these signals to an 8-bit digital signal with 12-bit resolution.
Four of the input channels 6Z can be selected by the software to operate as either voltage inputs or currant inputs. The other input channels 64 can only be operated as voltage inputs. Current and voltage input operation of the input channels 62 and 64 is controlled by multiplaxars (I~JXes) 66 and 68, which are selected by the software.
The voltage input channels can accept positive voltages in the range of 0-Z.5 Vdc and are applied to an auto-zeroable voltage adjustable gain amplifier 80. These signals can be processed in either an auto-ranging mode or a fixed gain mode, selectable by the software. It the auto-ranging soda is selected, the selected voltage input channel 6Z or 64 is ranged by values stored in an internal auto-ranging register to allow the gain to be automati-cally adjusted until the signal is at least one-halt of full scale but not in owrtlow. The ranged signal is than converted directly to an 8-bit digital value by an A/D ~8 and stored in an internal register. I! fixed ranging is selected, the voltage mode inputs 6Z or 64 can be operated at preselected gain settings, for example, l, 2, 4, 8 or 16. The preselected gain settings are provided by gain circuitry which includes a resistor network 84, connected to an inverting terminal of the voltage amplifier 80 by way of a ~ttnC i6. The ~JX 8s is controlled by the sott-ware. The output of the voltage amplifier 80 is connected to the 11/D circuitry 78 by way of another I~JX 88. The Ii~JX
8s is in the position shown in FIG. 34 when voltage gains other than one have been selected. However, when a gain of .pne is selected, the voltage amplifier 80 is discon-nected from the ranging circuitry and the input voltage channel 62 or 64 is applied directly to the A/D 78.
Current mode inputs 62 accept negative currents (e.g., currents flowing out of a MXO pin) with a -1.6 mA, for example, representing full scale. Unselected current input channels are tied to a digital ground pin (VSS) by the MUXes 68 which provide for make-before-break switch ing. The selected input current channel is connected to 1o an inverting input of an auto-zeroable current amplifier 90, referenced to analog ground (AVSS). The source fol-lower output is configured to maintain the inverting input at a virtual ground by providing currant to the selected channels through an adjustable current mirror 92. The current mirror 92 can be set by auto-ranging circuitry, which can be overwritten by the software, to one of the following exemplary ratios: 1/1, 1/2, 1/4, 1/8 or 1/1s.
The ratioed mirror output current is directed to the Mxo pin. Thus, the current flowing out of the lrDCO pin will be a programmable fraction o! the sue of the currents flowing out o! the selected current input channel pin. Two modes of operation are possible:
Non-integrating mode, 1,n external resistor (not shown) may be connected between the analog ground pin (AVSS) and the 1~0(p pin to convert the ratioad current to a voltage. T8ls voltage may then be converted to a digital value in a similar manner as discussed above. 1n this ~od~ o! operation, the default amplifier gain is set at tiaes one unless overwritten by the software.
Integrating mode. An external capacitor (not shown) may be connected between the analog ground pin (AVSS) and the I~DCO pin to integrate the ratioed current.
A shorting switch 96 is provided to discharge this capaci tor under prograa control. Integrated voltages are then converted by the A/p 78 as discussed above.
The voltage and current amplifiers 80 and 90 have offset voltage compensation circuitry 98 to compen-sate for oflsets inherent in CHOS amplifiers. These off-seta can be on the order ot, for example, ~20 millivolt (mV) which can a!lect the accuracy o! the least signifi-cant bits of the converted digital value. This circuitry 98 assures that the o!lset is always a negative voltage between 0.0 and 0.5 mV and forces the amplifier 80, 90 to have a positive output when the differential input voltage is zero. This offset correction can be done either auto-matically by the hardware or controlled by the software.
Sample and hold capability is provided !or the input channels 6I and 64. Hors specifically, the eight analog input channels 6Z and 64 are grouped into four pairs of channels 10Z, 104, 106 and 107. Each pair of channels 10Z, 104, 106 and 107 can be used as a single channel with sample and hold capability. A capacitor (not shown) may be connected between one channel input and the analog ground pin aVSS while the other channel of the pair is connected to a voltage input. 7~ saapla command permits the software to connect the two channels of each pair to-gether by way of sample and hold MLTXea 108, 110, 112 and 114, thus storing the input voltage on the capacitor of the adjacent channel. Only channel pairs in which both channels are eontigured for voltage mode operate in this fashion. X11 tour channel pairs 102, 104, 106 and 1o8 may ~ sampled siaultaneously.
There are other aspects of the inven-tion. hor example, the IC 10 is adapted to respond to eider analog signals or digital signals and provide a digital output signal. mother aspect relates to the tact that the IC 10 is provided with circuitry which enable it to sense the ambient teaperature in which it is disposed and generate a corresponding electrical signal.
Q~$,aTZNG HOD -a The ZC 10 has five modes o! operation. Two nor sal operating modes are single-chip and expanded. These are selected by strapping an EXPH pin to either the VSS or VDD pins. The three special operating modes are: emula tion, teat and self-check. These modes can only be selected by connecting certain pins to a voltage level twice VDD when the reset line is negated. They cannot be enabled by normal product operation. The operating mode of the IC 10 is determined by the electrical input levels of EXPN, IRQN, and TCAP pins when the device comes out of reset. These pins are sampled when the RESN pin transi-tions from an electrical low to high. Table 1 defines the pin input levels for various operating modes.
Table 1 OPERATING MODE S- -CTION
Single-Chip VDD VSS to VDD VSS to VDD
Expanded VSS VSS to VDD VSS to VDD
Emulation VDD*2 VSS to VDD VSS t° VDD
Teat VSS VDD*2 VDD
Sell-Check VSS VDD*2 VSS
The operating mode determines the function of certain device pins: ALE, PSEN, REN, WEN and PH2. The following sections discuss the behavior of the IC 1o in these different operating modes.
1. Sincle-chiOr-The single-chip mode i~ selected when the RESN
pin changes troa an electrical low to high and the EXPN
pin is at VDD. In this mode, Ports A and H operate as notes! bidirectional I/0 ports and the microprocessor exe-cutes code troy internal mask-programmed ROX. See Table 2 for operation of variable function device piny.
Z . ~andad trt~d.
The expanded mode is selected when the RESN pin changa~ troy err electrical low to high and the EXpN pin is at VSS. In this mode, port A becoae a multiplexed data/addres~ bus and Port H becomes the high-order address bus. This code requires program code to reside in an ex ternal maaory device. The internal code ROH is not avail able and all meaory locations above $4000 must be imple-sented with external device. See Table 2 for operation of variable function device pins.

so ~I~~~' b 3.~~ulation Mode The emulation mode is a special operating mode that is selected by setting the EXPN pin to a voltage level twice VDD when the RESN pin changes from an electri-cal low to high. This mode operates similar to the ex-panded mode, except that certain pin definitions change.
See Table 2 for operation of variable function device pins.
4. Test Mode The test mode is used for production testing of the IC 10. It is selected by setting IRQN at twice VDD
and TCAP at VDD when the RESN pin input rises. See Table 2 for operation o! variable function device pins.
5. Sell-check Med~
The self-check mode is used for production burn-in. It is selected by setting IRQN at twice VDD and TCAP
at VSS when the RESN pin input rises. Sae Table 2 for operation o! variable function device.
Ta-2~ PIN DEFINZTTnwt t~nn nnroirnTU.. .....,.._.,.
single- Ex- Emula- Tast self Pad chip panded tion check low ALE ALE ALE low PSEN high PSEN LIR LIR high REN high REN R/W R/W high ~1 high WEN E E high PgZ hr PH2 PH2 PH2 low wavr iuuicwl furl A~'lipDS
Since the IC 10 is intended for universal appli cation o! vsrious types of electrical equipment, certain configuration information is required to tailor the- IC to for a specific application. The configuration information is defined by either mask options, software, constants or run time configuration.
Regarding mask options, the contents of the RoM
38 may be specified at the time the IC 10 is manufactured.
Certain other cask configuration options may be determined by modification o! a single mask. These other configura-tion options include the dead-man subsystem 46, IRQN trig-~a cJ ~3 goring, an oscillator option, comparator hyateresis option and an SPI option. A predetermined hysteresis, for example, 20 millivolts (mV) or no hysteresis at all can be selected on an individual comparator basis.
A mask programmable option also permits selec-tion of the type of interrupt generated associated with an IRQN pin. One of two triggering methods may be selected as follows: 1) negative edge sensitive triggering only, or 2) both negative edge-sensitive and low level-sensitive triggering. It option 2) is selected, either type of in-put to the IRAN pin will produce an interrupt. The IC 10 can be configured to accept either a crystal/ceramic re-sonator input or an RC network to control the internal os-cillator. Hors specifically, the IC 10 can be configured by mask option to accept either a crystal/ceramic re-sonator input or an RC network to control the internal os-cillator. The internal clocks are derived by a divide-by-two of an internal oscillator, which operates with an AT-cut parallel re:onant quartz crystal resonator in the fre-quency range of 1 l~iz to 8 l~iz. Use of an external oscil lator is recommended when crystals outside the specified range are to be used. The crystal and components are be mounted as close as possible to the input pine to minimize start-up and stability problem. Table 1 shows recom sanded paraaeters for crystal resonators.
AM~C 8 CRYSTAL Cgg~IC
3 0 ~' Rg 400 75 10 ohas CO 5 ~ 40 1 8000 12000 4.3 pF
COSC1 15-40 15-30 30 pg COSCZ 15-30 15-25 30 pF
Rp 10 10 1-10 meQohms Q 30 40 1.25 103 A ceramic resonator may be a:ed in place of the crystal in cost-een~itive applications. The circuit shown in FIG. 35A is recoamended when wing a ceramic resonator.

s2 Itsp equivalent circuit is shown in FIG. 358. Table 3 shows the recommended parameters for various resonators.
A mask programmable oscillator option may be selected to permit the use o! a single external resistor R
between external oscillator pins OSC1 and OSC2 as illus trated in FIG. 35C. With this option, frequencies between 5 MFIz and 70 IQiz are practical. An external clock input should be used with either mask oscillator option. This external clock is connected to the OSC1 pin with the osc2 pin unconnected as shown in FIG. 35D.
Two mask options are available for the SPI.
This mask option selects one o! two serial peripheral in-terface data pin (MOSI, MISO) configurations.
Hidirectional Data Pins: This configuration causes the EPI data pins to change direction when master or slave operation is selected. The MOSI pin is an output when in the master mode and an input when in the slave mode. The MISo pin is an input when in the master mode and an output when in the slave mode.
Unidirectional Data Pins: This configuration forces the SPI data pins to operate indepen dently o! the mode o! the SPI. When this con figuration is selected, MOSI is always an output and MISO is always an input.
In addition to the cask options, software con-stants are also used !or configuration o! the IC 1o. More specifically, internal configuration registers are loaded by the sicroprocusor software lroa application dependent soltvare constants stored in the prograa ROM 38 or the EEPROM 40. A pair o! internal configuration registers (CFR, ACFR) are used to control these options in the IC
10. The CFR and ACFR registers are loaded on program ini-tialization and are not intended to be altered during nor-aal prograa operation and will be discussed in detail be-low.
Lastly, the IC to can be configured by run time configuration. with this option, configuration data is s3 ~.~~~u read from external devices through the microprocessor's I/O subsystem. This can be done by utilizing the in-put/output ports A, B, C or D in either a parallel or serial fashion.
CONFIGtn?~TION REGTST tea The configuration registers CFR and ACFR are used to specify various software configuration options available in the IC 10 architecture. These registers CFR, ACFR are programmed during software initialization to con-figure input/output pins to their appropriate function as well as setting other major configuration parameters. In order to avoid improper operation of the IC 10, the CFR
and ACFR registers should not be altered during normal operation.
The CFR register is a write only register. The ACFR register is a read-write register. The bit format for the CFR and ACFR registers is illustrated in FIG. 37.
Both the CFR and ACFR configuration registers are ini-tialized to zeros on power-up or reset. This defines the state of the IC 10 after power has been applied and before the microprocessor 30 alters it for the application.
The CFR register is a write-only register used to configure the comparator output options. Bits 4 and 5 are unimpluented. The remainder of the bits in the CFR
register are defined below.
C!R[7]: Communication Subsystem Master Enable.
Thin is a peraissive bit that allows the communication controller subeystea 29 to be switched into a master mode.
A sero in Ibis configuration bit prevents the communica-tion controller subsystea 29 frog entering into a master mode of operation. ~ one permits master operation. This bit is set to zero on reset.
CFR[6]: SPIOFF. This is s disable bit for the SPI subsystea. When sat, the SPI subsystea is disabled.
This bit is set to zero on reset, CFR[3...0]: Comparator Mode Control. These four contiguration bits enable the comparator outputs to be "ORed" with the least significant four bits of port C.
A zero in these configuration bits enables the OR opera-tion for the associated port pin. In this mode, each out-put pin will be low during device reset if the respective comparator input is above the threshold voltage (+1.25 V).
Reset will clear the microprocessor's PORT C output regis ter, making the output pin only a function of the compara tor input. When the microprocessor writes a "1" into the port output register, the output pin will be forced high independent of the state of the comparator input.
A one in these configuration bits disables the 'OR' operation. In this mode, the port pins are in a high impedance state after reset. The configuration bits are assigned in sequential order with CFR[0] controlling PCO/C1~0 and CF'R[3] controlling PC3/CMP3.
ACFR REGISTER
The ACF'R register is a 7-bit road-write register and is used to configure the analog subsystem. This register is set to zero on device reset or power-up. Bit 4 is unimplemented. Bit definitions of the ACFR register are as follows:
ACFR[7]: Clock Source. This bit selects the clock source !or the A/D 78 and the communication con-troller subsystsa 29 as well as the EEPROH charge pump 44.
They can be configured to use either an internally gener-ated clock or a clock derived fro' an external crystal os-cillator of the IC 10. It the crystal oscillator is selected (ACIrR[7] = 1), the oscillator frequency must be in the rsnge of Z-8 l~iz. other crystal values must use the internal clock source option (ACP'R[7] = 0)0. If the coaaunication controller subsystu is used, the external crystal option gust be selected (ACFR7] = 1). This bit is set to 0 (internal clock source) by reset. A delay of 10 milliseconds (ms) is required attar selecting the internal clock to perait the oscillator to stabilize. During the stabilization time, A/D 78 and EEPROH 40 operations should be perforsed.

f:~ ti l~
°' 6 5 ACFR(6]: Divider ratio. This bit selects the clock divider ratio !or the A/D 78 and the communication controller subsystem 29. There are two considerations when selecting external crystals and the ACFR[6] setting.
These considerations are; first, the clock input to the A/D 78 must be in the range o! 1-2 I~iz; and second, the clock input to the communication controller subsystem 29 gust be 1.8432 I~iz i! the communication controller carrier and bit rate standards are to be mat.
This bit selects a divider ratio o! 1/2 or 1/4, permitting crystals in the range o! 2-8 I~iz to be used.
I! the communication controller subsystem 29 is active, either a 7.3728 lrQiz or 3.6864 l~iz crystal must bs used.
The selection o! the external crystal and the state o!
ACPR(7] will determine the A/D conversion, autoranging and autozero times. Table 4 defines the configuration bit and dsaonstrates the elect on conversion times.

~~D CONVERTER CLOCR SOURCE
ACFR Crystal A/0 Conv.* Auto-zero**
[7, 6] Divider (l~iz) (~s) (~s) O X Internal 48 96 13 - 1,537 NA -1 0 1/Z Z 48 96 25 - 1,537 -1 0 1/Z 3.6864 26 - 15 - 884 1 1 1/4 4 48 - 25 - 1,537 1 1 1/4 7.3728 26 - 25 - 884 1 1 1/4~ 8 24 - 13 - 769 * auto-range time the Ths depends number on of gain steps required to range.

** The auto-zero tise depends on the amount o! oltset cor-rection required.
ACFR(5]: A/D power-down. This bit controls power down operation o! the A/D subsystem 78. When set, it will power up the A/0 subsystaa 78. When the bit is reset, the A/D subsystem 78 will power down. This bit is set to zero by reset on power-on. At least 100~s should be allowed for the converter to stabilize alter power-up.
ACFR(3...0] : HIJX3. . .I~JXO mode select. These configuration bits control the input mode of the analog input channels 6~ and 64. The input channels 62 (~0~..MtTX3) can be placed in either a voltage input mode or a currant input mode. A zero in these configuration bits selects the voltage mode, while a one selects the currant mode. These bits are assigned sequentially with ACFR[0] controlling MVXO and ACFR[3] controlling MUX3 as shown in Table 5.
AHALO INPLf't' MODE DEFTntr~rTnvc ACF'R[0] = 0 ~ voltage mode, 1 = current mode ACFR[1] = 0 = voltage mode, 1 = currant mode ACFR[Z] = 0 = voltage mode, 1 = currant mode ACFR[3] = 0 = voltage mode, 1 = current mode The microprocessor 30 is based on a Motorola type MC68HC05 architecture, a Von Neumann type machine, which places all data, program and I/O interlaces into a single address map. This reduces the nuabar o! special purpose instructions that must be supported and therefore results in a relatively small and easy to remember in to atruction set.
The microprocessor 3o is described in detail in -w- ----~ ~ra~oev ~V1 F xTTV ZJ.rs S' 1f11diT1T by MOtorCla, inc., copyrighted 1983.
The microprocessor 30 architecture is based on live regis Z5 tars: an accuaulator (A), an index register (X), a pro gram countes (PC), a stack pointer (SP) and a condition cods register (CC).
The accumulator is s register used general purpose 8-bit by the program !or arithmetic calculation 30 and data manipulations. A lull set o! raad/modily/writa instructions operate on this register. The accumulator is used in the register/mesory instructions !or data manipu-lation and arithaetic calculation. The index register is used in the index mode o! addressing or as an auxiliary 35 accumulator. It is an 8-bit register that can be loaded either directly or from memory, having its contents stored in aasory, or its contents compared to meaory. In index instructions, the index register provides an 8-bit value 67 ~~~~~t~
thatøis added to an instruction provided value to create an affective address. The index register is also used for limited calculations and data manipulation.
The program counter is a 16-bit register and contains the memory address of the next instruction that is to be fetched and executed. Normally, the program counter points to the next sequential instruction, how ever, it may be altered by interrupts or certain instruc tions. During an interrupt, the program counter is loaded with the appropriate interrupt vector. Jump and branch instructions may modify the program counter so that the next instruction to be executed is not necessarily the next instruction in memory.
The stack array or stack is an area in memory used for the temporary storage of important information.
It is essentially a sequence of R~rl~ locations used in a last-in-first-out (LIFO) fashion. The stack pointer al ways points to the next tree location on the stack. In terrupts and subroutines males use o! the stack to tem 2o porarily save important information. The stack pointer is used to autocratically store the return address (2 byte progracr counter) on subroutine calf and to automatically store all registers (5 bytes: h, X, PC and CC) during in terrupts. The stack starts at location SOOFF and extends downward 64 location.
The condition code register is a 5-bit register that indicates the results o! the instruction just exe-cuted, as well a~ the state o! the processor. These bits can be individually tested by a progracr instruction and specitied action taken a~ a result of their state. The following condition code bits are defined: halt-carry (H), interrupt crask (I), negative (N), zero (Z) and carry/borrow (C).
~~"~s ~r~...~
The cricroprocessor 30 is capable of addressing 66,536 bytes- of aeaory. Thus, the creaory space ranges troa $0000 to $FFFl. FIG. 36 is a diagracr o! meaory allo-cation for the IC 10.

~~~~.~~9 p.

1. ROM 38 The IC 10 memory map has three sections of mask programmable ROM 38 and accommodate up to 32,768 bytes of external ROM in the expanded mode at locations $8000 to $FFFF. This memory 38 is programmed at device manufac-ture. The three sections o! the ROM 38 are located as de-fined in Table 6.

$0M ASSIGNMENTS
$~SE ~$ FUNCTION
$EF00-SFEFF 4096 bytes User instruction memory $FF00-SFFEF 240 bytes Sell-check program $FFFO-SFFFF 16 bytes Vectors 2 . RAM 4 2 The IC 10 is configured with 208 bytes o! RAM
starting at location 50030 extending to $OOFF and can ac-comodate 16, 384 bytes o! external RAM at locations S4000 to S~FFF~ The top of this internal RAM S0030 to $OOFF
area is reserved for the stack. The stack starts at loca-tion SOOFF and extends downward a maximua o! 64 locations to $OOCO. Unused stack locations may be used by the pro-graa !or genesal storage. However, care must be exercised to avoid data being stored in these locations being over-written by stack operations.
2s 3. EpROM_4o The IC 10 has 2s6 bytes o! EEPROM 40 located at addruses $0100 through $O1FF.
The upper 16-bytes o! the meaory map are re served !or interrupt vectors. The address assignments for each are described below:
SPF1~E-FFFF : RESET VECTOR
This vector is used on processor reset. It has the highest priority o! the eight interrupts.
3s This vector is used during execution o! the SwI
instruction. It has the second-highest priority o! the eight interrupts.

a ar r re-r r r n :. EXTERNI~ ASYNCHROHOrrS Tarr~~~rro~,~
This interrupt is assigned the third-highest priority o! the eight interrupts. The external interrupt (IRQN pin) uses this vector.
SFFFB-FFF9: TIMER INT .RRrro~r This interrupt is assigned the fourth-highest priority of the eight interrupts. It is used by the timer 60.
SFFF6-FFF7: COMPARATOR SrT~ICVCT1~'wr TurTr~flrt~.n 1o This interrupt is assigned the fifth-highest priority of the eight interrupts. It is used by the com-parator subsystem 58.
SFFF4-FFFS: A/D SUBSYSTF~ ldT~'DDrtflT
This interrupt is assigned the sixth-highest priority of the eight interrupts. It is used by the A/D
78.
SFFFZ-FFF3: SERIAL PERrDWFOar. I~~~
This interrupt is assigned the seventh-highest priority o! the eight interrupts. It is used by the SPI
sub~ystaa 54.
SF9'FO-FFF1: INCOM CO lNTI'ITTI~id CONTRO L~ Tarrxronrro~r This interrupt is assigned the lowest priority of the sight interrupts. It is used by the communication controller 29.
5. DATA TRANBpER sum c~nwrronT_ Data tran~far and control functions are imple-~ent~d wing byte-vide register interfaca~ accessed by the microproce~aor 30 in its memory address space as defined in Tabl~ 7.

REGISTER ADD ESS lrLap F ~ Ii~ B

50000 PAD $0010 TSCH 50020 ADCR

50001 PHD 50011 TSCL $0021 AMLTX

50002 PCD 50012 TCR $0022 ADC

50003 PDD $0013 TSR 50023 ACFR

S0004 PAC 50014 TICH $0024 ADZ

SOOOS PHC 50015 TILL $0025 AIrtZ

50006 PCC $0016 TOCH 50026 AVSF

S000~ CMpST S0018 TCRH $0028 ICAFi $0009 CMPI $0019 TCRL $0029 ICAL

SOOOA SPCR SOOlA TARH $002A ICMO

SOOOB SPSR S001H TARL $0028 ICM1 $OOOC SPD $001C NVCR $002C ICM2 SOOOD $001D PWM $002D ICM3 $OOOE $OOlE CFR $002E ICCR

SOOOF $OOlF TEST $002F ICSR

$OFFO DMC

EEPROM CONTROL
The microprocessor 30 controls the operation of the EPROM 40 by a single read-write register NVCR, located in memory address space. FIG. 38 shows the format of this register. Reset clears this register to zero. This will configure the EEPROM 40 !or normal read operation. A de-scription of the bit assignments !or the NVCR register is provided below:
NVCR[7...5]: Unused. These bits are reserved !or device testing.
HVCR[4]: Hyte Erase Select (BYTE). This bit selects byte erase operations. When set, it overrides the row bit. It BYTE is set to a 1, erase operations effect the selected byte. I! BYTE is set to zero, erase opera-tions are either row or bulk.
IdVCR[3]: Row Erase Select (ROW). This bit selects row or bulk erase operations. It BYTE is set, this bit is ignored. It ROW is set to a l, erase opera-tion~ effect the selected row. It ROW is set to a 0, bulk erase is selected.
NVCR[Z]: EEPR014 Erase (ERASE). This bit con-trol~ erase operations in the following manner: It ERASE
i~ sit to a 1, erase mode is selected. It ERASE is set to a 0, noraal road or program mode is selected.
NVCR[1]: EEPROH Latch Control (EELAT). This bit controls EEPROH address and data latch operations as follows: It EELAT is set to a 1, address and data can be latched into the EEPROH 40 for programming or an erase operation. It EELAT is sat to a 0, data can bs read from the EEPROK 40. It an attempt is aada to sat both the EELAT and EEPGZ! bits in the same write cycle, neither will be set.

NVCR(0]: EEPROM Program Voltage Enable (EEPGN).
This bit determines the operating mode of the EEPROM 4o as follows: If the EEPGM is set to a 1, the charge pump 44 is on and the resulting high voltage is applied to the EEPROM array. It EEPGM is set to 0, the charge pump generator is off. If an attempt is made to set both the EELAT and the EEPGM in the same write cycle, neither will be set. If a write to a EEPROM address is performed while the EEPGM bit is sat, the write is ignored and the pro-10~ gramming operation currently in progress is not disturbed.
These two safeguards prevent accidental EEPROM 40 changes.
EEPROM OPT
Specifications for the EEPROM 40 era provided in Appendix A. An internal charge pump 44 avoids the neces sity of supplying a high voltage !or erase and program wing. To reduce programming time, bulk, row and byte erase operations are supported.
The erase state of an EPROM byte is $FF. Pro gramming changes ones to zeros. I! any bit in a location needs to be changed frog a zero to a one, the byte must be erased in a separate operation before it is reprogrammed.
I! a new byte has no ones in bit positions which were al-ready programsed to zero, it is acceptable to program the new data without erasing the EEPROH byte first.
Programming and erasure of the EEPROM 40 relies on an internal high voltage charge pump 44. The clock source for the charge pump 44 is the same as the A/D sub-system and is selected by ACFR[7,6] as discussed above.
Clock frequencies below 2 MHz reduce the efficiency of the charge pump 44 which increases the time required to pro-gram or ara~e a location. The recommended program and era:e time is 10 as when the selected clock is 2 MHz and should be increased to as much as 20 ma when the clock is between 1 l4~iz and 2 MHz. At least 10 ms should be allowed after changing the clock source for the charge pump 44 to stabilize.
The EEPROH 40 operation is controlled by the IdVCR register. Various operations are performed by the G

EEPROM 40 as described below. Other processor operations can continue to be performed during EEPROM programming and erasure provided these operations do not require a read of the data from the EEPROM 40. The EEPROM 40 is discon-nected from the internal read/data bus 34 during program and erase operations.
To read data from the EEPROM 40, the EELAT bit must be zero. When this bit is cleared, the remaining bite in the NVCR register have no meaning or effect and the EEPROM 40 may be read as i! it were a normal ROM.
During EEPROM 40 programming, the ROW and BYTE
bits are not used. The zero bits in a byte must be erased by a separate erase operation prior to programming. The following sequence o! operations is required to initiate a programming cycle as follows:
1. Set the EEL~T bit with EEPGM = 0 2. Store data to the EEPROM memory location 3. Set the EEPGM bit to turn on the high voltage 4. Wait 10 ms 5. Reset both EEPGM and EEhI~T bits to return to normal operation (clear NVCR) The following sequence o! operations is required to initiate a bulk erase o! the EEPROM memory as follows:
1. Set the ER~r.SE and EEL~rT bits with EEPGM = 0 a. Write any data to any EEPROM address 3. Set the EEPGM bit to turn on the high voltage 4. wait 1o ma 5. Reset ERJ~.SE, EEIJ,T and EEPGM bit to return to normal operation (clear IdVCR).
~ row in the EEPROM 40 ie a group o! 16 bytes whose starting address is SXXNO and whose ending address i= SXXNF. The x~s indicate donut care addreaa bits. The N is the row number. This type o! erase operation saves time compared to byte erase operations when large sections 0! EEPROM are to be erased. The sequence of operations required to 'initiate a row erase in the EEPROM 40 is as follows:
1. Set the ROW, ERJ~SE and EEL7~T bits with EEPGM = o 2. Write any data to any EEPROM address in the selected row 3. Set the EEPGM bit to turn on the high voltage 4. Wait 10 ms 5. Reset ROW, ERASE, EELAT and EEPGM bit to return to normal operation (clear NVCR) DE_~D-MAN SUBcvcT~at d~
The dead-man circuitry 46 monitors the micropro cessor 30 for proper operation. This !unction is a mask enabled option that interacts with the microprocessor 30 through a single register (DMC) located at address $OFFO.
The dead-man circuitry may implemented as a 17-bit ripple counter that provides a timeout period o! 32.8 milli-seconds at a bus rate of 4 MHz (262,144 oscillator cycles). I! the counter overflows, a processor reset will occur and the device will be reinitialized.
The dead-man timer is reset by writing a zero to DMC[0). This will reset the counter and begin the timeout period again. The location o! the DMC register was chosen such that a normal bit manipulation instruction cannot re-set the timer. Only extended or indexed, 16-bit offset addrasaing modes can access this location.
r The dead-man subsystem is controlled by a 1-bit register (DMC) located in memory address space. FIG. 39 delinea the register s format.
DMC[0): Dead-man Reset. This write-only bit is used to reset the dead-man timer. writing a zero to it will reset the dead-man counter and restart the dead-man 3o timeout time.
AN~InG SUHSYSTEm I~ gacE R .r:rcT~a Ths microprocessor 30 interlace consists of seven registers (PrDZ, AMZ, AMUX, ~rCSF, AVSP, ADC, and ADCR) located in the memory address space. The format of these registers is shown in FIG. 40.
ADZ: l~/D Auto-zero Value. This 6-bit read-write register contains the o!lset correction value for the voltage input amplifier 80. The ADZ register is 3 .

lowed with the correction value at the completion of an auto-zero sequence. A value of zero represents the in-trinsic positive offset built into the amplifier 80. As the ADZ value increases, the offset decreases. A least-significant-bit represents approximately 0.5 mV offset.
The correction value may be changed by writing to this register. Write operations to the ADZ register are in-tended for diagnostic and verification purposes and are not intended in normal operation. The auto-zero sequence should provide the proper offset value for nominal device operation. At the completion of the auto-zero, the offset of the amplifier 80 should be in the range of 0 to -0.5 mV.
AMZ: Amplifier Auto-zero Value. This 6-bit read-write register contains the offset correction value for the current amplifier 90. The AHZ register will be loaded with the correction value at the completion of an auto-zero sequence. As the AMZ value increases, the off-set decreases. A least-significant-bit represents ap-proximately 0.5 mV offset. The correction value may be changed by writing into this register. write operations to this register are intended for diagnostic and verifica-tion purposes and are not intsnded in normal operation.
The auto-zero sequence should provide the proper offset value for noainal device operation. At the completion of the auto-zero, the offset o! the amplifier should be in the range o! 0 to -0.5 mV.
ACSF: Current Scale Factor. This read-write register i~ used to control operation o! the current input 3o auto-ranging. The value written into this register deter mines the current subsystes auto-ranging operating mode.
It a zero is written, the current aubsystes is placed in auto-ranging mode. A nonzero value inhibits auto-ranging and sets the current mirror 92 into a fixed scale value.
Table 8 detine~ possible values for ACSF write operations.
Values other than these will cause unpredictable opera-tion.

~~.~~~~~8 a This register is not a true read-write register.
The value read from it is not necessarily the value that was written into it. Writing a zero into ACSF enables auto-ranging, however, a zero will never be read from the 5 ACSF register. There are only five possible values that will be read: $10, $08, $04, $02, and $0l.
The value read from this register is one of the scale factors required to properly scale the 8-bit A/D
output. Five values era possible: x1, x2, x4, x8 and 10 x16. Scab factors are shown in Table 8.

CURRENT SCALE FACTOR CONTROL VALUES
~w~ SOFTWARE
ACS j7...01 MODE SCALE FACTOR
15 $00 Auto-ranging enable $10 Divide by 16 x 16 S08 Divide by 8 x 8 $04 Divide by 4 x 4 $0Z Divide by 2 x 2 20 $O1 Divide by 1 x 1 AVSF: Voltage Scab Factor. This read-write register is wed to control operation of the voltage input auto-ranging. The velum written into this register deter-min~a the voltage amplifier 80 auto-ranging operating 25 mode. It a zero is written, the voltage amplifier 8o is placid in auto-ranging mode. A nonzaro value inhibits auto-ranging and seta the voltage .ampliti~r 80 in a fixed gain mode of operation. Table 9 delinu legal values for AVSF writs op~ra~tion~. values other than th~s~ will cause 30 unpredictable operation.
Thia register is not a true read-writs register.
The value read from it is not nsc~s~arily the value that was written into it. writing a zero into AVSF enables auto-ranging, howw~r, a zero will nw~r b~ read from 35 AVSF. Th~r~ are only five possibly values that will be read: 510, $08, $04, $0Z and SOl.

s ~J~~j~3~
~s Q

A/DVOLTAGE ~I~fDT.TFTFD CONTROL VALUES
HARDWARE SOFTWARE
ACSF[7...t~,l MODE SCALE FACTOR
S00 Auto-ranging enable S01 x 16 gain x 1 $02 x 8 gain x 2 $04 x 4 gain x 4 $08 x 2 gain x g $10 x 1 gain x 16 The value read from this register is one of the scale factors required to properly scale the A/D output.
Five values are possible: x1, x2, x4, x8 and x16. Scale factors are shown in Table 9. This register should not be read or written to while a conversion is in progress.
AI~JX: Input Multiplexer Controls. This 8-bit road-write register is used to select the MUXes 66 and 68 connected to the voltage and current input channels 62 and 64. The register is divided into two 4-bit fields; one for controlling the voltage input channels and the other for controlling the current input channels. It is also used to initiate the A/D conversion process. writing to this register will initiate an A/D conversion.
~tJX[3...0]: A/D Channel Select. These four bits control operation of the voltage input channels 62 and 64. Th~ae bits are decoded as shown in Table 10.
Values indicated as "reserved" are dedicated to test and verilication and should not be selected during normal operation. When the currant channels 62 are selected (AIi~JX[3...0] ~ 1000), auto-ranging of the voltage ampli-fier 80 will be inhibited and the gain set to x1. If a nonzero value has previously been written into the AVSF
register, the selected gain will be used instead of an x1 gain factor.

a A/D CHANNEL SELECT DECODE
AI~JX AMUX
f3...01 SOURCE f3. .O1 SOURCE
0000 MUXO 1000 MXO (Current Channel) 0001 MUXl 1001 Reserved 0010 MUX2 1010 Reserved 0011 MUX3 1011 Reserved 0100 MUX4 1100 Reserved 0101 1~TX5 1101 Reserved 0110 MUX6 1110 Temp. Sensor 0111 MUX7 1111 AVSS (0 volts) AMUX[7...4]: Current MUX Select. These four bits control operation of the current input channels 62.
Each bit controls a channel independent of the other three bits. Bits are assigned sequentially with AMUX[4] as-signed to input pin MUXO and AHUX[7] assigned to input pin MUX3. Those bits have no effect it the associated channel is configured !or voltage mode by the CFR register. A
zero in AMUX[7...4] connects the appropriate input pins) to digital ground (VSS), while a one connects the pins) to the current mirror 92 output. The currents can be summed by selecting multiple current inputs. I! all four bits o! this field are zero, no input channels are con-nected to the current mirror 92 output. Since the invert-ing input o! the current amplifier 90 remains connected to the current mirror 92 output, the currant amplifier 90 output will be low, and the current mirror 92 will have no current flowing out o! it.
ADC: A/D CONVERTER OUTPUT This read-only register is used to return the 8-bit output value. The least-significant bit is in ADC[0]. This value must be multiplied by the voltage and current scale factors found in ACSF and AVSF. Depending on the mode o! operation, both scale factors may not be needed:
Voltace In~~uts: The ADC register should be mul-tiplied by AVST for all voltage inputs. The contents of a!
ACSF register should not be used to scale a voltage read-ing.
Current Inputs: The ADC register should be mul-tiplied by the value in the ACSF register and then the AVSF register for scaling of the current subsystem output (MSO) . I! the voltage amplifier 80 is set to auto-rang-ing, the AVSF software scale factor will always be x16, since the voltage hardware gain will be forced to x1 by the selection of MXO.
If an input voltage is converted that is not in the range of AVSS to VREF, the A/D converter will return either $00 (voltages less than AGND) or $FF (voltages greater than VREF). No additional indication is provided.
AD~R: A/D Subsystem Control This byte-wide register i~ used to control operation of the A/D 78. It is implemented as a read-write register to permit read modify-write instructions to properly manipulate bits.
All command bits will read as zero. Control bits will read the current value of the control bit.
ADCR[O]: Unused. This bit is not used. The ADCR[0] bit will always read zero.
ADCR[i]: Sample Inputs. This control bit is used to close the four MUXu 108, 110, 112 and 114 that connect the pairs of channels 102, 104, 106 and 108 to-gather to Ior~ the sample and hold function. The channels are closed when ADCR[i] ~ 1 and open when ADCR[1] = o.
ADCR[i] is set to zero by device reset. Each of the aaaple and hold switches 108, 110, 112 and 1i4 will close only it both channels it is associated with are configured 3o in the voltage mode.
ADCR[Z]: Initiate Auto-Zero Sequence. When this command bit is written with a one, the voltage and current amplifiers 80 and 90 will initiate an autozero se-quence. When the sequence is completed, the ADCR[6] bit will be set to a one. An interrupt will be generated, if enabled, at the completion o! the autozero sequence. The ADCR[Z] bit will always read 'zero'.

Q
ADCR[3]: Integrator Reset. When this control bit is written with a one, the MUX 96 disconnects the MXo pin from the current mirror 92 and shorts lrDCO to analog ground. The MUX 96 will remain shorted as long as this bit remains set. To open the MUX 96 a zero must be writ-t~n to ADCR[3]. This bit will read the present state of th~ I~JX 9 6 .
ADCR(4]: Enable Interrupt. This control bit enables int~rrupts from the A/D subsystem 78. When the ADCR(1] bit is s~t to on~, int~rrupts era enabled. The ADCR(4] bit will read the present state o! the interrupt enable.
ADCR(5]: Aeknowl~dg~ Int~rrupt And Operation Compl~t~. This command bit resets the operation complete flags when written with a one. It will reset ADCR[6...7], removing the interrupt request from the processor.
ADCR(5] should b~ written with a one prior to initiation o! another conversion. This bit will always read as a Z~r0.
ADCR[6]: Auto-Z~ro S~qu~nc~ Complete. This read-only status bit indicates the completion of an auto-z~ro sequence. It will b~ set to a one altar completion o! the auto-zero cycle. Registers ADZ and AHZ will b~ up-dated with the new value o! o!!s~t correction calculated by the auto-zero sequence. This bit is reset by writing to the ADCR(S] bit with a one. The ADCR(6] bit cannot be vritt~n.
~(7]: Conversion Compl~t~. This read-only statue bit indicates the complstion o! an A/D conversion cycle. It will b~ set to a one a!t~r completion o! the 1~/D conversion and indicates that data is available in the ADC, ACSP, and AVSF registers. It is reset Dy writing the ~rDCR(s] bit with a one. This bit cannot b~ written.
AlD SUHSYSTE1~I OPERATION
The A/D subsystem should b~ initialized during the power-up routine. The following initialization opera-tions are required.

so ~ ~r~~
The ACFR register should be written with the ap propriate value to select the proper operating mode of the PiUX4...MUXl inputs. Care should be used when placing an input channel in the current mode, since this will produce a low-impedance on the input pin.
The clock source and divider ratio should be selected with the ACFR['7,6] bits based on the applica-tion~s crystal value. I! the RC oscillator mask option is selected, the clock source should be set to internal ~ (ACFR[7] = 0). The ACFR[5] bit should be written with a one to enable A/D operation.
The control register (ADCR) should be written with an appropriate value. Bits 1, 3 and 4 should be set to establish initial operation conditions. An auto-zero sequence should be initiated by setting the bit ADCR[2]
1. This will cause the offset voltages in the voltage and current amplifiers 80 and 90 to be canceled and the ADZ
and AMZ registers to be sat to the correct values.
The two scale factor registers (ACSF and AVSF) should be initialized. I! auto-ranging is desired, a zero should be written into both registers, otherwise the re quired scale factors should be selected.
OPERATION WITH VOLTAGE INPUTS
To initiate a conversion of a voltage input, the AMUX register should be written with a value that contains the desired input channel in the low-order lour bits and the present currant switch selection in the high-order lour bite. This will start the conversion of the selected voltage input. When the conversion is complete, an in terrupt will be generated (1! enabled) and the ADCR[7] bit will be set. The ADCR(5] bit should be written with a one to clear the interrupt and acknowledge the operation com-plete flag. This will reset the ADCR[7] bit. The conver-sion value is read lros ADC register and then multiplied by the value in the AVSF register to produce a 12-bit value. It should be noted that a voltagt gain factor of x! produces a scale factor of x16. Moreover, it is not intended to write to the AVS! register prior to each con-a version. The ADCR(7] bit must be cleared after every con-version operation by writing to the ADCR(5] bit with a one.
W
To initiate a conversion of a current input, the AMtTX register should be written with a value that contains $8 in the low-order four bits and the present current switch selection in the high-order four bits. This will start the conversion of the 1~c0 input. when the conver-sion is complete, an interrupt will be generated (if en-abled) and ADCR[7] will be set. ADCR(5] should be written with a one to clear the interrupt and acknowledge the operation complete flag. This will reset ADCR[7]. The conversion value is read from the ADC register and then multiplied by AVSF and ACSF to produce a 16-bit value. If voltage auto-ranging has been enabled by writing AVSF with a zero, it is not necessary to multiply the result by AVSF
as long as a 12-bit result is desired. It should be noted that a voltage gain factor of X1 produces a scale factor of X16. As long as AvSF is not written with an overriding gain factor, the X16 scale !actor can be ignored for cur-rent conversions.
There is no need to write AVSF or ACSF prior to each conversion. ADCR[7] must be cleared alter every con version operating by writing ADCR[5] with a one.
The A/D subsystem generates a synchronous inter-rupt at vector address $iFF4-1FF5. The interrupt must be acknowledged prior to resetting the 1 bit in order to not reprocess the interrupt.
QUADCOMPA_~~TOR S YSTEK OPER_~TTON
1. OUADCOI~ARaTeRa SUgSYSTF~
r Four individual inverting comparators are avail-able. The non-inverting input o! each is referenced to +1.25 volts. The comparators are discussed in detail be-3s low. Specifications for the subsystaa are provided in Ap-pendix H. The comparator output states can be read from a register (CI~sT) and can also be directly connected to the least signilicant four output pins of port C. One com-~~.~~~~;

para~or, 200 interrupts on both rising and falling output signals while the other three comparators interrupt only on rising outputs.
The quadcomparator subsystem 58 is controlled by 4 bits of the Configuration Register as defined in FIG.
37.
CFR[3...0]: Comparator Mode Control. These tour configuration bits enable the comparator outputs to be ORed with the least-significant tour bits of port C. A
zero in a configuration bit enables the OR operation for the associated port pin. In this mode, each output pin will be low during device reset it the respective compara-tor input is above the threshold voltage (+1.25 V). A re-set will clear the port C output register making the out-put pin only a !unction of the comparator input. when the microprocessor 30 writes a 1 into this port output regis-ter bit, the corresponding output pin will be forced high independent of the state of the comparator input.
one in those configuration bits disables the OR operation. In this mode, the port pins behave as nor mal port pins. The configuration bits are assigned se quentially, with CFR[0] controlling PCO/CMPO and CFR(3) controlling PC3/C1~3. See Table 11 for assignments.
T1~H_ HL
TOR MODE CONTROL
CFR[3]: PC3/CI~3 CFR[2].: PC2/CMP2 CFR(1] : PCl/Cl~l CFR[0] PCO/CMPO
The cosparator subsystem 58 communicates with the aicroproceesor 30 through a set of two control and status registers (C1KPI and CMPST) located in memory ad-drese space. The state of each comparator output can be read through the CI~ST register. An external interrupt facility is provided to generate interrupts on selected edge o! the comparator outputs. These comparators have approximately ZO mV of hystarasis. FIG. 41 ahow~ the for-mat of these registers.

83 ~~~.~~rd~~
CMPI REGISTER
a.
CMPI[7...4]: Interrupt Acknowledge. These four command bits are used to reset the interrupt request generated by the quadcomparator subsystem 58. They always read as zero. When a one is written into a command bit, the corresponding interrupt request is cleared. These four bits are not read-write registers. The interrupt re quest must be reset prior to clearing the 1-bit to prevent reprocessing the interrupt. Bit assignments are defined in Table 12.

CMPI f 7 . . . 41 BTT acSTC:~t~rre CMPI[4]: CPO
CMPI[5]: CP1 CMPI[6]: CP2 CMPI[7]: CP3 CMPI[3...0]: Interrupt Enable. These four con-trol bits are used to enable the eomparator interrupts. A
one enables a comparator interrupt, while a zero disables it. They are true enables in that transitions prior to the enable will be ignored. Clearing the enable with an interrupt pending will remove the interrupt request.
These lour bits are implemented as true read-write regis-ters. Hit assignments are defined in Table 13.

CMPf3...01 BIT ASSIGNMENT
CMP[O]: CPO
CMP[1]: CPl CMP[2]: CPZ
CMP[3]: CP3 CMPST REGISTER
CI~ST[7...4]: Interrupt Request. These four read-only status bits indicate which comparator inter-rupt s) are active. They are read to determine the cause 0! the microprocessor interrupt. A one indicates an in-terrupt request !or its respective comparator output. Bit auignsents are defined in Table 14.

~~.~J~~

TAB-CMpST[4]: CPO
CMpST[5]: CPl CMpST[6]: CP2 CMpST[7]: CP3 C1~ST[3...0]: Comparator Output. These four read-only status bits indicate the state of the four com parator outputs. A one indicates the comparator output is high and that the comparator input is below the threshold.
Bit assignments are defined in Table 15.

CMPST[O]: CPO
CMpST[1]: CPl CI4pST [ 2 ] : CP2 CIKPST [ 3 ] : CP3 The comparator subsystem 30 generates a syn-chronous interrupt at vector address $FFF6-$FFF7.
2. H+ COMP~~a't'~Q Sn The H+ comparator 50 is discussed in detail be low. This comparator is provided !or power supply genera tion (see FIG. 76H). The negative input o! this compara tor is connected to the VREF pin (+2.5 V nominal). The positive pin is HSENSB. The comparator output is located at HORIVE. Comparator specif ications are provided in Ap pandix H.
3A COI~A_Q~Tnn 49 The A comparator 49 is utilized with the com-munication controller 29 receiver circuits. Both invert-ing (ANEG) and non-inverting [APOS) inputs are available as input pine. The output is AOL1T. This comparator pri-marily operates at input voltages near VR.EF. Specifica-tions !or this comparator are provided in Appendix C.
4. PWX SUecveT~nr 61 A pulse width modulated output 61 may be pro vided by circuitry on output pin PWlt. This output is a periodic signal whose high-to-low ratio is controlled by the 8-bit value stored in the Pwi~i register, The input to the 8-bit pulse width modulator is the processor s phase 2 cloc,~c divided by 4, which results in a PWM period of 0.2778 ms when the 3.6864 MHz crystal is used. The PWM
base frequency is the crystal frequency divided by 1024.
The PwM subsystem is not controlled by the Configuration Control Register.
INTERFACE REGISTERS
The PWM subsystem 61 is controlled by a single 8-bit register (PWM) located in memory address space.
FIG. 41 defines the register's format.
pWN: PULSE WIDTH MOD n.~TflD DaTTn The ratio o! high-to-low signal levels on the PWH pin is determined by the value in the PWM register.
The eight bits o! the PWH are taken as the numerator (N) o! a fraction whose denominator is 256. This traction de-termines what proportion o! the time the PwiK pin will be high. I! N = 0, the PWI~i pin will remain low. I! N = $80, the duty cycle will be 50~. Alter a react, both the PwM
register and the internal counter register will be set to zero and the PWli output will be low. When the PwM regis-ter is written with a non-zero value, the PWM output will go high two PH2 cycles alter the write is completed. The output will remain high for the specified width and then go low !or the remainder o! the PWirt cycle. The output pulse will repeat itsel! continuously within the PWM cycle until a new value is written to the Pwht register. The new pulse width will become valid alter the completion o! the current PWI~t cycle. The Pwlrt register is double-buffered such that a new value written into the PwM register will take elect only at the start o! a Pwlt count sequence.
This will prevent producing spurious output pulse widths.
I! the value written to the PWlri register is zero, the output will stay low alter the current PWH cycle is completed. A zero value in the PwM register will dis-able the PWI~I until a non-zero value is written. Start-up 0! the PWI~t alter a non-zero value i~ written will always be two PHa clock cycles alter completion o! the write to the PWM register. This will ensure the start o! the PWM
cycle at a defined point in time.

('Y r a 6 ~ ~'~ " '' 5. ~ROCt~,tvtwt~gLE TIMER 60 The IC 10 contains a single 16-bit programmable timer 60 with dual output compare registers. The timer is driven by the output of a fixed divide-by-four prescaler operating from the microprocessor 30 phase 2 clock. It can be used for many purposes, including input waveform measurements, while simultaneously generating an output waveform. Pulse widths can vary from several microseconds to many seconds. The timer 60 is also capable of generat-ing periodic interrupts or indicating passage of an arbi-trary number of internal clock cycles. A block diagram of the timer is shown in FIG. 42. Timing diagrams are shown in FIGS. 43A-43D.
Because the timer has a 16-bit architecture, each specific functional capability is represented by two registers. These registers contain the high and low byte o! that function. Generally, accessing the low byte of a specific timer function allows full control of that func tion; however, an access of the high byte inhibits that specific timer function until the low byte is also ac-cessed. The 1-bit in the condition code register should be set while aanipulating both the high and low byte register of a specific timer function to ensure that an interrupt does not occur. This prevents interrupts from occurring between the time that the high and low bytes are accu:ed.
The key element in the programmable timer is a 16-bit free running counter preceded by a prescaler which divides the ~icrocontroller internal phase 2 clock by four. The preacaler gives the timer a resolution of 2 . 00 ~s sousing a crystal frequency of 4 I4tia. The counter is clocked to increasing values during the low portion of the internal phase 2 clock. Software can read the counter at any time without affecting its value.
The double byte free running counter can be read !roe either of two locations: the counter register (TCRi~I, TEL) . or the alternate counter register (TARH, TARL) . A
read sequence containing only a read of the least aignifi-cant~byta of the counter register will receive the counter value at the time of the read. If a read of the counter at either location first addresses the most significant byte, it causes the least significant byte to be trans-ferred to a buffer. This buffer value remains fixed after the first most significant byte read, even if the user reads the most significant byte several times. The buffer is accessed when reading the counter register (TCRL) or alternate counter register (TARL) least significant byte, and thus completes a read sequence of the total counter value. In reading either the counter register or alter-nate counter register, if the most significant byte is read, the least significant byte must also be read in or-der to complete the sequence.
The free running counter cannot be loaded or stopped by the program. During a power-on-reset or device reset, the counter is set to $FFFC and begins running af-ter the oscillator start-up delay. Because the counter is 16 bits and is preceded by a fixed divide-by-four prescaler, the value in the counter repeats every 262, 144 1~PU phase 2 clock cycles. When the counter rolls over from SFFFP to $0000, the timer overflow flag bit (TOF) is set. An interrupt can also be enabled when counter rollover occurs by setting its interrupt enable bit (TOIE) , The programmable timer capabilities are provided by using the following twelve addressable 8-bit registers.
Note that the names high and low represent the signifi cance o! the byte. The format o! these registers is shown in FIG. 44.
The tiger has two 16-bit output compare regis-ters. Each consists o! two 8-bit registers. The primary output coapare register consists o! TOCIi and TOOL, with TOCB the cost significant byte. The secondary output com-pare register is TSCH and TSCL. These output compare registers can be used for several purposes such as con-trolling output wavelorms or indicating when a period of tiae has elapsed. These registers are unique in that all ~xi~~~r bits are readable and writeable and are not altered by the timer hardware. Reset does not affect the contents of these registers, and it the compare functions) are not utilized, the four bytes of these registers can be used as storage locations.
The contents of each output compare register is compared with the contents of the free running counter every fourth rising edge of the phase 2 clock. If a match is found, the corresponding output compare flag (POCF or SOCF) bit is set and the corresponding output level (PLVL
or SLVL) bit is clocked to the appropriate output level latch. The values in the output compare register and the output level bit should be changed after each successful comparison in order to control an output waveform or es-tablish a new elapsed timeout. J,n interrupt can also ac-company a successful output compare, provided the cor-responding interrupt enable bit, OCIE, is set.
l~ttar a processor write cycle to the output com pare register containing the most significant byte (TOCH
or TSCfI), the corresponding output compare function is in hibited until the least significant byte is also written.
The user must write both bytes it the most significant byte is written. ~r write made only to the least signifi-cant byte will not inhibit the compare function. The free running counter is updated every tour internal phase 2 clock cycles due to the internal prescaler. The minimum time required to update the output compare register is a !unction of the software program rather than the internal hardware.
A processor write may be made to either byte of the output compare register without affecting the other byte. 1~ valid output compare must occur before the output level bit becomes available at the output compare pin:
TCMP.
Because neither the output compare flag (POCF
SOCF) or output compare registers are affected by reset, cue gust be exercised when initializing the output com-~~.t~~<f s pare function with software. The following procedure is recommended:
1. Write to the high byte of the output compare register to inhibit further compares until the low byte is written.
2. Read the timer status register to clear the out-put compare flag if it is already set.
3. Write to the low byte of the output compare register to enable the output compare function with the output compare flag clear.
The objective of this procedure is to prevent the output compare flag from being set between the time it is read and the write to the output compare register. A
software example is shown balow.

B6 13 LD~1 TSTAT ARM OCF HIT IF SET
8F 17 STX OCMPLD REl~DY FOR NEXT COMPlrR.E
The two 8-bit registers (TICFI, TILL) which make up the 16-bit input capture register, are read-only and are used to latch the value of the free running counter alter a detined transition is sensed by the input capture edge detector. The level transition which triggers the counter tranater is defined by the input edge bit (IEDG) .
Reset does not affect the contents o! the input capture registers.
The result obtained by an input capture will be ot~ gore than the value of the tree running counter on the rising edge of the phase 2 processor clock preceding the external transition (refer to the timing diagraa shown in FIG. 12). This delay is required for internal syn-chronization. Resolution is affected by the prescaler al-lowing the tiger to only increment every tour phase 2 clock cycles.
The tree running counter contents are trans !erred to the input capture register on the proper signal transition regardless of whether the input capture flag (iCF) is set or clear. The input capture register always contains the free running counter value which corresponds to the most recent input capture.
After a read of the most significant byte (TICH) of the input capture register, counter transfer is in s hibited until the least significant byte of the register is also read. This characteristic forces the minimum pulse period attainable to be determined by the time used in the capture software routine and its interaction with the main program. For example, a polling routine using 10~ instructions such as BRSET, BRA, LDA, STA, INCX, CMPX, and eEQ might take 34 internal phase 2 cycles to complete.
The free running counter increments every four processor clock cycles due to the prescaler.
A read of the least significant byte (TILL) of 15 the input capture register does not inhibit the free run ning counter transfer. Minimum pulse periods are ones which allow software to read the least significant byte and perform needed operations. There is no conflict be tween the read o! the input capture register and the free 20 running counter transfer since they occur on opposite edges of the internal processor clock.
TIMER CONTROL EGICT
The timer control register (TCR) is a read-write register which contains five control bits. Three of these 25 bits control interrupts associated with each of the three flag bits loured in the timer status register. The other two bits control: 1) which edge is significant to the capture edge detector; and 2) the next value to be clocked to the output level latch in response to a successful out-30 put compare. The timer control register and the free run ning counter are the only sections o! the time affected by reset. The output compare pin (SCMP) is forced low during external reset and stay low until a valid coapare changes it. The tiger control register bit assignment is defined 35 in FIG. 44.
TCR[0]: Primary Output Level (PLVL). The value o! the priaary output level bit (PLVL) is clocked into its output level latch by a succusful output compare and will Z~.

a appear on the output compare pin PCMP. PLVL and the pri-many output level latch are cleared by reset. A zero in PLVL produces a low output level on PC1KP.
TCR[1]: Input Edge Polarity (IEDG). The value of the input edge (IEDG) bit determines which level tran sition on the PD7/TCAP pin will trigger a free running counter transfer to the input capture register. Reset does not a!lect the IEDG bit. A zero selects the falling edge.
10' TCR[2]: Secondary Output Level (SLVL). The value o! the secondary output level bit (SLVL) is clocked into its output level latch by a successful output compare and will appear on the secondary output compare pin SC1KP.' SLVL and the primary output level latch are cleared by re-set. A zero in SLVL produces a low output level on SCi~.
TCR[~]: Secondary Output Compare Interrupt En-able (SCIE). It the secondary output compare interrupt enable (SCIE) bit is set, a timer interrupt is enabled whenever the SOCF status flag is set. It the SCIE bit is clear, the interrupt is inhibited. This bit is cleared by reset.
TCR[5]: Timer Overtlow Interrupt Enable (TOIE).
I! the timer overflow interrupt enable (TOIE) bit is set, a timer interrupt is enabled whenever the TOF status flag (in the tiaar status register) is set. I! the TOIE bit is clear, the interrupt is inhibited. The TOIE bit is clured by reset.
TCR(6]: Primary Output Co'pare Interrupt Enable (OCIE). It the priaary output co'pare interrupt enable (OCIE) bit is set, a tiser interrupt is enabled whenever the POCP status flag is set. I! the OCIE bit is clear, the interrupt is inhibited. This bit is cleared by reset.
TCR(7]: Input Capture Interrupt Enable (ICIE).
It the input capture interrupt enable (ICIE) bit is set, a tiaer.interrupt is enabled when the ICF status flag (in the tiger status register) is set. I! the ICIE bit is clear, the interrupt is inhibited. The ICIE bit is cleared by reset.

TIMER STATUS REGISTER
The timer status register (TSR) is a 4-bit register containing read-only status information. These four bits indicate the following:
s A proper transition has taken place at the TCAP
pin with an accompanying transfer of the free running counter contents to the input capture register.
s 7~ match has been found between the free running counter and one of the output compare registers.
The free running counter contains $FFFF (timer overflow).
The timer status register is illustrated in FIG.
44. The timing diagrams shown in FIG. 43 illustrate the timing relationships to the timer status register bits.
TSR[4]: Secondary Output Compare Flag (SOCF).
The secondary output compare flag (SOCF) is set when the primary output compare register matches the contents of the free running counter. The SOCF is cleared by access-2o ing the tiger status register (with SoCF set) and then writing the low byte o! the secondary output compare register. Reset does not a!lect the secondary output com-pare flag.
TSR[5]: Timer Overflow Flag (TOF). The timer overflow flag (ROF) bit is set by a transition o! the free running counter frog SFFFF to S0000. It is cleared by ac cessing the tiger status register (with TOF set) followed by an access o! the free running counter least significant byte. Reset does not a!lect the TOF bit.
TSR[6]: Primary Output Compare Flag (POCF).
The priaary output compare flag (POCF) is set when the priaary output compare register matches the contents of the free running counter. The POCF is cleared by access-ing the ti'ar status register (with POCF sat) and then writing the low byte o! the primary output compare regis-ter. Ruat' does not a!lect the primary output compare flag.

t~.~

TSR[7]: Input Capture Flag (ICF). The input capture flag (ICF) is set when the selected edge has been sensed by the input capture edge detector. It is cleared by a processor access of the timer status register (with ICF set) followed by accessing the low byte of the input capture register. Reset does not affect the input compare flag.
Accessing the timer status register satisfies the first condition required to clear any status bits which happen to be set during the access. The only re maining step is to provide an access of the register which is associated with the status bit. Typically, this pre-sents no problem for the input capture and output compare function.
A problem can occur when using the timer over-flow function and reading the free running counter at ran-dom times to measure an elapsed time. without incorporat-ing the proper precautions into software, the timer over-t low flag could unintentionally be cleared if: 1) the timer status register is read or written when TOF is set;
and 2) the least significant byte of the free running counter is read but not for the purpose of servicing the flag. The counter alternate register contains the same value, as the tree running counter; therefore, this alter-nate register can be read at any time without affecting the timer overflow flag in the timer status register.
During the WAIT instruction, the programmable tiger continues to operate normally and may generate an interrupt to trigger the CPU out of the waft state. The STOP instruction has been disabled in the IC 10.
5.SERIAL PERIPI~IER~ INTERFACE 5~
The serial peripheral interface (SPI) subsystem 54 is designed to provide efficient connection of peri-pheral devices that communicate over a serial bus. It may also be used for interprocessor cosmunication in a multi-processor systea. The SPI supports several versatile operating modes that permit connection of devices that coa~unicate using various protocols. The SPI is essen-tiakly an 8-bit shift register with separate pins for in-coming and outgoing data, a pin for clock, and a fourth pin for device select functions. The following features are supported by the SPI: full duplex, three-wire syn-chronous transfers; master or slave operation; four pro-grammable master bit rates; programmable clock polarity and phase; end o! transmission interrupt flag; write col-lision flag protection; and master-master mode fault pro-tection capability.
The SPI is controlled by three registers SPD, SPSR and SPCR located at memory addresses: $0C, SOB and $01~, respectively.
The SPI can be used in two basic types of sys tems: single master and mufti-master. FIG. 45 illus trates the basic connections for both of these system types. 1~ shown, the MOSI, MISO and SCR pins are all wired to equivalent pins on each o! the four devices. The master device generates the SCR clock, and the slaves all receive it. Slave device selection is accomplished by three port pins wired to the three individual slave select pins on the slave devices. J~ slave device is selected when the master pulls its SSN pin low. l~s the master transmits data on MOSI (output), selected slaves receive it on MOSI (input). Care must be taken if multiple slaves are selected to avoid bus contention on MISO. Only one slave device can drive the MISO pin. ors data is shifted out o! the faster on MOSI, it is shifted in on MISO. If the slave device is receive only, it needs no connection to lIISO.
7~ aore complex mufti-master system is shown in FIG. 46. here the :lava select lines are generated by more than one potential master device. Only one master may control the select lines at any one time. ~n exchange o! master control must be implemented using a handshake method through the I/O ports or by an exchange o! coded messages through the serial peripheral interlace system.
The major device control that is used in this system is the MSTR bit in the SPCR and the MODF bit in the SPSR.

Four pins are connected to the serial peripheral interface subsystem 54. Operation of two of these pins (MOSI and MISO) is dependent on the SPI data pin mask op-tion selected.
MOSI; Master Out Slave In. Operation of this pin is determined by the SPI data pin mask option selected. If the bidirectional option is selected, the MOSI pin is bidirectional and is configured as a data out-put in a master mode device and as a data input in a slave mode device. If the unidirectional option is selected, the MOSI pin is always an output. Data is transferred serially from a master to a slave device on this line;
most significant bit first. The timing diagrams of FIB.
47 show the relationship between data and clock (SCK). As shown, four possible timing relationships may be chosen by using control bits CPOL and CPI~iar. The master device al-ways allows data to b~ applied on the MOSI pin a half-cycle belor~ the clock edgy in order for the slave de-vice s) to latch the data. It should b~ noted that both the master and slave devices must b~ programmed for simi-lar timing modes for proper data transfer.
When the master device transmits data to a slave dwica via the MOSI line, the slave responds by sending data to the aast~r dsvica using its MISO pin if the bi-directional option is selected; or its MOSI pin, if the unidirectional option is selected. This full duplex transmission is synchronized with the same clock edge for both transmission and reception of data. The internal data strobe always uses the opposite phase of the clock from the adg~ used to shift data in and out. The byte transmitted is replaced by the byte r~c~ivad and elimi-nates the need for separate transmit-empty and receiver-full status bits. a singly status bit (SPIF) is used to signify that the I/0 operation is complst~.
Configuration of the MOSI pin is a function of the 1KSTR bit in the SPCR and the s~l~ctsd SPI data pin mask option. It the mask option selected is unidirec-tional, the MOSI pin is always an output. I! the mask op-s tion selected is bidirectional, the MOSI pin is an output when the MSTR bit is a one and an input when the MSTR bit is a zero.
MISO: Master In Slave Out. Operation of this pin is determined by the SPI data pin mask option. If the bidirectional mask option is selected, the MIXO pin is bidirectional and is configured as a data input in a mas-ter mode device and as a data output in a slave mode de-vice. If the unidirectional mask option is selected, the HISO pin is always an input. In this manner, data is transferred serially from a slave to a master, most sig-nificant bit first. when configured as a slave, the MISO
and MOSI pins era placed in the high-impedance state if not selected by a low on the SSN pin. As shown in FIG.
17, four timing ralationships are possible by using the control bits CPOL and CP1~. The master device always ap-plies data on the MOSI line a half-cycle before the selected clock edge on SCR in order for the slave to have adequate data setup time.
When the master mode device transmits data to a slave mode device via the master's MOSI pin, the slave de-vice responds by sending data to the master via the caster's HISO pin. This full duplex transmission synchro-nizes both data out and data in with the same clock edge of the SCl~ supplied by the master device. A single status bit (SPI!) in the SPSR is used to signify that the I/o operation is complete.
In the master mode device, the MSTR control bit in the SPCR is sat to a one by the program to configure 3o the device to receive data on its MISO pin. In the slave device, its MISO pin (or MOSI pin of unidirsctional mask option is selected) is enabled by the low level on the ssN
pin. If SSN is high in a slave mode device, the MISO and I~iOSI pins are placed in the high-impedance atata.
SC1C: Serial Clock. The serial clock is used to synchronize the movement of data both in and out of the device through its MOSI and XISO pins. The master and slave devices are capable of exchanging a data byte of in-formation during a sequence of eight clock pulses. Since the SCK is generated by the master mode device, the SCK
pin becomes an input on all slave devices and synchronizes slave data transfer. The type of clock and its relation-s ship to data are controlled by the CPOL and CPHA bits in the SPCR discussed below. Refer to FIG. 47 for timing.
The master mode device generates the SCK through a circuit driven by the internal processor clock. Two bits (SPRO and SPR1) in the SPCR of the master device select the clock rate. The master device uses the SCK to latch incoming slave device data on the MISO line and shifts out data to the slave mode device on the MOSI pin.
Both master and slave mode devices must be operated in the same timing mode as controlled by the CPOL and CPI~ia bits in the SPCR. In the slave mode device, SPRO and SPR1 have no eftect on the operation of the SPI.
SSN: Slave Select. This low-true input pin is used to enable slave roods device data transfer. To ensure that data will be accepted by a slave mode device, the SSN
pin must be low prior to occurrence of SCR and must remain low until attar the last (eighth) SCR cycle. FIG. 47 il luatratea the relationship between SCR and the data for different coabinations of CPIih and CPOL. When SSN is first pulled low in a slave mode device, the following evanta xcur:
1. The appropriate output pin i~ driven with the first data bit. If the SPI data pin option is bidirectional, this is the MISO pin. If the unidirectional option is selected, this is the MOSI pin.
2. The slave mode device is prevented from writing to its data register when CPII1~ ~ 0.
The description of the wCOL status flag in the SPSR contains more information of the elects that the SSN
input and CPIih have on the I/O data register. 1~ high on SSN in the slave mode device forces the appropriate output pin to the high-impedance state. l~lso, SCR and the appro-priate input pin are ignored by a slave mode device when its SSN pin is high.
Whsn a device is in the master mode, it con stantly monitors its SSN input for a low level. The master device will become a slave mode device any time its SSN input is driven low. This ensures that there is only one master controlling the SSN line for a particular sys-tem. When the SSN pin is detected to be low, it clears the KSTR bit in the SPCR. Also, control bit SPE in the SPCR is cleared, which causes the SPI to be disabled. The HODF flag bit in the SPSR is also sat to indicate to the program that another device is attempting to become a master. Two devices attempting to be masters is normally the result o! a software error; however, a system could be constructed which would employ this error detection to provide a 'backup master' to restart a faulted system.
INTERFACE REGISTERS
The SPI is controlled by three registers: SPD, SPSR, and SPCR. These registers provide control, status and data storage functions !or the SPI. FIG. 48 defines the registers formats.
SPD: Serial Data Register. This 8-bit read-write register is used to transmit and receive data on the synchronous serial bus. only a writs to this register will initiate transmission and reception o! another byte and this will only occur in the master mode device. A
slave mode device writing to its SPD register will not initiate a transmission. At the completion o! transmit-ting a byte o! data, the SPIF status bit is set in both the master and slave's SPSR. A write or read o! the SPD, alter accessing the SPSR with SPIF set will clear the SPIF.
During the clock cycle that the SPIF bit is being set, a copy o! the received data byte in the shift register is moved to a butler. When the program reads the SPR, the butter is actually read. During an overrun con-dition, when the master device has sent several bytes of data and the slave device has not internally responded to a clear the first SPIF, only the first byte is contained in the receive buffer o! the slave device; all others are lost. The program may read the buffer at any time. The first SPIF must be cleared by the time a second transfer of data from the shift register to the read buffer is ini-tiated, or an overrun condition will exist.
A write to the SPD is not buffered and places data directly into the shift register for transmission.
The ability to access the SPD is limited when a transmis-sion is in progress. It is important to read the discus-sion defining the WCOL and SPIF status bits to understand the limits on using the SPD.
SPSR: S$I Status Register. This 3-bit, read only register is used to indicate the operational status of the SPI. Two of these status flags (SPIF and MODF) will generate an interrupt when sat if interrupts are en-abled by the SPIE control bit in the SPCR.
SPSR[7] - SPIF: Transfer Flag. This status flag indicated that a data transfer between the device and an external device has been completed. With the comple tion of the transfer. SPIF is set, and it SPIE = l, an SPI
interrupt is generated. During the clock cycle that sPIF
i~ being ut, a copy of the received data byte in the shift register is moved to the receive buffer register.
When the SPD is read, it is the receive buffer register that is actually read. During an overrun condition, when the master .device has sent several bytes of data and the slave device has not responded to the first SPIF, only the first byte sent is contained in the receive buffer regis tar and all others era lost.
The transfer of data is initiated by the master mode device writing to its SPD. Clearing the SPIF is ac-complished by a software sequence o! accessing the SPSR
while SPI1~ is sat, followed by a write or read of the SPD.
While SPI! is set, all writes to the SPD era inhibited un-til the SPSR i~ read. This occurs in the master device.
In the slave device, SPIF can be cleared using a similar sequence during a second transmission; however, it must be loo Z~.~.~~;.J~Ca cleared before the second SPIF in order to prevent an overrun condition. The SPIF bit is cleared by reset.
SPSR[6] - WCOL: Write Collision. This opera tion flag is set when an attempt is made to write to the SPI data register (SPD) while data transfer is taking place. The transfer continues uninterrupted, and the write operation will be unsuccessful. A read collision will never occur, since the received data byte is placed in the receiver buffer register in which access is always . synchronous with the processor operation. If a write col-lision occurs, WOOL is set, but no SPI interrupt is gener-ated until the transfer has been completed. The WOOL bit is a status flag only.
Clearing the SCOL is accomplished by a software sequence of accessing the SPSR while WCOL is set, followed by 1) a road of the SPD prior to the SPIF bit being set, or 2) a read or write of the SPD after the SPIF bit is sat. A write to the SPD prior to the SPIF bit being set will result in generation of another WCOL status flag.
Both SPIF and WCOL bits will be cleared in the same se-quence. It a second transfer has started while trying to clear the previously set SPIF and WCOL bits with a clear-ing sequence containing a write to the SPD only the SPIF
bit will be cleared.
A collision of a write to the SPD, while an ex-ternal data transfer is taking place can occur in both the master mode and the slave mode of operation, although, with proper programming the master device should have suf-ficient intoraation to preclude this collision. collision in the master mode device is defined as a write of the SPD
while the internal clock (SCR) is in the process o! trans-fer. The signal on the SSN pin must always be high on the master mode device.
A collision in a slave device is defined in two separate modes.
One problem arises in a slave device when the CPHA control bit is zero. When CPHI~r is zero, data is latched with the occurrence o! the first ~ x .""-, ? t.:

clock transition. The slave device does not have any way of knowing when that transition will occur; therefore, the slave device colli-sion occurs when it attempts to write the SPD
after its SSN pin has been pulled low. The SSN
pin of the slave freezes the data in its SPD and does not allow it to be altered if the CPHA bit is zero. The master device must raise the SSN
pin of slave devices between each byte it trans ' fern.
The second collision mode is defined for the state of the CPI~iar equal to one. With CPHA set, the slave device will be receiving a clock (SCK)~
edge prior to the latch o! the first data trans-fer. This first clock edge will freeze the data in the slave's SPD and drive the most signifi-cant bit of the register onto the MISO pin of the slave mode device. The SSN pin low state enables the slave device, but the MISO pin en-able does not take place until the first SCK
clock edge. The WCOL bit will only be set if the SPO is accessed while a transfer is taking place. Hy definition of the second collision mode, a master device might hold a slave's SSN
pin low during a transfer of several bytes of data without a problem.
l~ special case of WCOL occurs in the slave de-vice. This happens when the master device starts a trans-!er sequence ( an edge or SCR f or CPIiar = 1; or an act ive SSN transition for CPIi~ = 0) at the same time the slave device processor is writing to its SPD. In this case, the data byte written by the slave processor is lost and the previous contents of the SPD will be transferred to the caster mode device's SPD. Because the master mode device receives back the last byte transmitted to the slave, the banter device can detect that a fatal collision occurred if the software protocol is appropriately designed.

a Since the slave mode device is operating asyn-chronously with the master device, the wCOL bit may be used as an indicator of a collision occurrence. The soft-ware communication protocol should be designed to accommo-date the collisions that may be generated by this asyn-chronous operation.
The WCOL bit is cleared by reset.
SPSR[4J - MODF: Mode Fault. The function of this flag is defined for the master mode of operation. If the device is a slave, the MODE bit will be prevented from toggling from a zero to a one; however, this does not pre-vent the device from being in the slave mode with MODE
set. The MODF bit is normally zero and is set only when the master device has its SSN pin pulled low. Setting the MODF bit affects the internal SPI subsystem in the follow-ing ways:
1. MODF is set and an SPI interrupt is generated if SPIE = 1.
2. The SPE bit is forced to zero. This blocks all output drive on the SCR, MOSI and MISO pins.
The SPI subsystem is disabled.
3. The MSTR bit is forced to zero, thus placing the device in the slave mode.
Clearing the MODF is accomplished by a software sequence of accessing the SPSR while MODF is set followed by a write to the SPCR. Control bits SPE and MSTR may be restored to their original set state during this clearing sequence, or after the MODF bit has been cleared. The hardware does not allow the prograa to set the SPI and MSTR bit while MODF is a one, unless it is during the proper clearing sequence. The MODF flag bit indicates that there might have been a multi-master conflict for systea control and allows a proper exit frog system opera tion to a reset or default system state. The MODF bit is cleared by react.
SPCR: SPI Control Register. This 7-bit is used to control operation of the SPI subsystea. It is imple-103 ~~~, ,,,G.o .~
' ' !~
mented as a read/write register to permit read-modify-write instructions to properly manipulate bits.
SPCR[7] - SPIE: Interrupt Enable. When this bit is a one, it allows the occurrence of a processor in s terrupt. An interrupt will be generated when either SPIF
or MODF is set. If SPIE is zero, it will not inhibit the setting of these status bits, but an interrupt will not be generated. The SPIE bit is cleared to zero by reset.
SPCR[6] - SPE: SPI Enable. When this bit is set to one, the SPI subsystem is enabled. SPI pins that are defined as outputs are enabled. When SPE is zero, all SPI pin drive is inhibited. This bit is cleared to zero by reset.
SPCR[4] - MSTR: Master Enable. The master en able bit (MSTR) determines whether the SPI is in master or slave mode. It the MSTR bit is zero, the device is in slave mode. It the master mode is selected (MSTR = 1), the function of the SCR pin changes from an input to an output, and the function of the MISO and MOSI pins are re versed. This allows multi-master systems to be con-structed without external logic to reconfigure I/O pins.
The MSTR bit is cleared by reset, placing the SPI in the slave mode t on power-up.
SPCR[3] - CPOL: Clock Polarity. The clock polarity bit controls the normal, or steady state, level o! the clock when data is not being transterrsd. The CPOL
bit affects both the master and slave modes o! operation.
It gust be used in conjunction with the clock phase con trol bit (CPIi~) to produce the wanted clock-to-data rela tionship between a master and slave device. When CPOL is zero, it produces a steady state low on the SCR pin of the master mode device. It CPOL is one, a high is output on SCR when not transferring data. The CPOL bit is not af-fected by reset.
SPCR[2] - CPFiJ~: Clock Phase. This bit controls the relationship between the data on the MISO and MOSI
pins and the clock produced or received at the SCR pin.
This control bit has effect in both the master and slave l04 modes of operation. It must be used in conjunction with the clock polarity control bit (CPOL) to produce the de-sired clock-to-data relationship. The CPH~ bit selects the clock edge used to capture data in the shift register.
If CPHA is a zero, data is captured on the falling edge of SCR. If CPFIA is a one, data is captured on the falling edge. See FIG. 47 for waveform details. CPHA is not af-fected by reset.
SPCR[L-0] - SPRL, SPRO: Bit Rata. These two control bits select one of four communication bit rates to be used as SCK if the device is in master mode. They have no effect in slave mode, since the clock is generated ex ternally by the master device. The slave mode is capable o! shifting data in and out at a maximum rate which is equal to the processor's P2 (bus) clock. Table 16 defines the encoding of these two control bits SPRl and SPRO are not affected by reset.

SPI BIT RATE
Processor Crystal ~p$1 SPRO Clock Divide Hv: Clock Divide By:

6. PORT D
In the non-expanded mode, Port D 57 is an 8-bit bidirectional input-output port. The eight Port D pins can be individually programmed as input or output. In the expanded mesory node, Port D 57 contains the external 8-bit data buy multiplexed with the low-order eight address line.
HON-EXPANDED MODE
When the sxpansion control pin (EXPN) is high, Port D 57 operates as a fully programmable I/O port.
The Port D subsystem 57 communicates with the aicrocontroller through a set of two registers (PDD, PDC) looted in memory address space. The direction of each port bit is determined by PDC, while the state of the port pins is controlled by PDD. See FIG. 49.
PDCf7...01: PORT DIRECT OM
These eight, read-write register bits are used to control the direction o! the corresponding port pin.
The port pin is an input if the port direction bit is taro. At reset, the port direction bits are cleared to zero, defining the port pins as inputs.
PDD f 7 . . . 0 ], : PORT DAT1~
Those eight, read-write register bits are used to read the state o! the port pin it an input, and to con-trol the state of a port pin i! it is an output. A zero corresponds to an electrical low on the port pin. Bits era assigned sequentially, with PDD[0] controlling pin PDO. Device react does not a!lect the data register.
FIG. 50 illustrates the parallel port I/O circuitry.
Those bits era not true read-writs register bits under all conditions. I! the direction is out (PDC[n] = 1), the PDD[n] bit operates as a true read-writs register. I! the direction is in (PDD[n] = 0) , the read source is the port pin, not the port data register bit.
PORT OPERATION
Each o! the eight bits o! the port operates in dependently o! the others. The following paragraphs de scribe the operation o! a single port bit.
Each port pin can be programmed to be either an input or output as determined by the appropriate port direction register bit. A pin is configured as an input i! its associated port direction register bit is set to zero. At power-on or reset, all port direction register bits era cleared, which configure the eight port pins as inputs. when a port direction register bit is sat, the port pin baco~as an output, driving the state o! the port data register bit onto the port pin. A one in the port data register causes a high on the port pin. When the port data register is written, the eight data bits are latched in the port data register.

Z~_~

When the port data register is read, the source of the data is determined by the port direction register as follows:
If the port pin is configured as an output, the read operation data source is the port data register, not the port pin.
~ If the port pin is configured as an input, the read operation data source is the port pin it-self. This prevents read-modify-write opera-tions Eros altering the state of output pins that may be loaded by external circuitry.
Whenever a port pin's direction is changed to output, its data register should be loaded with the de sired output state prior to direction change. Read modify-write operations can altar the state of data regis ter bits configured as inputs.
When the expansion control pin (EXPN) is low, Port D 57 is used to multiplex both the data bus and the low-order eight address bits. The Port D pins change de finition on the falling edge of the OSl signal. They are outputs and contain address information when PH2 is as-serted. When PH2 is negated, they are bidirectional and contain data.
Addresses should be latched on the falling edge o! the address latch enable signal; ALE. The address latch should be implemented with a transparent latch that is transparent while in clock signal is high.
The port pins are driven with output data while PHZ is low. Output data may be latched on the rising edge o! the low-true write strobe signal; WEN. I! the memory cycle is a read operation, the port pins are tri-stated while PHZ is low. External memory devices should use one of the low-true read strobe signal: REN or PSEN, to drive rsad data onto the Port D pins. Read data will be latched on the rising edge o! REN or PEEN. The REN read strobe is used !or memory resources (typically Rte) in the memory address range o! S4000 to $7FFF. The PSEN read r c strobe is used for memory resources (typically PROM) in the memory address range of $8000 to $FFFF. It the read operation is from an internal memory resource, Port D will be driven with the contents of the internal data bus.

In the non-expanded mode, Port B 53 is an 8-bit bidirectional input-output port. The eight Port B pins can be individually programmed as input or output. In the expanded memory mode, Port B contains the high-order eight address lines.
NON-EXPANDEr_~ trtnnp If the IC 10 is in the non-expanded (single chip) mode, operation of this port is identical to the operation of Port D 57. In this mode, port B is an 8-bit bidirectional input-output port. The eight Port B pins can be individually programmed as input or output.
INTERFACE REGTSTERS
The Port H subsystem 53 communicates with the microcontroller through a sat of two registers (PBD, PBC) located in memory address space. The direction of each port bit is determined by PBC, while the state of the port pins is controlled by PBD. See FIG. 51.
PBC[7...0]: Port Direction. These eight read write register bits are used to control the direction of the corresponding port pin. The port pin is an input if the port direction bit is zero. At reset, the port direc-tion bits are cleared to zero, defining the port pins as inputs.
PHD[7...0]: Port Data. These eight read-write register bits are used to read the state o! the port pin it an input, and to control the stets of a port pin if it is configured as an output. A zero corresponds to an electrical low on the port pin. Hits are assigned sequen tially, with PBD[0] controlling pin PBO. Device reset does not a!lect the data register.
EXPANDED MODE
I! the IC 10 is in the expanded mode (EXPN low), Port B 53 is an output port containing the high-order ad-2~~~»~ ~3 dress lines. Addresses will change at 'the rising edge of PIi2 .

8. PORT C
Port C 56 is an 8-bit bidirectional input-output port. The eight Port C pins can be individually pro grammed as input or output. Four can be assigned spe cialized output functions by the configuration register, CFR.
CONFIGURATION REGISTER
Four CFR bits control the function of the low-order four Port C pins as shown in FIG. 37.
CFR[3...0]: Comparator Moda Control. These four write-only configuration bits enable the comparator outputs to ba ORed with the least-significant four bits of the Port C data register: PCD[3...0]. A zero in these configuration bits enables the OR operation and sets the respective port control register (PCC[3...0]) bit to a one. Device reset places the four port pins in output mode, clear: the port data register, and OR's the four data register outputs with the respective comparator out-puts. Thus, the four least significant port pins will be driven low it the comparator inputs are above the thresh-old of +1.25 V. Since the comparator~s inverting input is connected to the comparator input pin,. there will be an inversion between respective comparator input pins and Port C pins. I! a one is written into a Port c data register bit; PCD[3...0], the port pin will be driven high independent o! the comparator input level. A low-order Port C pin cannot be placed in the input mode (PCD[3...0]~0) ft its corresponding comparator mode con-trol (CFR[3...0]) bit is reset.
A one in these configuration bits disables the OR operation. In this mode, the low order four port pins operate a~ normal bi-directional I/O pins and are only a function of the port data register (PCD) and port control register (PCC). The configuration bits are assigned in sequential order with CFR[0] controlling PCO/Cl~o and CFR[3] controlling PC3/C1~3.

2i~9~~~

°' INTERFACE REGISTERS
The Port C subsystem communicates with the microcontroller through a set of two registers (PCC, PCD) located in memory address space. The direction of each port bit is determined by PCC, while the state of the port pins is controlled by PCD. See FIG. 52.
PCC[7...0]: Port C Direction. These eight read-write register bits are used to control the direction of the corresponding port pin. The port pin is an input if the port direction bit is zero. At reset, PCC[7...4) are set to zero and PCC [ 3 . . . 0 ] era set to one. This de fines the low order four port pins as outputs and the high order four pins as inputs. Bit assignments are in ascend ing order with PCC[0] assigned to pin PCO and PCC[7] as signed to pin PC7.
PCC[7...4]. The high-order nibble of the port control register operates as a normal sat of bidirectional port control bits. The following conditions apply:
Reset clears PCC[7...4].
~ writing a zero to a PCC[7...4] bit will cause the corresponding port pin to become an input, with its state readable by the respective bit in the PCD register.
Writing a one to a PCC[7...4] bit will cause the corresponding port pin to become an output, with its state driven by the last state written to the respective bit in the PCO register.
Reading PCC[7...4] will reflect the current state of those bits allowing for bit manipula tion using read-modify-write instruction.
PCC[3...0]. The low-order nibble of the port control register operates differently fros the high-order depending on the state of the configuration control regis-ter bits CFR[3...0]. The following conditions apply:
~ A zero in one o! the lower four bits of the CFR
(CFR[3...0]) will sat the corresponding bit in the PCC.

2~~~~'~~

Since device reset clears CFR[3...0], the low-order nibble of the port control register (PCC[3...0]) will be set after reset.
Writing a zero to one of the lower four bits of the PCC (with the corresponding CFR bit set), will cause the respective port pin to become an input, with the pings state readable in the data register PCD.
s Writing a one to one of the lower four bits of 10~ the PCC (with the corresponding CFR bit set), will cause the respective port pin to become an output, with its state driven by the last state written to the appropriate PCD bit.
Writing a one to one of the lower four bits of the PCC will be ignored it the respective bit in the CFR is clear.
s Reading the lower tour bits of the PCC will re f lect the current state of those bits as stored in the PCC allowing for bit manipulation using read-modify-write instructions.
PCD[7...0]: Port C Data. These eight read/write register bits are used to read the state of the port pins it configured as an input, and to control the state of port pins i! it is configured as an output. A
zero corresponds to an electrical low on the port pin.
Hits are assigned sequentially, with PCD[0] controlling pin PCO. Device reset clears the low order tour bits of the data register: PCD[3...0]. The high order tour bits are not changed by reset.
PCD[7...4]. The high-order nibble of PCD oper-ates as a normal bidirectional port data register. The following conditions apply:
Reset does not affect the upper tour bits of the PCD register.
s ~r read of the upper tour bits of the PCD will reflect the state of the respective port pin if the corresponding PCC bit is clear (input mode).

' ~ A read of the upper four bits of the PCD will reflect the last state of the respective bit in the PCD if the corresponding PCC bit is set (output mode).

9. PORT A
Port A 52 is an 8-bit bidirectional input-output port. The eight Port A pins can be individually pro grammed as input or output. Operation of this port is identical to the operation of Port D in the non-expanded ' mode .
The Port A subsystem 52 communicates with the microprocessor 30 through a set o! two registers (PAC, PAD) located in memory address space. The direction of each port bit is determined by PAC, while the state of the port pins is controlled by PAD. See FIG. 53.
PAC(7...0]: Port Direction. These eight read-write register bits are used to control the direction of the corresponding port pin. The port pin is an input if the port direction bit is zero. At react, the port direc-tion bits are cleared to zero, defining the port pins as inputs.
P11D(7...0]: Port Data. These eight read-write register bits are used to read the state of the port pin i! an input, and to control the state of a port pin if it is an output. A zero corresponds to an electrical low on the port pin. Hits are assigned sequentially, with pAD(0~
controlling pin PAO. Device react does not affect the data rsgiater.
The communication controller 29 (hereinafter identified as ICC) provides the microprocessor 30 access to an INCOM network as defined in detail in U.S. Patent Ho. 4,644,566. It pro-vides modern (aodulator/demodulator) functions, serializa-tion/duerialization o! messages, and implements the re-quired network protocol. The microprocessor 30 communi-cates with the coamunication controller 29 through eight ~~~~?~8 inta~face registers located in memory address space. Four registers are used to transfer INCOM messages between the controller and the microprocessor, while the other four are used to set the communication address, speed, modula-tion method and to control transmit/receive operations of the INCOM communication controller (ICC) 29.
The ICC is capable of operating both as a master and slave controller. Master operation is inhibited un less a permissive flag is set in the configuration regis tar.
The ICC supports the fast status request message that will reduce network response time. The transmit and receive registers for the ICC are independent of each other. This allows a message such as the fast status to be periodically updated in the transmit registers. Then, when a fast status request is received, the ICC can trans-mit the response without processor intervention.
Details of the INCOM network protocol are de-fined below.
w.vavr iyy~f11V1'1 1(LC~ i'~R
The ICC 29 is configured by the CFR and ACFR as shown in FIG. 37.
CFR[7]: ICC Master Enable. This permissive bit allows the ICC Z9 to be switched into the master mode. A
taro in this configuration bit prevents the ICC from en t~ring the master mode of operation. A one permits master operation. This bit is sat to taro on reset inhibiting the ICC frog being placed in master mode.
AC~R(6]: Divider Ratio. This bit selects the clock divider ratio for the A/D and ICC subsystems. If the ICC subsystem is active, either a 7.3728 MHz or 3,6864 MHz crystal must be used. If a 7.3728 MHz crystal is em ployad, ACFR(6] must be set to a one.
INTERFACE REGTSTERB
The microprocessor 30 interface to the ICC
consists of iight registers (ICAH, ICAL, ICM3, ICM2, ICMl, ICNO, ICSR and ICCR) located in memory address space. The toraat o! these registers is shown in FIG. 54.

2~~~2~~

ICAIi, ICAL: Address Registers. These two byte-wide, read/write registers are used set the communication bit rate, modulation method and the 12-bit INCOM address for the ICC. FIG. 54 defines the bit assignments for these registers. Both of these registers are set to zero by reset and power-up. These registers should not be al-tered during normal ICC operation.
ICAIi[7, 6] : Bit Rate. These two bits determine the communication bit rate of the ICC. Table 17 defines the decoding of this field. These bits are set to zero on power-up or reset.

BAUD RATE DECODE
~7.61 ASR OPERATION $OpE~TION
5 0 0 300 bps 19.2 kbps 0 1 1200 bps 38.4 kbps 1 0 4800 bps 76.8 kbps 1 1 9600 bps 153.6 kbps ICAH[5...4]: Modulation Method. These two bits deteraine the modulation method used by the INCOM con-troller. Table 18 datin:.s the meaning of this field.

These bits are set to zero on power-up or reset.

MODULATION METfiOD DECODE
ICAHL .41 1 0 Haseband 1 1 Baseband ICAN[3...0]: INCOM Address Bits 11...8. These tour bits deteraine the high-order tour bits o! the INCOM

address. They are set to zero by rest or power-up.

ICAL[7...0]: INCOM Address Bits 7..Ø This byte-wide register determines the low-order eight bits of the INCOM address. It is set to zero by reset or power-up.
ICN3...ICMO: Message Registers. These tour byte-wide read/write registers are used to transfer INCOH

messages between the ICC 29 and the microprocessor 30.
They are not true read-write registers, since read opera-tions access the receive buffer register, while write operations write to the transmitter buffer register. The same value will not necessarily be read from a message register location that had previously been written to that register location. For this reason, the read-modify-write instructions should not be used to manipulate these regis-ters. The mapping of INCOM message bits is shown in FIG.
55. The transmitter buffer register is loaded when these registers are written. Reset clears the transmitter buffer register to all zeros.
ICM3[7...0]. This 8-bit register contains mes-sage bits 26 through 19.
ICM2[7...0]. This 8-bit register contains mes-sage bits 18 through 11.
ICMl[7...0]. This 8-bit register contains mes-sage bits 10 through 3.
ICMO['7]. This is the control bit 2 of the INCOM
message. The register bit is not a true read-write regis ter, since read operations access the receive buffer re gister, while write operations write to the transmitter butter register. The same value will not necessarily be read troa this bit that had previously been written.
ICMO[6...2]. These Live bits are reserved for test. During normal operation they will contain arbitrary bite patterns that should be ignored by the program.
Writing to thsa will have no effect on the operation of the INCOM subsystu unless the IC 10 is in the test mode.
ICKO[1...0]. These two bits contain the two status bits that are transmitted in 826(ICHO[1]) and B25 (ICMO[0]) o! the reply massage. These bits are imple mented as a true read-write register. The same contents will be read troa them as was previously written. These bits are cleared by reset.
ICSR: Status Register. This byte-wide, read-only register contains ICC status flags that the microcom-puter needs to coaaunicate with the INCOM communication ~1~~ ~~~

contgoller. FIG. 54 defines bit assignments in the status register.
ICSR[7]: Busy. This bit will read as a one whenever the ICC is busy transmitting or receiving a mes s sage on the INCOM network.
ICSR[6]: Interface Enabled. This bit will read as a one when the ICC interface is enabled. An ICC in master mode will always have this bit set, since its in-terface is always capable of transmission.
ICSR[5]: Transmitter Active. This bit will read as a one when the ICC is in the process of transmit-ting a message. The transmission may have bean initiated by software, or by the receipt of a message requiring re-ply or tact status request. A transmission request (ICCR[0] = 1) should not be issued when ICSR[5] = 1.
Under these conditions, the transmission request will be ignored.
ICSR[4]: Fast Status Sent. This bit will read as a one after the completion of the fast status message transmission. It is cleared by reset and by writing ICCR[3] = 1.
ICSR[3]: Transmit Operation Complete. This bit will read as a one after the completion of a message transaission. It is cleared by reset and by writing ICCR[Z] = 1.
ICSR[Z]: Receive Operation Complete. This bit will read as a one after a received message has been loaded into the receive message buffer. It is cleared by react and writing ICCR[i] = 1. The ICC will not begin ra ceiving a new ae~aage until ICSR[Z] = 0.
ICSR(i]: BCH Error. This bit will read as a one alter the receipt of a message containing a BCH error.
It is cleared by reset and by writing ICCR[1] = 1.
ICSR(0]: overrun. This bit will be set if the message register has not been released (ICSR[2] = 1) when a new message is ready to be loaded into the message re gister. This status bit is cleared by reset and by writ ing ICCR[1] ~ 1.

ICCR: Control Register. This byte-wide, read-write register is used to control operation of the ICC 29.
It has been designed as a read-write register to permit read-modify-write instructions to operate on it correctly.
FZG. 54 defines bits assignments in the control register.
This register is implemented with two types of register bits: command and control. Command bits will initiate a !unction when written with a one. They will always read as zero. Control bits can be set and cleared by the soft-ware. They will read the current value of the control bit. All control bits are set to zero by reset.
ICCR[7]: Enable Interrupts. This control bit enables interrupt operation o! the ICC. When set, trans-mit and receive operations will generate an interrupt.
ICCR[6]: Fast Statue Transmit Enable. This control bit enables automatic response to the fast status request message type. When set, the message stored in the transmit butler registers will be transmitted following the receipt of the fast status request. The fast status request message has the control bit set (H2=1), an in-struction field of 3, a command field of 0, and a subcom-mand !field of either 0 or 1. The address must match the ICC addre:a and the massage BCIi must be correct. Only de-vices configured as slaves will respond to the fast status request. i~ihenever the fast status message stored in the transmit buffer is being updated, ICCR[6] should be reset prior to any change in the buffer registers.
ICCR[5]: Master Mode. This bit, when set, places the INCOM communication controller in the master mode of operation. In the master mode, the ICC is permit ted to transit at any time and will receive all messages, independent of their address. Reply and fast status operations are disabled. ICCR[5] may only be sat if C1~R [ 7 ] = 1. Attupts to write a one into ICCR [ 5 j with C!R[7] ~ 0 will fail. This bit is set to zero (slave mode) by reset or power-up ICCR[4]. This control bit is not used.

ICCR(3]: Acknowledge Fast Status Sent. This command bit is used to reset ISCR[4]. ICSR(4] will be re-set whenever ICCR(3] is written with a one. ICCR[3] will always read as zero.
ICCR[2]: Acknowledge Transmit Complete. This command bit is used to reset ICSR[3]. ICSR[3] will be re-set whenever ICCR[2] is written with a one. ICCR[2] will always read as zero.
ICCR[1]: Acknowledge Received Message. This co'mand bit is used to reset ICSR[2]. ICSR[2] will be re set whenever iCCR(1] is written with a one. ICCR[1] will always read as zero.
ICCR[0]: Initiate Transmit. This command bit, when written with a one, will initiate transmission of the massage stored in the transmit butter registers:
ICH3...ICMO. The transmitter must not be buoy (ICSR(5]=o) and the interface must be enabled ( ICSR ( 6 ] ~l ) in order to transmit a message.
INCOM NETWOR_%
All communication on the INCOM network is in 33-bit messages. The massages are asynchronous transmissions that begin with two start bits and terminate with a single atop bit. Both carrier-based and baseband modulation mode: are supported.
It the carrier-based modulation method is selected, two coapatible modulation schemes era possible:
frequency-spilt keying (FSK), and amplitude-shift keying (1181C) .
a 1,.S1C: The amplitude-shift keying modulation scheae uau a 115.2 kHz carrier. A message bit is a one it carrier is present and a zero it carrier is not present.
FSR: The frequency-shift keying modulation scheme a:es two carrier frequencies. A message bit is a one it the carrier frequency is 115.2 kHz and a zero it the carrier frequency is 92.16 kHs.

These modulation methods are compatible since the INCOM controller only uses FSK !or transmission; the receiver's digital demodulator only correlates with the 115.2 kHz carrier. The presence o! the 92.16 kHz carrier is not required for correct demodulation. The use of the second carrier frequency is to provide limiter capture in poorly terminated environments when operated at the higher bits rates. An INCOM controller in FSK mode will properly receive messages transmitted in ASK mode as long as echoes art properly terminated. The interval between messages must be a zero (no carrier). During the following discus-sions, the terms one and zero will be used to denote the state o! the communication line to eliminate confusion be-tween ASK/FSR and baseband modulation.
BASEBAND MODUL7~TTnu I! baseband modulation is selected, the messages are transmitted without carrier in a non-return-to-zero lormat. I! the receiver input (RX) is an electrical low, the message bit ie a zero, and a one i! an electrical high. The transmit output (TX) is an electrical high for a one and electrical low for a zero. The interval between messages must be a zero (RX low). During the following discussions, the terms one and zero will be used to denote the state of the communication line to eliminate confusion between AS1C and baseband modulation.
COMMI1NICATION H T R~T_ES
INCOX networks may be conligurad to operate at a variety of bit rates depending on the transmission mode selected. Table 19 dalinee the possible communication bit rates o! INCOM networks. only one transmission mode and one bit rate may be selected !or a given INCOM network.
TAH
SMISSION MODES wrn BIT RATEc 300 H/S 19.2 RB/8 1200 H/S 38.4 RH/S
4800 b/~ 76.8 kb/s 9600 b/~ 153.6 kb/s a As mentioned above, the ASK and FSK modulation methods are compatible. ASK systems may require network restrictions not applicable to FSK at the higher bit rates.
MESSAGE FORMAT
All INCOH messages are 33 bits in length and have the following characteristics:
s The first two bits era start bits and must be ones.
s The third bit is the control bit that determines the basic message type.
s A five-bit BCH error check code is transmitted starting with the 29th bit.
The last bit of the message is a stop bit that must be a zero.
The bit rata and modulation method can vary de-pending on system options selected, however, the 33-bit, asynchronous message format is common. In the following discussions, the first transmitted bit is numbered B0, and the last bit o! the message (stop bit) is B32. The INCOM
message format is shown in FIG. 55.
START HITS: H0. B1 Each network message begins with two start bits.
These bits are used to frame the message and must be ones.
The interaauage gap is filled with zeros. A new message may start iamediately following the stop bit.
C9~OL HIT: B2 The control bit delines the meaning of massage bite B3 through H26. If BZ is a one, the message is a control message and will be interpreted by the ICC. If B2 is a zero, the message is a data message and will not be interpreted by the ICC. Transmission rights (bus master-ship token) can be exchanged only by control messages (HZ~1).

The meaning of thin massage field is determined by the control bit BZ.

A control bit sat to one indicates that bits B3 through B26 contain instruction, command, sub-command, and address fields that are to be in-terpreted by the ICC.
s ~r zero control bit designates that bits B3 through B26 contain arbitrary data, and are part of a higher-level message protocol. These mes-sages will not be interpreted by the ICC.
BCH ERROR CHECK CODE~ B27-B31 ~ Haaaaga bits B27 through B31 contain a five bit error check calculated using the BCH 31, 26 coda. The ICC
computes a BCH remainder based upon bits B2 through B26 of the 33 bit message. The resulting message (82...H30) will always have as a root the BCIi generator polynomial: XS +
XZ + 1. This error check code has a Hamming distance of 3 and will detect all random double bit errors and all burst errors up to 5 bits in length. Several examples of mes-sages with correct BCH follow:
s -bch- -sc- - addr - -cf- ins c st STOP HIT: B3Z
Each massage ends with a stop bit that is always a zero.
CONTROL I~iESSI~rGES
Messages with the control bit (82) sat (one) are defined as control messages.
INSTRUCTION FIELD: B6-H3 The instruction field consists o! message bits B6 through H9. This field provides for certain primitive slave control functions and also implements the hardware level transmission rights protocol. This field is inter prated as shown in Table 20.

~~.Q~~~~

IN STRUCTION DECODE

INSTRUC TION FIELD D .FTNTTT~IwIe $~ ,~ RUCTION

0 0 0 0 0 Shed Load 1 0 0 0 8 Shad Load With Reply 0 0 0 1 1 Restore Load 1 0 0 1 9 Restore Load With Reply 0 0 1 0 2 Disable Interlace lOV 1 0 1 0 A Disable Interlace with Reply 0 0 1 1 3 Enable Interface 1 0 1 1 H Block Addressing 1 1 0 0 C Reserved, Block Addressing 0 1 0 0 4 Reserved, Block Shad 0 1 0 1 5 Block Restors 1 1 0 1 D Reserved, Universal Addressing 0 1 1 0 6 Reserved 1 1 1 0 E Reserved, Universal Addressing 0 1 1 1 7 No Opsration 1 1 1 1 F No Operation with Reply NETWOR K ADD ESg B22-H11 Message bits B11 through H22 contain the network addre ss in control messages. B11 is the least significant byte of the address. Address comparisons in control mes-sagas depend on the contents of the instruction field.

Host control message nstructions use all twelve bits o!
i addre ss information determine the recipient of the mes-to sage. Certain instructions use fewer than twelve.
BLOCi~ AD
E

D
SS INSTRUCTteNa The following instructions use only B22 through B15 ! the address 0 field in address comparisons:
$4-Block Shed, $C-Reserved, and $5-Block Restore. Thess three in-structions addressing in which the low-order use block tour bits o! address information era ignored. These com-sands can be received by up to sixteen different non-master etwork.
dsvices on the n UNIVERSAL ADDRESS INS't~uttr~rrnurc The following instructions employ no addressing:
SD-Reserved, S6-Scram, and $E-Reserved. These instruc-tions will be received by all devices on the network.
COMMAND FIELD: B7-H n This 4-bit field defines the command in control musages. It is unused in all other control message typo. The definition of the commands defined by this field is determined by a higher-level software protocol and will vary from product to product with the exception of two status commands. The fast status ICC hardware in terprets and executes to status commands defined by a com mand of zero with either subcommand of zero or one. These two status command control messages are universally de fined for all product types as follows:
SINGLE MESSAGE STATUS
A control message with an instruction of $3, command of 0, and subcommand of 0 is defined to be a Single Message Status request. The addressed device is expected to respond as defined below.
EXTENDED STATUS
A control message with an instruction of $3, command of 0, and subcommand of 1 is def fined to be an Ex tended Status request. The addressed device is expected to respond as defined below.
D FIELD: 826-H23 This 4-bit ffield deffines the subcommand in con-trol messages. It is unused in all other control message types. The definition of the subcommands defined by this field is determined by a higher-level software protocol and will vary frog product to product with the exception of subcoaaands zero and one in the status command.
~UL~iICATION NETWORK A_~gTTRITT~11~1 The INCOH network is a multi-drop communication bus on which a number o! devices may transmit. Bus arbi tration is performed by both hardware and software proto cols. The network is arbitrated by a token-passing scheme in which control of bus transmission rights is defined by the message type and message contents. The arbitration protocol assumes a single network controller (network master) that is defined by system configuration. Multiple devices may be capable of performing the network master function, however, only one may be active at any given time.
The network master has several means of dis-tributing bus transmission rights:
Sanding a control message to a slave device re questing a reply. If the message did not re quest a reply, bus transmission rights remain with the network master. If the message re quires reply, the slave will begin to transmit a single reply control message within one bit time of the receipt of the requesting message. In this case, bus transmission rights were granted to the slave for one message only. A slave is not able to transmit a message without receiving a control message requiring reply or having its interface enabled by the appropriate control message.
Sending a control message to a slave controller enabling its interlace. In this case, bus transmission rights era passed to the enabled slave device. The slave may transact as many messages as the software protocol requires. Its intertace will remain enabled until it receives a disable interface control message or detacta a control message (H2 set) to another address.
The software communication protocol determines when bus transmission rights are returned to the network master controller or, possibly, another slave device. The hardware-level arbitration protocol will prevent more than two slave de-vfcu with di!lerent address from having their interlaces enabled at any one time.

a All slave devices conforming to the INCOM stan-dard protocol are required to reply to valid status re-quests. Three status request transaction sequences have been defined. All products era required to support the first type, and can optionally support all three.
All slave devices are required to produce a re ply when they receive control messages that require a re ply. This assumes that the control message contains an address match and correct HCH. The reply status message will have bit H2 sat and contain a two-bit status in bits H25 and H26. Hits H3 through H24 are undefined, but for many implementations, will be an echo o! the status re-quest message. As expected, the HCH will be computed on whatever bit pattern is sent in H1 through H26. The typi-cal datinition o! the two status bits is shown in Table 21.
TAB
~pAgp STATUS DEFINrTTnNc 0 0 O!! or Ready (Normal, Inactive) 0 1 On or Running (Normal, Active) 1 0 Trip (Abnormal, Inactive) 1 1 Alarm (Abnormal, Active) The lollowing INCOH control messages will pro-duce a reply:
0 aaa 0 8 1 Shed Load, Reply 0 aaa 0 9 1 Restore Load, Reply 0 aaa 0 A 1 Disable Interlace, Reply 0 aaa 0 P 1 Status Reply Request Transmission rights are returned to the device that requested the status at the completion o! the reply message.
SINGLE MESSAGE STATUS
So» slave devices are capable o! responding to a single aeseage status request. liicroprocessor-based slaves are generally able to respond to this request. The sing2~a status request message is: "0 aaa 0 3 1". This massage consists of the enable interface instruction, a command of zero, and a subcommand of zero. The slave has two options for its response:
The slave can return a reply message containing two status bits as described above.
The slave can return a product status message as described below.
Transmission rights are returned to the device that requested the status at the completion of the single response message.
PRODUCT STATUS MESSAGE
The product status message has the following format ~ H2: 0 H8-3: 6-bit Division Code H12-9: 4-bit Communication software version B18-13: 6-bit Product ID
H21-19: 3-bit Product Specific Status ~ B26-22: 5-bit Standard Status Code DIVISION CODE
This 6-bit field identifiu the product division or company. The following era exemplary cods:
H8-3 pjY~
00 Reserved O1 Westinghouse Electrical Components (Aaheville) 0~ Westinghouse Breaker Components (Beaver) ~~~rw~R~ vExslc,~
This 4-bit Held may be used to define the coamunication software code version number that the slave product is using.
PRODUCT ID
This 6-bit field may b~ used to define the spa-cific product within the division or coapany code.
PRODUCT SPECIFIC STATUS
This 3-bit field may be used by each product to define status conditions that are unique to the product.

STANDARD STATUS COD
This 5-bit field may be used as the standard status coda as defined in Table 21. Each product should attempt to define four operating states that best match these definitions.
EXTENDED STAT T,C~
Some slave devices are capable of responding to an extended status request. Microprocessor-based slaves are generally able to respond to this request. The ex-tended status request message is: "1 aaa 0 3 1". This massage consists of the enable interface instruction, a command of zero, and a subcommand of one. The slave should respond with two messages:
The tirat message should be a product status message as described above.
The second should be an acknowledge message of the following format: "0 aaa 1 3 1". The 12 address bits should be the address of the slave.
Transmission rights are returned to the device that requested the status at the completion of the second response message.
The INCOH communication controller 29 has two operational codes: Master and Slave. In general, a given application will configure the ICC 29 to operate in only one of those codes; although it is possible to use it with a systea that persits multiple masters on a single com-aunication network. The following avctions describe a typical prograaaing interface to the ICC.
A~ part of the IC l0 initialization software, certain paraaeters must be set in the ICC Z9;
Configuration: Appropriate values must be loaded into CFR(7] (ICC Master Enable) and ACFR(6] (Divider Ratio). CFR[7] must be sat if the application is peraitted to operate in the master mode. I! it is not set, the ICC cannot 12~
be placed in the master mode. ACFR[6] must be set as a function of the crystal frequency.
Communication Parameters: ICAIi[7...4] must be set for the appropriate bit rate and modulation method. These values should not be changed dur-ing normal operation of the IC l0.
INCOM Address: If the IC 10 is configured as a INCOM slave, ICAIi[3...0] and ICAL[7...0} must be loaded with the network address of the slave.
INCOM masters do not need an address.
Mode: If the IC 10 is a master on the INCOM
network, ICCR[5] must be sat. When ICCR[5] is set, the ICC will receive all messages on the network regardless of address. Setting ICCR[5]
will also permit the ICC to transmit massages.
This will be indicated by the interface enabled status bit being sat (ICSR[6]).
Interrupts: I! the ICC subsystem fa to be in terrupt driven, ICCR[7] must be set to enable the interrupt. ICSR[3] and ICSR[2] are used to indicate interrupt requests. They may be polled in non-interrupt driven systems.
Once these parameters have been set, the ICC can function correctly in the communication network.
RECEIVE OPERATION
Operation o! the IC 10 depends on the operating mode o! the ICC 29. The receiver operates di!lerently de-pending on whether it is in the master mode.
MasTER MODE
I! the IC 10 is in the master mode, it will re-ceive all INCOM network messages regardless o! control msssage addressing. In the master mode, its interlace is enabled (ICSR[6] ~ 1) at all times. This will cause all network messages to be received.
SLAVE MODE
I! the IC 10 has been con! igured as a slave, it will only receive control messages that match its address.
Data messages will be received only i! the ICU s interface ~l~~~a~

is onabled (ICSR[6] = 1) when the message is processed. A
slave device's interface is enabled by a specific control message type containing the slave's address. Its inter-taco will be disabled whenever a different slave device is enabled. All control messages with the correct address match will be received.
MESSAGE PROCESSTN-When a message is received by the ICC the fol-lowing events occur:
1. When two start bits are detected, the serial bit stream is framed and shifted into a buffer re-gister.
2. If the message is a control message, the BCH and address are checked and the instruction exe cuted.
3. If ICSR[6] is set, or if the massage is an in-struction message for this INCOH address, the message will be loaded into the receive buffer registers addressed as ICH3 through ICMO. This will only take place i! ICSR[2] is reset. If ICSR[Z] is set, the message will be discarded and ICSR[0] (receiver overrun) set.
While a message is being received, ICSR[7]
(busy) will be read as a one. Alter'the message has been processed, ICSR[Z] will be set indicating that a new mes sage has been placed into the receiver buffer. If an ad ditional ausage that must be loaded into the receiver bu!!sr registers is processed before the previous message hay been acknowledged by the software, a receiver overrun will occur.
I! interrupts are enabled, the setting of ICSR[Z] will generate an interrupt. The software should read the ICC status register and retrieve the new message from the receive message registers. ~rtter the message has been read and status checked for overrun (ICSR[0]), BCH
error (ICSR(1]), and receive operation complete (ICSR[2]), the software should set ICCR[1] to acknowledge the re-ceived message. This will reset ICSR[2...0] and free the receive buffer for the next message. Clearing ICSR(2) will reset the interrupt request.
TRANSMIT

OPERATTnnt The IC 10 software is only permitted to transmit a message on the INCOM
network when the ICC interface i s enabled (ICSR[6] = 1). To transmit a message, the soft-ware should perform the following operations:

1. Reset ICCR[6] to disable fast status operation.

This is required since the transmitter buffer registers are also used for fast status opera-tion.

2. Load the message to be transmitted into the mes-sage registers ICM3 through ICM1. Load ICMO[7]

with the transmitted message control bit. Read-modify-write instructions should not be used for the operation. ICMO[1...0] must contain the correct reply status information at all times during this load operation.

3. Read ICSR[5] to verify that the transmitter is not busy and initiate the transmission by set-ting ICCR[0]. This will cause the transmitter to begin operation. The transmitter buffer re-gisters can be changed as soon as the transmit-ter active bit (ICSR[5]) indicates the transmis-sion is in progress.

The software should poll ICSR[3] for the trans-aittar co'plete flag or wait for the ICC inter-rupt at the end o! the transmission. ICSR[3]

and the interrupt request are cleared by Batting ICCR[2].

REPLY STATUS OP ITT~ftit The ICC generates reply status messages autonoaously in response to INCOM network reply requests.
The reply statue auaage will contain ICMO[1,0] in message bits B26 and B25. The software should reload appropriate values in these two message register bits whenever the ap-plication~s status changes.

FAST STATUS OPERATION
The IC 10 will transmit a fast status reply mes-sage automatically on receipt of the appropriate INCOM
control massage it properly programmed by the application software. In order for the fast status to be transmitted, the software should do the following:
1. Reset ICCR[6j to disable the transmission of fast status.
2. Update the fast status message in the message 10' registers (ICH3...ICMO).
3. Set ICCR[6j to enable the transmission of fast status.
Whenever a fast status request is processed by.
the ICC, the message in the transmitter buffer registers will be sent if ICCR[6j is set. Whenever a new massage is placed in the buffer registers, the ICCR[6j bit must be reset during the load operation to prevent ~ data tearing' o! the new message.
INTERRUPT VECTOR
The INCOK interrupts have the lowest priority in the microcontroller. They are assigned with a vector ad-dress o! $FFFO-FFFi. The interrupts) must be ac-knowledged set ICCR[i, 2 or 3j prior to resetting the I
bit in the processor to avoid reprocessing the interrupt.
~NAt~ SU9SYSTEM SCIiEI~TICS
The analog subsystaa for the IC 10 is illus trated in FIGS. 56-112. More specifically, FIGS. 56-71 illustrate the digital control logic while FIGS. 72-84 il lustrata the analog circuitry. FIGS. 85-112 illustrate 3o the digital logic !or the ICC 29.
DIGITS CONTROL LOGIC
The quadcomparator subsystaa 58 includes tour individual comparators 200, 20Z, 204 and 206 (FIGS. 57 and 74). Each o! these comparators 200, 202, 204 and 206 is reteranced to a predetermined voltage, for example, +1.25 Vdc connected to a non-inverting input (FIG. 74). Input signals are applied to external pins CPO, CP1, CP2 and CP3 illustrated in FIG. 56.
The comparator subsystem 58 communicates with the microprocessor 30 through two registers CMPI and CMpST
located in memory address space. An internal interrupt facility is provided to generate interrupts on selected edges of the comparator outputs Q0, Ql, Q2 and Q3. The comparator outputs Q0, Q1, Q2 and Q3 are read at the data bus DATA[3...0]. More specifically, each of the compara-for outputs Q0, Q1, Q2 and Q3 is connected to a pair of serially coupled inverters 208 and 210 (FIG. 57); 212 and 214; 216 and 218; and 220 and 222, respectively, for high gain. The outputs o! the inverters 210, 214, 218 and 222 are applied to tristate devices 224, 226, 228 and 230.
The outputs of these tristate devices are connected to the data bus DATA[3...0] as CMPST[3...0]. These comparator outputs comprise the CMPST[3...0] status bits. Reading of these status bits is controlled by a read signal RDCMPSTh, which is active any time the microprocessor 30 addresses the C1~ST register and initiates a road as discussed be-low.
The Ci~I register is used for interrupt control.
Hcre spscitically, CMPI [3...0] is used to enable inter-rupts whip ClsPI[7...4] is used to reset the interrupt re-quest generated by the comparator subsystem 58. The in-terrupt request must be reset prior to clearing the I bit to prevent reprocessing the interrupt.
An interrupt request signal INTREQ is generated on s~l~ctad edge of the comparator outputs QO, Ql, Q2 and Q3. More sp~citically, the INTREQ signal is generated on rising and telling output states of the comparator 200 and on rising output states or the comparators 202, 204 and 206. This INTREQ signal is available at the output of a quad NOR gate Z3Z. Comparator interrupt request signals REQOh, REQlh, REQZh and REQ3h are applied to the inputs of the NOR gate Z3Z. These interrupt request signals are available as outputs at an OR gate 234 for the comparator Z00 and Clip-flops 236, 238 and 240 for the comparators ~f0~~~

202,.204 and 206, respectively. More specifically, the interrupt request signal REQOh for the comparator 200 is generated at the output of the dual input OR gate 234.
The inputs to the OR gate 234 are from flip-flops 242 and 244. The output QO of the comparator 200 is applied to the clock input CK of the flip-flop 244 by way of the high gain inverters 208 and 210. The complement of this sig-nal, available at the output of the inverter 208 is ap-plied to the clock input CK of the flip-flop 242. The outputs of the flip-flops 242 and 244 are applied to the OR gate 234 to generate the REQOh signal on rising and falling output states of the comparator 200. The Ql. Q2 and Q3 outputs of the comparators 202, 204 and 206 are ap-plied to clock inputs CK of the flip-flops 236, 238 and 240 to generate the REQlh, REQ2h and REQ3h signals. These REQOh, REQlh, REQ2h and REQ3h signals may be read as sta-tus bits CMPST[7...4] by the microprocessor 30 on the data bus DATA[7...4]. More specifically, the REQOh, REQlh, REQ2h and REQ3h signals are applied to tristate devices 246, 248, 250 and 252. The outputs of these tristate de-vices 246, 248, 250 and 252 are connected to the data bus DATA[7...4]. The tristate devices 246, 248, 250 and 252 are under the control of the RDC1~~STh signal.
Four command bits Ci~I [ 7 . . . 4 ] are used to reset the interrupt request. These command bits CMPI[7...4] are used to reset the flip-flops 236, 238, 240, 242 and 244, which generate the REQOh, REQlh,. REQ2h and REQ3h signals.
These coaaand bits CI~PI[7...4] may be written by way of the data bt~s 011TH[7...4] and are applied to dual input NaND gates 254, 256, 258 and 260 along with a WRCMPIh sig-nal. These NAND gates will only be enabled when the microprocessor 30 addresses the C'MPI register and initi-ates a write. The outputs of the NAND gates 254, 256, 258 and 260 are applied to tri-input AND gates 262, 264, 266 and 268. The outputs of these AND gates are applied to the reset inputs ~ o! the flip-flops 236, 238, 240, 242 and 244. The other two inputs to the AND gates 262, 264, 266 and 268 are the reset signal RESETb from the 2~ ~~?58 micr9processor 30, available at the output of an inverter 270 and interrupt enable signals ENAOh, ENAlh, ENA2h and ENA3h, available at Q outputs of flip-flops 272, 274, 276 and 278. The ENAOh, ENAlh, ENA2h and ENA3h signals allow the interrupt request to be cleared once acknowledged.
The RESE'Tb signal allows the microprocessor 30 to reset these flip-flops. In order to prevent reprocessing of the interrupt, a wRCi~Ib signal, available at an output of an inverter 271, is applied to the b inputs of the interrupt enable flip-flops 272, 274, 276 and 278. These flip-flops are thus reset after the write signal WRCMPIh becomes in-active.
The command bits C1~I [ 7 . . . 4 ] are always read as zero on the data bus DATA[7...4]. Hore specifically, these command bits are read at the output of the tristate devices 278, 280, 282 and 284. The input to these tri-state devices is connected to digital ground. The tri-state devices 278, 280, 282 and 284 are under the control o! a RDCMPIh signal which indicates that the microproces-sor 30 has addressed the C1~I register and initiated a read.
Four status bits CMPI[3...0] are used to read cosparator interrupt enable signals ENAOh, ENAlh, ENA2h and ENA3h. These signals ENAOh, atAlh, ENA2h and ENA3h are available at the Q outputs o! the !lip-!lops 272, 274, Z76 and Z78. These outputs are connected to tristate de-vices 286, Z88, 290 and 292. The outputs o! these tri-state devices are connected to the data bus DATA[3...0].
The tristate devices 286, 288, 290 and 29Z era under the control o! a RDCI~Ih signal which indicates that the sicroprocesaor 30 has addressed the C1~I register and has initiated a read.
The RDC~STh. RDCI~~Ih and WRCKPIh signals are generated by a comparator decode systea 294. The compara for decode subsystea 294 decodes addresses applied to the internal address bus ADDR[4:0] to allow the registers CI~ST and CHPI to be written to and read. More spe-cifically, as illustrated in Table 7 the CliPST register is located at address location $0008. Thus, when the address S0008 is placed on the internal address bus ADDR[4...0) an AND gate 296 (FIG. 58) is enabled. More specifically, the AND gate 296 is an eight input AND gate. The address in-s puts ADDR[4,2,1,0], identified as A4h, A2h, Alh and AOh, are applied through inverters 298, 300, 302 and 304. The address bit ADDR[3], identified as A3h, is applied directly to an input of the AND gate 296. Also applied to the AND gate 296 are signals SELh, DISAHLEb and PH2h from the microprocessor 30.
The hexadecimal address S0008 corresponds to bi-nary bits 00010. When the address 01000 is applied to the address inputs of A4h, A3h, A2h, Alh and AOh, the output of the AND gate 296 will be a logical one indicating that the CHPST register is being addressed by the microproces-sor 30. More specifically, the RDCMPSTh signal is avail-able at an output of a dual input AND gate 306 (FIG. 57).
The inputs to the AND gate 306 are Q outputs of flip-flops 308 and 310. The CMPSTh signal, available at the output of the AND gate 296, is applied to a D input of the flip-tlop 310. Timing for this flip-flop is provided by the 'icroprocessor~s phase 2 clock signal PH2h. More spe-cifically, a PH2h signal is applied to a clock input CK of the flip-flop 310 by way of a pair of serially connected inverters 31Z and 314. An inverted phase 2 clock signal, available at the output o! the inverter 312 is applied to a ~ input of the flip-flop 310. A READh signal, avail-able trosi the microprocessor internal control bus CPOCTL[3...0], is applied to a D input of a flip-flop 308.
The REaIDh signal indicates that the microprocessor 30 is requesting a read operation. Timing signals are applied to the clock C!~ and ZR inputs of the flip-flop 308 in the sage manner as the flip-Clop 310. Thus, whenever the aicroprocessor 30 addresses $0008, the signal RDC'ISPST is generated at the output of the AND gate 306, which indi-cates that the microprocessor 30 is reading the CI~ST re-gister.

- An RDCMPIh signal is available at the output of a dual input AND gate 316. The output of the flip-flop 308 is applied to one input of the AND gate 316 which in-dicates that the microprocessor 30 has initiated a read.
The other input to the AND gate 316 is a Q output of a flip-flop 318. A CI~Ih decode signal is applied to a D
input of the flip-flop 318. Timing control for the flip-flop 318 is identical to the flip-flops 308 and 310. The CI~Ih signal is available at an output of an AND gate 320 (FIG. 53). Tha circuitry including the AND gate 320 and the inverters 298, 300, 302 and 304 generates the CMPIh signal whenever the microprocessor 30 addresses $0009.
The WRCIKPIh signal is available at an output of a dual input AND gate 322 (FIG. 57). One input to the AND
gate 322 is the output of the flip-Clop 318, which indi cates that the CMPI register was addressed. The other in-put to the AND gate 3 2 2 is from a NOR gate 3 2 4 . The NOR
gate 324 is used to develop a microprocessor write signal.
Specifically, the output of the flip-flop 308 is applied to one input of the NOR gate 324. The output signal from the NOR gate 324 will be low during write operations. The other input is troy the phase 2 clock, available at the output o! the inverter 278.
Four configuration bits CFR[3...0] from the con tiguration register CFR are used for comparator mode con trol. These configuration bits CFR[3...0] allow the out puts of the comparators 200, 202, 204, 206 to be ORed with port C. A zero enables the OR operation while a one dis ables it. Hore specifically, the CFR register is a write only register and includes the-tlip-flops 326, 328, 330 and 332. The D inputs of these flip-flops are connected to the data bus DATA[3...0]. The Q outputs of these tlip-tlops are tied to an internal bus CFR[3...0] which allows the OR operation. These flip-flops 326, 328, 330 and 332 are under the control of a dual input NAND gate 334, which enables the write operation. One input to the NAND gate 334 is from the output o! the NOR gate 324 which indicates a write operation. The other input is from a flip-flop 336.. A CFRh signal, which indicates that the microproces-sor 30 addressed the CFR register is applied to the D in-put of the flip-flop 336.
The CFRh signal is a decode signal and is avail s able at the output of an AND gate 338 (FIG. 58). The AND
gate 338 and the inverter 298 decode the address bus ADDR(4...0] to enable the AND gate 338 and generate the CFRh signal any time the microprocessor 30 addresses $OOlE.
The microprocessor 30 is adapted to reset the comparator subsystem 58. More specifically, a reset sig-nal RESETb from the computer control bus CPUCTL[3...0] by way of the inverter 270 is applied to the AND gates 262, 264, 266 and 268 to reset the flip-flops 236, 238, 240, 242 and 244. The RESETb signal is also applied to the flip-flops 272, 274, 276, 278, 308, 310, 318, 326, 328, 330, 332 and 336 to allow the microprocessor 30 to reset the comparator subsystea 58.
2. PROCESSOR HUS INTERFACE LOGIC
The microprocessor 30 communicates with the ana-log control system by way ot, for example, seven registers ADCR, AMUX, ACFR, ADZ, AMZ, AVSF and ACFR located in meaory addrua space as identified in Table 7. The format of the registers is illustrated in FIG. 40. These regis-tars are selected by register select flip-flops 350, 352, 354, 356, 358, 360 and 362, illustrated in FIG. 60. These regi~tar~ are all read-write registers and are decoded by a register decode subsystem 364 illustrated in FIG. 61.
These registers may be decoded in various manners, for exaaple, seven programmable logic arrays (PLA) 366, 368, 370, 37Z, 374, 376 and 378 may be provided. Each of these PLA's includes address inputs ADOR[4..:0] applied either directly or by way of inverters 366, 368, 370, 372 and 374 as shown in FIG. 61 and three control signals SELh, DIS-ABLE and PHZh. The SELh signal corresponds to the micro-processor's AHABSh signal. The ANABSh signal is a regis-ter select signal from the microprocessor master chip ad-dress decoder which enables address decoding on a block basis. The DISAHLEb signal corresponds to the micropro cessor's IOOFF signal, used to disable all I/O devices during a test mode. The IOOFF signal is available at the output of a buffer 375. The PFi2 signal is the micropro cessor's phase 2 clock.
The outputs of the PLA's 366, 368, 370, 372, 374, 376 and 378 represents register select signals ADZh, AHZh, AVSFh, ACSFh, ADCRh, AMUXh and ACFRh indicating that a particular register has been addressed by the micropro-cassor 30. For example it the address $0020 is placed on the address bus ADDR[4...0], the register ADCR will be selected. Similarly, when the addresses of the other registers are placed on the address bus ADDR[4...0] in ac-cordance with Table 7, those registers will be selected.
The output signals from the PLA's 366, 368, 370, 372, 374, 376 and 378 are applied to D inputs of the register select flip-flops 350, 352, 354, 356, 358, 360 and 362. Timing for those register select flip-flops is provided by the phase 2 clock signal PFI2h, applied to the clock inputs C1C of these flip-flops 350, 352, 354, 356, 358, 360 and 362 through a pair of inverters 380 and 382 and an inverted phase 2 clock signal, available at the output o! the inverter 380, applied to the 2!R inputs of these flip-flops. 1~ reset signal RESETh from the micro-processor control bus CPUCTL[3...0] is applied to the re-set inputs ~ o! the Clip-flops 350, 352, 354, 356, 358, 360 and 36Z by way of an inverter 384 to sat these flip-flops to zero on reset. The output of the register select flip-tlop~ 350, 352, 354, 356, 358, 360 and 362 are the register select signals ADCRh, AHU~Qi, ACFRh, ADZh, AHZh, AVSFh and ACSFh.
The ACFR register is a read-write register uti-lized by the A/O subsystu 78. This register includes the flip-flops identified by the reference numerals 386, 388, 390, 392, 394, 396 and 398 (FIG. 60).
The ACFR register can be read or written to by the microprocessor 30. Hore specifically, the D inputs of the .flip-flops 386, 388, 390, 392, 394, 396 and 398 are tied respectively to the data bus DATA[7...0] to allow the microprocessor 30 to write to this register. The output Q
of these flip-flops are also tied to the data bus DATA[7...0] by way of the tristate devices 408, 410, 412, 414, 416, 418, 420 and 422 to allow this register to be read.
During read operations the tristate devices 408, 410, 412, 414, 416, 418, 420 and 422 are under the control of a read control HAND gate 424 and a read-write control flip-flop 426 to allow the Q outputs of these flip-flops to be tied to the data bus DATA[7...0] and read by the microprocessor 30. The tristate device 422 for the ACFR[4] bit has its input tied to ground. Thus, this bit will always read zero.
A read signal is developed by the NAND gate 424.
The NAND gate 424 is a two input NAND gate and is under the control o! the read-write control flip-flop 426 and the ACFR select flip-flop 354. A read signal READh from internal control bus CPUCTL[3...0] is applied to a D input of the read-write control flip-flop 426. Timing for this flip-flop is a phase 2 clock sign PH2h, applied to a clock input C1C and an inverted phase 2 clock signal from the output o! the inverter 380, applied to the ~!C input of a flip-flop 4Z6. The Q output of the flip-flop 426 is a read clock signal RDCLi~, which is applied to the NAND
gate 4Z4. Thus, any time the microprocessor 30 addresses the ACFR register (eg. $0023) and places a read signal READh on the computer control bus CPUCTL[3...0], the flip-flops 386, 388, 390, 392, 394, 396 and 398 as well as the ACFR[4] bit will be read.
During write operations the tristate devices 408, 410, 41Z, 414, 416, 418 and 420 are norsally in a high impedance state. A write control signal is applied to the ~ inputs of these flip-flops. The write control signal is under the control of a write control NOR gate 4Z8 and a HAND gate 430. The NOR gate 428 is a two input NOR gate with a one input from the phase 2 clock PH2h and one input from the read-write control flip-flop 426. The output of the NOR gate 428 is a write signal WRCLKh. The write signal WRCLKh is applied to one input of the dual input HAND gate 430. The other input to the NAND gate 430 is the ACFR register select signal ACFRh. The output of the HAND gate 430 is then applied b inputs of the ACFR
flip-flops 386, 388, 390, 392, 394, 396 and 398. The data bus DATA[7...5] and DATA[3...0] are applied to the D in-puts of these flip-flops to allow the microprocessor 30 to write to them. The bit ACFR[4] is tied to digital ground.
The ACFR register may be reset by the micropro cessor 30. More specifically, a reset signal RESETh from the control bus CPUCTL[3...0] is applied to the reset in puts R of the flip-flops 386, 388, 390, 392, 394, 396 and 398 through an inverter 432.
As heretofore discussed, the ACFR register is a configuration register used to configure the A/D subsystem 78. Thus the ~ outputs of the flip-flops 386, 388, 390, 392, 394, 396 and 398 are connected to inverters 434, 436, 438, 440, 44Z, 444 and 446. The outputs of the inverters 434, 436, 438 and 440 are tied to an internal bus ACFR[3...0]. The outputs of the inverters 444 and 446 are tied to an internal bus ACFR[7,6]. The output of the in verter 44Z i~ applied to an internal bus ACFR[5] and also is used a~ a signal ADPUh.
The RDChRh signal, available at the output of the read write control Clip-flop 426, is used to develop a state machine clock signal SMCLRh for use in auto-zero and auto-range etate machines to be discussed below. The SXCLICh signal is available at the output of a butter 447.
The input to the butler is a MUX 448. The MOX 448 allows for inputs from an external clock source signal CLItSRCh under the control of test circuitry applied to its select input SL. During normal operation, the SMCLR signal is developed by a flip-flop 450. Timing for the Clip-flop 450 i~ an inverted phase 2 clock signal, available at the output of the inverter 380. This flip-flop 450 may be re-set by the microprocessor 30. An output from a HAND gate 452 is applied to a D input of the flip-flop 450. The NAND~gate 452 is a dual input NAND gate. One input to the HAND gate 452 is the RDCLKh signal. The other input to the HAND gate 452 is an output of an OR gate 454. The in-s puts to the OR gate 454 are the ACSFh and AVSFh signals which indicate that the AVSF or ACSF registers have been addressed by the microprocessor 30 to allow the SMCLK sig-nal to be generated when autozero and autoranging opera-tions are initiated.
, ADCR REGISTER
The ADCR register is used to control the opera-tion of the A/D subsystem 78. This register is a byte wide read-write register. This register includes the flip-flops 458, 460, 462, 464 and 466 (FIG. 62). Three.
bits ADCR[5), ADCR[2) and ADCR[Oj era tied to ground and will always read zero. More apecitfcally, the bit ADCR[0]
is tied to ground and to the input of a tristate device 468~ The output of the tristate device 468 is tied to the data bus DATA[O]. The bit ADCR[2] is also tied to ground and to a tristate device 470. The output of the tristate device 470 is tied to the data bus DATA[2). Similarly, the bit ADCR[5j is also tied to ground and to a tristate device 472. The output of the tristate device 472 is con-nected to the data bus DATA[5j.
The balance of the bits may also be read by the microprocessor 30. More specifically, the ~ outputs of the flip-!lops 458, 460, 462, 464 and 466 are coupled to tristate devices 474, 476, 478, 480 and 482. The outputs o! these tristate devices are applied to the data bus DATA[1,3,4,6,7].
The tristate devices 468, 470, 472, 474, 476, 478, 480 and 48Z for all the bits ADCR[7...0] era under the control of the read control NaldD gate 484. Normally these tristate devices are in a high impedance state.
However, during a read operation the NaND gate 484 enables these tristate devices to connect the I~DCR[7...0] bits the data bus DATA[7...Oj. The Nip gate 484 is a two input N~ gate~ A ADCRh signal is applied to one input. This signal is a decode signal for the ADCR register. More specifically, the ADCR register is located at memory ad-dress S0020. Thus, any time this address is written by the microprocessor 30 the ADCRh signal will be active.
The other input to the NAND gate 484 is the RDCLKh signal discussed above. Thus, any time the microprocessor 30 ad-dresses the ADCR register and initiates a read, the NAND
gate 484 will be enabled.
The bits ADCR[1], ADCR[3] and ADCR[4] are con trol bits which may be written by the microprocessor 30.
Hore specifically, the D inputs of the flip-flops 458, 460 and 462 are connected to the data bus DATA[1, 3, 4]. The inputs of these flip-flops are connected to an output of a dual input HAND gate 486. The ADCRh signal is applied to one input of the NAND gate 486 which indicates that the ADCR registers have been addressed by the microprocessor 30. A WRCLt~ signal is applied to the other input. Thus, the NAND gate 486 will be enabled any time the micropro cessor 30 addresses the ADCR register and initiates a 2o write operation.
The bits ADCR [ 7 J and ADCR [ 6 ] are read only sta-tue bits generated by the flip-flops 464 and 466. These bite indicate that the auto-zero sequence is complete and the A/D conversion is complete. These flip-flops 464 and 466 are clocked by the phase 2 clock signal PH2h by way of two inverters 488 and 490. Status signal: EOCh and EOAZh, representative of the status of the A/D subsystem 78 and the auto-zero operation, are applied to the D inputs of these flip-flops 464 and 466 by way o! control circuitry di~cu~aed below. Xore specifically, an end of auto-zero signal EOJ~Zh, which indicates that the auto-zero process has finished ie applied to an ~ input of a flip-flop 492 by way of an inverter 494. The Q output of the flip-flop 49Z is applied to a D input of a delay flip-flop 496. The Q output of the flip-flop 496 is applied to the D input of the flip-flop 464 by way of a buffer amplifier 498. The ~
output o! the flip-flop 464 ii applied to the data bus 2i~~~~8 DATA~(6] by way of the tristate device 480 and a buffer am-plifier 500 to generate an end of auto-zero flag.
An EOCh signal is applied to the flip-flop 466, by way of an inverter 504 and control circuitry discussed below. The EOCh signal indicates the end of the A/D con version process. The output of the inverter 504 is ap-plied to an 3 input of a flip-flop 502. A Q output of the flip-flop 502 is applied to a D input o! a flip-flop 506.
The Q output of the delay flip-flop 506 is applied to the D input of the flip-flop 466 by way of a buffer amplifier 508. The output of the buffer 508 is applied to the D in put of the flip-flop 466. The ~ output of the flip-flop 466 is applied to the data bus DATA[7] by way of the tri state device 482 and a buffer amplifier 510 to generate an A/D conversion complete flag.
Timing for the flip-flops 496 and 506 is an in-vetted phase 2 clock signal available at the output of the inverter 488. The flip-flops 496 and 506 as well as the flip-flops 464 and 466 may be reset by the microprocessor 30 by way of a RESLTb signal available at the output of an inverter 516.
ACFR [ 5 ] is a command bit which resets the auto-zero coaplete and A/D conversion complete flags and resets the bits ACFR[6] and ACFR[7] to remove an A/D interrupt request SYIlb signal from the microprocessor 30. The com-mand bit AClR[5] is available on the data bus DATA[5] and is applied. to one input of a dual input NAHD gate 512.
The other input to the NAND gate 512 is a non-inverting output o! the NAND gate 486 which enables write operations to the ADCR register. The output o! the NAND gate 512 is applied to one input o! a dual input AND gate 514. The other input to the AND gate 514 is the microprocessor re-set signal RESETb available at the output o! the inverter 516. The output o! the AND gate 514 is applied to the re-set inputs 1~ o! the flip-flops 49Z and 50Z to react the coaplete flag and remove the A/D interrupt SY1H.
The A/D interrupt signal SYIlb is generated at an output of a three input NAND gate 516 at the completion of the autozero sequence and the A/D conversion when the A/D interrupt ADCR[4] is enabled. One input to the NAND
gate 516 is an output o! a two input OR gate 518. The in-puts to the OR gate 518 are status bits ADCR[6] and ADCR[7], available at outputs o! bu!lera 517 and 519, which are connected to the Q outputs o! the flip-flops 4s4 and 466. These bits ADCR[7,6] indicate that the autozero operation is complete and the A/D conversion is complete, respectively. Another input to HAND gate 516 is the ADCR[4] bit, which indicates an interrupt enable. The third input is from the test circuitry, normally used only during testing.
The ADCR[2] is a command bit which may be writ ten by the microprocessor 30 and is used to initiate an A/D sequence. This bit is available on the data bus DATA[2] and is applied to dual input HAND gate 520. The other input to the HAND gate 520 is from the ADCR register write control HAND gate 486. The output of the HAND gate 520 to an 3' input of a flip-flop 52Z. A Q output o! the Clip-flop 52Z is applied to a D input of a Clip-Clop 524.
The output of the Clip-Clop 524 is applied to a D input of another flip-tlop 5Z6 by way of a bulter 525. A Q output of the delay flip-Clop 526 is used to generate the start auto-zero signal STAZh. More specifically, the Q output 0! the Clip-!lop 5~6 is applied to one input of a dual in-put AND gate 5Z8 by way of a butler 530. The other input to the AMD gate 5Z8 is from teat circuitry. The output of the A1~1D gate 5Z8 is the STAZh signal.
The ST~rZh signal is cleared when the auto-zero state machine is busy (AZHSYh). More specifically, a AZ8SYh signal is applied to one input o! a two input AND
gate 530 by way o! an inverter 531. A RESETb signal from the microprocessor 30 is applied to the other input. The output of the AND gate 530 is applied to an ~ input of the flip-flop 5Z? to reset this flip-Clop. Timing for the flip-!lops 5Z4 and 5Z6 is provided by the SMCLRh signal available at an output of an inverter 527. The flip-flops 524 and 526 are reset by the microprocessor 30 by way of the RESETb signal applied to their reset inputs ~.
Ths ADCR[1] bit is used to control the four sample and hold switches 108, 110, 112 and 114. Spe cilically the D output of the flip-flop 458 is applied to one input o! a AND gate 532. The other input to the AND
gate 532 is from the test circuit. The output o! the AND
gate 532 is a signal SAMPh which controls the sample and hold switches 108, 110, 112 and 114.
The ADCR[3] bit available at the Q output of the flip-flop 462. This bit ACDR[3] is used to control the integrator reset. More specifically, the output o! the flip-flop 462 is applied to a two input AND gate 534. The othsr input to the AND gate 534 is from the test circuit.
The output o! the AND gate 534 is an integrator reset sig-nal INTRESh. This signal is applied to a bu!!er 757 (FIG.
66) to generate a DISCI3h signal and applied to the switch 96 (FIG. 81). When this signal is high, the switch 96 (FIG. 34) disconnects and MXO pin from the current mirror 92 and shorts the MXO pin to analog ground AVSS. The switch 96 remains shorted as long as this bit remains set.
Ths shorting :witch 96 may be open by writing a zero to ~RI3l. Thin signal is also raprasentatiw o! the pre-sent state o! the switch 96.
2 5 INPUT , An S bit read-write register AMUX(7...0] is used to control the voltage and current input MUXa: 62 and 64.
This register is divided into 2 lour bit fields, one field controls the voltage inputs and generates a signal V!!OX ( 3 . . . 0 j and the other controls the current inputs and generates a signal ClnTX [ 3 . . . O j .
The VMUX[3...Oj signal is developed by the llip-llop~ 536, 538, 540 and 542 (FIG. 62). The D inputs o!
these flip-flops are tied to the data bus DATA[3...0] to allow the microprocessor 30 to write to them. The output o! these flip-flops are tied to the data bus DJ~TA[3...0]
by way o! triatats devices 544, 546, 548 and 550 to allow the content: o! these flip-llop~ to be read by the micro-~~~~w~~

processor 30. The tristate devices 544, 546, 548 and 550 are normally in a high impedance state and are under the control of a HAND gate 552. The NAND gate 522 is a two input HAND gate. A read clock signal RDCLKh is applied to one input. The RDCLKh signal indicates a read operation by the microprocessor 30 as previously discussed. An AMtTXh signal is applied to the other input. The AMUXh signal represents that this register has been addressed by the microprocessor 30. More specifically the AMUX regis-ter is located in memory at S0021. Thus any time the mi croprocessor 30 addresses this location, the AMUXh signal will be active high. This allows the microprocessor 30 to read the contents of the flip-flops 536, 538, 540 and 542 by connecting the Q outputs of these flip-flops to the data bus DATA[3...0].
Write operations to the flip-flops 536, 538, 540 and 542 are controlled by a HAND gate 554. This HAND gate 554 generates a write WRMUXb that is applied to the ~ in-puts of these flip-flops. The NAND gate 554 is a three input NAND gate. The write clock WRCLIQs signal is applied to one input. An Al4UXh signal is applied to another in-put. Lastly, a signal ARBSYh is applied to the NAND gate 554. The ARBSYh signal represents that the auto-ranging aystaa is bu:y and will be discussed in detail below.
The flip-flops 536, 538, 540 and 542 are reset by a RESETb signal, applied to the reset inputs ~ of these llip-flops. The RESETb signal allows the microprocessor to reset these flip-flops.
The CRUX[3...0] signal is developed by the flip 30 llop~ 556, 558, 560 and 562. The D inputs o! these flip flops are tied to the data bus DATA[7...4] to allow the microprocessor 30 to write to them. The Q outputs of these flip-flops are connected to the data bus DATA[7...4) by way o! tristate devices 564, 566, 568 and 570 for read operations. The tristate devices 564, 566, 568 and 570 are normally in a high impedance state and are under the control o! the HAI~iD gate 552 which allow: these flip-flops to be read when the microprocessor 30 initiates a read J r:r operation and places the address $0021 on the address bus ADDR[4...0]. Write operations to the flip-flops 556, 558, 560 and 562 are controlled by the HAND gate 554 in a simi-lar manner as the flip-flops 536, 538, 540 and 542.
The outputs of the flip-flops 536, 538, 540, 542, 556, 558, 560 and 562 are used to generate a MUX con-trol signal I~iJXCTL[26...0] to control the current and voltage IKLTXes 66 and 68 as shown in FIGS. 34 and 73. More specifically, the voltage channel MUXes 66 and 68 are con-trolled by the control signal VMUX[3...0] while current channel MUXes 66 are controlled by the CMUX[3...0] signal.
These signals are decoded by a series of inverters, oR
gates and AND gates (FIG. 60) to generate the MUX control signals MUXCTL to allow individual control of each of the voltage and currant channel MUXes. More specifically, the c~L?X[3...0] signal is applied to one input of dual input AND gates 572, 574, 576 and 578.. The other inputs to these AND gates is a signal CaZh, which is applied to the AIdD gates 572,574,576 and 578 by way o! an inverter 580.
The signal CalZh indicates that the current amplifier 90 is being auto-zeroed. The signal C~rZh is also used to develop a signal CSHItTh, used in the A/D subsystem 78.
The signal CSI~tTh is available at the output o! an in-verter 581, serially coupled to the inverter 580.
The output o! the N1~ND gate 572 is coupled to a chain of serially connected inverters 582, 584, 586, 588, 590, 59Z and 594. The output o! the NJ~ND gate 574 is coupled to s chain of serially connected inverters 596, 598, 600, 60Z, 604, 606 and 608. Similarly, the output of the N71ND gate 576 is coupled to a chain of serially con=
nectsd inverters 610, 61Z, 614, 616, 618, 620 and 622.
Lastly, the NJ~IdD gate 578 is applied to a chain o! seri-ally connected inverters 624, 626, 628, 630, 632, 634 and 636. The outputs o! the inverters 594,608, 62Z and 636 are coupled to inputs o! dual input OR gates 638, 640, 642 and 644. The other inputs to these OR gates are the out-puts lroa the inverters 582, 596, 610 and 624. The out-puts o! the N~ND gates 572, 574, 576 and 578 are applied 147 2~~~;~~~
to inputs of dual input OR gates 646, 648, 650 and 652, respectively. The other inputs to these OR gates are from the outputs of the inverters 592, 606, 620 and 632, re-spectively. The outputs of the OR gates 638, 640, 642, 644, 646, 648, 650 and 652 are applied to inputs of dual input AND gates 654, 656, 658, 660, 662, 664, 666 and 668, respectively. The ACFR[3...0] bits from the internal bus ACFR[3...0] are applied to the other inputs of these AND
gates to control whether the input MUXes 66 (FIG. 34) are in a current mode or a voltage mode. A one in these bits selects the current mode. More specifically, the ACFR[3]
bit is applied to the inputs of the AND gates 654 and 656.
The ACFR [ 2 j bit is appl ied to the inputs of the AND gates 658 and 660. The ACFR[ 1 j bit is applied to the inputs of the AND gates 608 and 610. Lastly, the ACFR[Oj bit is ap-plied to the inputs of the AND gates 666 and 668. The outputs of the AND gates 654, 656, 658, 660, 662, 664, 666 and 668 are decode signals MUXCTL[20...13] which allow for individual control of the current channel MUXes 66.
The sample and hold switches 108, 110, 112 and 114 may also be individually controlled by decode cir-cuitry (FIG. 65j which includes inverters 670, 672, 674, 676, 678 and 680 and three input AND gates 682 and 684.
Hore specifically, the inverters 670, 672, 674 and 676 are tied to the internal ACFR bus ACFR[3...0]. The outputs of the inverters 670 and 672 are applied to the AND gate 682 along with .a SAI~h signal, available at its output of the N71ND gate 53Z (PIG. 62) , which indicates that the command bit AChR[1] has been set. The outputs of the inverters 674 and 676 are applied to the AND gate 684 along with the SA~h signal. The SAIKPh signal is also tied to the in-verters 678 and 680 and the AND gate 684 represent the signal l~J7cCTh[24...21] which allows individual control of the sample and hold switches 108, 110, 112 and 114.
The V1~JX[ 3 . . . 0 ] bits of the AMUX register con-trol the vortage channel I~iU)ces 66 and 68 to generate the I~JXCTL[ 10. . .0] signals and to generate the 1~J7CCTL[ 10. . . o ) signals (FIG. 66) . The I~iU)CCTL[26, 25, 12, iij signals 148 ~~~~~~8 (also shown on FIG. 61) are used during a testing mode.
More specifically, the voltage channel MUXes 66 and 68 are selected by the VMUX[3...0] bits. These four bits are de-coded by circuitry (FIG. 66) consisting of buffer ampli-fiers 686, 688, 690 and 692; inverters 694, 696, 698 and 700; quad input AND gates 702; 704, 706, 708, 710, 712, 714, 716, 718, 720 and 722; and dual input AND gates 724, 726, 728, 730, 732, 734, 736, 738, 740, 742 and 744. More specif ically, the VI4LTX [ 3 . . . o ] bits are appl ied to the in-puts of the buffer amplifiers 686, 688, 690 and 692 and the inverters 694, 696, 698 and 700. The output of the buffer amplifier 686 is applied to the inputs of the AND
gates 718, 720 and 722. The output of the inverter 694 is applied to the inputs of the AND gates 702, 704, 706, 708, 710, 712, 714 and 716. The output of the buffer amplifier 688 is applied to the inputs o! the AND gates 710, 712, 714, 716, 718 and 722. The output of the inverter 696 is applied to the inputs of the AND gates 702, 704, 706, 708 and 720. The output of the buffer amplifier 690 is ap-plied to the inputs of the AND gates 706, 708, 714, 716, 718 and 72Z. The output of the inverter 698 is applied to the inputs o! the 1,ND gates 702, 704, 710, 712 and 720.
The output o! the butler amplifier 692 is applied to the inputs o! the l~rND gates 704, 708, 71Z, 716 and 722.
Lastly, the output o! the inverter 700 is applied to the inputs o! the J,ND gates 70Z, 706, 710, 714, 718 and 720.
The output o! the 7,ND gates 702, 704, 706, 708, 710, 712, 714 , 716 , 718 , 7 Z 0 and 7 2 Z are a decode o! the A14LJ7C [ 3 . . . 0 ]
bits. The outputs o! these AND gates era applied to the inputs o! dual input PrND gates 724, 726, 728, 730, 732, 734, 736, 738, 740, 74Z and 744. The other input to these 1,ND gates is an 1~ND gate 746 which controls write opera-tion to the voltage channel I~itTXea. Hors specifically, the J~HD gate 746 is a two input AND gate. Pr WRHtTXb signal is applied to one input. The other input to the AND gate 746 is a signal VI~Zh by way of an inverter 748. Pre will be di:cussed below, the signal vl~Zh is active high and indi-cates when the voltage amplifier 80 is being zeroed. This signal will be discussed in detail below. The outputs of the~AND gates 724, 726, 728, 730, 732, 734, 736, 738 and 740 represent the signal MUXCTL[9...0].
As will ba discussed in more detail below, auto s ranging is inhibited when the current channels are selected. Thus the output of the AND gate 720 is used to generate a signal I~cOSELh by way o! an inverter 750. More specifically, a signal MXOh, which indicates that a cur rent channel has bean selected, is generated at the output 0! an AND gate 752 and corresponds to an MUXCTL[9] signal.
The AHD gate 752 is a two input AND gate. One input is from the test circuit. The other input is from an OR gate 754. The OR gate 754 is a two input OR gate. One input is from a test circuit. The other input is from the AND
gate 742 which indicates that the current channels have been selected.
A VNULLh signal is developed at the output of a bu!!er amplifier 756. This VNULLh signal controls the zero switch 86 (FIGS. 62 and 80) for auto-zeroing the voltage amplifier 80. The input to the bu!!er amplifier 756 is an output o! a two input AND gate 758. Ona input to the AND gate 758 is the VAZh signal indicating that the amplifier 80 is being autozeroed. The other input to the AND gate 758 is the same signal VAZh applied by way o! in verters 748 and 760.
An 7lGNDh signal is developed at the output o! a two-input OR gate 761. one input to the OR gate 761 is the output lro~ the AND gate 744. The VNULLh signal is applied to the other input. The AGNDh signal is used to connect the voltage channels to ground (FIG. 34). The AGNDh signal is available as IdtTXCTL[lOJ.
Lastly. I~JXCTL[26,25,12,11) are used in conjunc-tion with the test circuit.
ALTO-2ER0 REGTSTERS At~Z AND A14Z
The ADZ and AI~IZ registers are used in connection with the auto-zero logic 98 !or the current and voltage amplifiers 80 and 90. The auto-zero logic corrects for the offsets in the amplifiers 80 and 90 created by using a CMOS process to fabricate these amplifiers.
ADZ REGISTER
The ADZ register (FIG. 67) is a 6 bit read-write register which contains the offset correction value for the voltage amplifier 80. As previously discussed, write operations to this register era intended for diagnostic and verification purposes only. This register is loaded with the correction value for the amplifier 80 at the com . pletion at the auto-zero sequence.
More specifically, the data bus DATA[7...0] is applied to this register, which includes the flip-flops 762, 764, 766, 768, 770 and 772. MUXes 774, 776, 778,.
780, 782 and 784 allow the inputs of these f lip-f lops to be connected to either the data bus DATA[7...0] or to a zero bus 2ER0[5...0]. The ZERO[5...0] bus contains the offset correction value for the amplifier being zeroed and allows the offset correction value to be written to the ADZ register. The data bus DATA[5...0] bits era applied to the A inputs of the MUXes 774, 776, 778, 780, 782 and 784. The ZERO[5...0] bus signal is applied to the B in-puts of these MtJXes. A ZERSEL signal is applied to the select inputs SEL of these MLJXes. The ZERSEL signal con-trols whether the ADZ register is loaded from the data bus or the zero bus. The outputs of the IKIJXes 774, 776, 778, 780 and 782 are applied to the D inputs of the flip-flops 762, 764, 766, 768, 770 and 772. The ZERSELh signal (FIG.
69), generated by the autozero state machine, indicates that the state machine is in state S3. In state s3, the offset correction value is latched into the ADZ register by the VZCLIas signal discussed below. The VZCLRh signal is developed by the auto-zero state machine and is avail able at the output of a buffer 803. This signal vZCLKh is applied to the ~ inputs of the flip-flops 762, 764, 766, 768, 770 and 772.
The Q outputs o! the flip-flops 762, 764, 766, 768, 770 and 772 are applied to an internal bus VZERO [ 5 . . . 0 ] !or the voltage ampl if ier 80. The Q outputs of these flip-flops are also connected ~ ~he ~a~a bus DATA[5...0] by way of tristate devices 786, 788, 790, 792, 794 and 796 to allow these flip-flops to be read by the microprocessor 30. The bits ADZ[7,6] will always read zero since these bits are tied to the data bus DATA[7,6]
by way of tristate devices 798 and 800 which have grounded inputs.
The tristate devices 786, 788, 790, 792, 794, 796, 798 and 800 are normally in a high impedance state except during a read operation when these tristate devices allow the outputs of the flip-flops 762, 764, 766, 768, 770 and 772 to be tied to the data bus DATA[7...0]. These tristata devices are under the control of a NAND gate 802.
The output of the NAND gate 802 is a signal VZRDb signal, which represents a ADZ register read. The inputs to the NAND gate 802 are the signals RDChIQi and ADZh. The read signal RDCLIQi is the read clock signal. The ADZh signal represents a microprocessor 30 has placed the address 50024 on the address bus ADDR[4...0]. The ADZ register is loaded by a VZCLKh signal.
The ADZ register can also be reset by the micro-processor 30. More specifically, a RESb signal is applied to the reset inputs of the flip-flops 762, 764, 766, 768, 770 and 77Z.
The AHZ register is a 6 bit read-write register that contains the olfset correction value for the current airror asplitier 9Z. This register is loaded with the correction value at the completion o! the auto-zero se-quence. write operations to this register are diagnostic and verification purposes only.
The AHZ register includes the flip-flops 804, 806, 808, 810, 81Z and 814. The Q outputs o! the flip-llops 804, 806, 808, 810, SlZ and 814 are tied to an in-ternal bus CZERO(5...0].. The data bus DATA(7...0] is ap-plied to the inputs of these flip-!lops by way o! the Ii~JXes 774, 776, 778, 780, 78Z and 784. These flip-flops are clocked by a CZCLRb signal, developed by the auto-zero ~~~3~~~8 state machine to latch the offset correction value for the current amplifier 90 (FIG. 34) into the AMZ register in state S7 (see Appendix C) as discussed below. The CZCLKb signal is applied to ~ inputs of these flip-flops by way of a buffer 816. A RESb signal is applied to the reset inputs R of these flip-flops for reset. The Q outputs of the flip-flops 804, 806, 808, 810, 812 and 814 are applied to the data bus DATA[5...0] by way of tristate devices 818, 820, 822, 824, 826 and 828 The bits AMZ[7,6] are not used and will always read zero. More spacilically, the data bus DATA[7, 6] are connected to the outputs of tri-state devices 830 and 832, respectively. The input to the tristata devices 830 and 832 are connected to ground.
Thus the AMZ[7, 6] bits will always read zero.
All of the tristate devices 818, 820, 822, 824, 826, 828, 830 and 832 are under the control of a NAND gate 834. The output o! the HAND gate 834 is a signal CZRDb, which indicates an AMZ register read. The RDCLI~ is ap-plied to one input of the NAND gate 834. An AMZh signal is applied to the other input. The AMZh represents that the microprocessor 30 has written to the address $0025.
The flip-flops 804, 806, 808, 810, 812 and 814 are reset by the sicroprocessor 30. More specifically, a RESb signal is applied to the reset inputs of these flip flops.
AUTO-ZERO STATE MACHINE
An auto-zero state machine ~gensratas the se-quencing required to zero the voltage and current ampli-fiers 80 and 90 by adjusting the internal bias currents.
During an autozero sequence, the amplifier's inputs and output are isolated from associated circuitry and the in-puts are grounded. The autozero state machine differen-tially varies the bias currents in discrete steps by way o! current divider (FIG. 83) until the amplifier's output changes states. The autozeroable amplifier is illustrated in FIG. 84.' The nusber o! steps corresponding to the change in state represents the oflsat correction value.

This correction value stored in the ADZ and AMZ registers is discussed above.
The auto-zero state machine is illustrated in l~ppendix C, Figure C-1. An auto-zero state transition table, state diagram and transition table state equations are provided in Appendix C.
The state machine includes three state register flip-!lops 836, 838 and 840 as wall as NAND gates 842, 844, 846, 848, 850, 852, 854, 856, 858, 860, 862, 864 and 866 and as shown in FIG. 69. There are also various in-puts to and outputs from the state machine as discussed below. The state register flip-!lops produce state vari-ables ROh, ROb, Rlh, Rlb, R2h and R2b. These variables, along with the variables ROd, Rld and R2d, available at the outputs o! the NPrND gates 848, 856 and 866 are used in the development o! the state equations in Appendix C.
Each o! the state register flip-!lops is clocked by the SHCLIQ: signal discussed above. The reset signal RESETb, available at the output o! an inverter 892 is ap plied to the reset inputs R o! these flip-!lops.
The outputs o! the state register flip-!lope 836, 838 and 840 define the auto-zero state assignments !or the auto-zero state machine in accordance with Table Cl, Appendix C. Eight permissible states SO-S7 are ds lined as lollovs:
SO - Idle. The state machine is in the idle state waiting !or a start auto-zero signal STAZh to become active. The stag machine also stays idle as long as the auto-range state aachine is busy. This interlock between the two state aachines is implemented by an auto-range busy signal aR8SYh. When the STI~Zh signal is active, the 1~R8SYh signal is inactive to allow the state machine to transfer to state Sl.
S1 - Five microsecond delay. States S1 and S2 lots a loop which is repeated until either the voltage am plitier output changes stag or a counter 868 ruches a final count. During state S1, a tiae request signal TII~tEQh is active and a time out signal TIHOUTh is moni ~~.~~~~8 toted. The act of the time request signal going active, which happens in state S1, triggers a five microsecond timer (FIG. 68). When the timer has timed five micro-seconds, the time out TIMOUTh signal becomes active. This causes the state machine to transfer to either state S2 or state S3. The state machine proceeds to state S2 if the amplifier output VAMph signal is still high indicating that the bias current is not yet high enough and the counter 868 has not yet reached the final count as indi-Gated by a FULLb flag. The state machine transfers to state S3 i! either the amplifier output signal goes low or the counter 868 reaches the final count. The five microsecond delay permits the amplifier 80 to settle and reach a stable output.
S2 - clock counter. State S2 is entered when the counter 868 does not yet contain a full count and the count is not enough to generate a suitable bias current for zeroing the amplifier 80. The clock counter signal is active in state S2 incrementing the count value by one.
The state machine always transfers to state S1 on the next clock pulse.
S3 - Latch PrDZ value. State S3 is reached when either the voltage amplifier output has switched states or the counter 868 has reached the final count. The present count in the counter is latched into the 1~DZ register by activating the vZCLI~ signal for one state time. The state aachine always transfers to state S4.
S4 - Clear Counter. State S4 sets up the state aachine tos auto-zeroing the current amplifier 90. The counter is cleared via a ZERRESb signal and the MUXes 774, 776, 778, 780, 78Z and 784 on the output o! the counter switch to direct the count value to the PrMZ register and the current amplifier 90. The state machine always trans-fers to state S5.
S5 - Five microsecond delay. State 85 and S6 are analogous to state s1 and S2 except the counter is ap-plied to the currant amplifier 90 and the output of the current amplifier CAI~h is examined to transfer from SS to S7.
S6 - Clock Counter. This state is identical to state S2. The state machine always transfers to state S5.
S7 - Latch AMZ value. This state is analogous to state S3. Current contents of the counter 868 are latched in the AMZ register. Additionally the end of auto-zero signal EOAZh is active to indicate that the auto-zero operation has been complete.

There are various input and on signals to the auto-zero machine identified as follows:
ARBSYh - Auto-Range Husy. This signal is active high when the auto-range state machine is not idle. This 15 signal is applied to the state machine by way of an in verter 870. This signal is discussed in connection with the auto-range state machine.
STAZh - Start Auto-Zero. This signal is active high when the initiate auto-zero bit of a command register 20 has been written. This signal is available at the output of the AND gate 528 (FIG. 62).
VAHPb - Voltage Amp Output. This signal is ac tive low when the voltage amplifier 80 has been auto zeroad. The VAIKPh signal is the output signal of the 25 voltage aaplilier 80 and is applied to a flip-flop 889.
CAHPb - Current Amp Output. This signal is low when the current amplifier 90 has been auto-zeroed. The C7l~h signal is available at the output of the current am-plifier 90. This signal is applied to a flip-flop 891.
30 TIliOLITh - Time Out. This signal is active high when a live microsecond time delay has expired. The TIXOUTh signal is available at the output of a NAND gate 870 (FIG. 68). The NAND gate 870 is a two input NAND
gate. One input is active during the test mode. The 35 other input is troa a five microsecond timer comprised of flip-!lops 87Z, 874, 876 and 878 and a NAND gate 880. The Q outputs of these flip-flops are tied to the inputs of the HAND gate 880. The flip-flops 842 and 846 are con-figured such that their ~ outputs are connected to their D
inputs. The ~ outputs of the flip-flops 872, 874 and 876 era applied to clock inputs CK of the flip-flops 874, 876 and 878, respectively. A TIMOUTb signal, available at an output of an inverter 871 (FIG. 69), is also applied to the auto-zero state machine. The TIMOUTb signal is ap-plied to the NAND Bats 860. The SMCLKh signal is applied to an inverter 882, whose output is connected to the clock input CK of the flip-flop 872. The reset inputs 1~ of the flip-flops 872, 874, 876 and 878 are controlled by an OR
Bata 884. The OR gate 884 is a two input OR gate. Sig-nals AZTIMh and ARTIMh era applied to the inputs.
FULLb - Counter Full. This signal is active low when the counter 868, used to set a bias current, is at 111111 count. This is a maximum bias count value. The counter 868 includes the flip-flops 872, 874, 876, 878, 880 and 882 and a HAND gate 884 connected as shown in FIG.
64. The Q outputs of the flip-flops 872, 874, 876, 878, 880 and 882 are connected to the NAND gate 884, which is the FULLb flag and the internal ZERO[5...0] bus. The FVLLb flag is applied to the inputs of the HAND gates 842 and 854. It is also applied to an OR gate 886 by way of an inverter 888. The other input to the OR gate 886 is frog a flip-llop 889. The input to the flip-flop 889 is a signal VAI~h, which is the output of the voltage amplifier 80. This Clip-llop is clocked by the complement of the S~ICLI~ signal, available at an inverter 890. The flip tlop 889 is rust by a RESETb signal available at the out put of an inverter 892. The output of the OR gate 886 is applied to the Hl~ND gate 858.
RESETh - Raaat. This signal is active high to reset the state ragi~ter flip-flops 836, 838 and 840 dur-ing systaa reset. This signal is available at the output of the inwrtar 89Z.
AUTO-ZERO OUTPUTS
ZERRESb - Zaro Counter Rasat. This signal is active low to reset the counter 868 used to generate the i5~ 2~.d~~~8 bias counter and is active in states SO and S4. This sig-nal is available at an output of a NAND gate 894.
ZERCLKh - Zero Counter Clock. This signal is active high to increment the bias current counter 868.
This signal is active in states S2 and S6 and is available at an output o! a NAND gate 896.
TIIrDtEQh - Time Request. This signal is active high to request a five microsecond delay period and is ac-tive in states S1 and S5. This signal is available at an output of a two input OR gate 898. Outputs from AND gates 900 and 902 are applied to inputs of the OR gate 898. In-puts to these AND gates are from the state machine.
AZBSYh - Auto-Zero Busy. This signal is active high to indicate when an auto-zero operation is active.
This signal is used to interlock the auto-zero and auto range state machines. The AZBSYh signal also inhibits the microprocessor 30 from writing to the auto-zero registers by disabling NAND gates which decode the register select signals AMZh and ADZh and the write clock signal wRCLKh.
This signal is active in states S1, S2, S3, S4, S5, S6 and S7. This signal is available at an inverting output of a NAND gate 904. The output o! the NAND gate 904 is also the ZERSEhh signal.
EOAZh - End o! Auto-Zero. This signal is in ac tive high signal which cats the flip-flop 492 (FIG. 62) in the ADCR atatu~ register to indicate an auto-zero process has finished. It also clears a flip-flop which generates the STAZh signal and is active in S7. This signal is available at an output o! an AND gate 906.
GZh - Curzent Auto-Zero. This signal is active high to indicate when the current amplifier 90 is being auto-zeroed. This signal is active in states S4, S5 and S6. This signal is available at an inverting output of a two input NOR gate 908. Inputs to the NOR gate 908 are from AND gates 910 and 912 which is connected to the state machine.
VAZh - Voltage Auto-Zero. This signal is active high to indicate when the voltage amplifier 80 is being 158 '~~~~~~~
autorzeroed and is active in state S1 and S2. This signal is available at a non-inverting output of a HAND gate 914.
CZCLI~1 - Current Zero Register Clock. This sig nal is active low to clock the AMZ register for the cur s rent amplifier bias count. When the state machine is idle, this signal is generated by decoding the register select signal AI~Zh and the write clock signal wRCLKh.
When the state machine is active, S7 is decoded to gener-ate a clock pulse to the register. This signal is active in S7. This signal is available at an output of a two in put AND gate 916. One input to the AND gate 916 is an in verting output from the NOR gate 908. The other input is from a three input NAND gate 918. One input to the HAND
gate 918 is a non-inverting output of the HAND gate 904.
The other inputs are the AMZh and WRCLICh signals.
VZCLRh - Voltage Zero Register Clock. This sig-nal is active low to clock the ADZ register !or the volt-age amplifier bias count. When the state machine is idle, this signal is generated by decoding the register select signal ADZh and the write clock signal WRCLIQs. When the state machine is active, S3 is decoded to generate a clock pulse to the register. This signal is active in S3. This signal is available at an output of two input AND gate 9Z0. An inverting output of the N1~ND gate 914 is applied to one input. The other input is from a three input NAND
gste 9Z2. The l~DZh and wRCLi~ signals are applied to two inputs. 1~ non-inverting output from the NAND gate 904 is applied to the other input.
~rZST bus - Auto-Zero States. This three bit wide bus contains the auto-zero state machine flip-!lops 836, 838 and 840. This bus is used to permit reading of these flip-flops during a test condition.
~L~14-2ER0 STATE I~CHZNE OPERJ~TION
The auto-zero state machine auto-zeros the volt age 80 and current amplifier 90. The auto-zero function is initiated by the software setting a bit in the command register which includes the flip-flops 5Z2, 524 and 526.
Nore specifically, referring to FIG. 6Z when the command bit ADCR[2] is written to the HAND gate 520 this, in turn, controls the flip-flops 522, 524 and 526 to generate the start auto-zero signal STAZh at the output of the AND gate 528.
The voltage 80 and current 90 amplifiers are auto-zeroed in the following manner. After the generation of the start auto-taro signal STAZh. First, the six bit counter 868 (FIG. 67) is cleared. This counter 868 is cleared by the ZERRESb signal in states SO and S4. After the six bit counter 868 is cleared the voltage amplifier 80 is placed in the auto-zero state by shorting its nonin-verting input to ground by way of the MUXas 66 and 86.
This is accomplished by the VAZh signal which is active in atatas Sl and S2. This signal generates an AGNDh signal at the output of an OR gate 924, which shorts the nonin-varting input of the voltage amplifier 80 to ground by way of the IKtJXe: 66 and 86. MUX 88 removes the internal com-pensation from the voltage amplifier 80. Next the output of the counter 868 is gated onto a ZERO[5...0] bus for the amplifier being auto-zeroed. Subsequently, a five aicrosecond delay is timed. This is accomplished by the circuitry in FIG. 68 previously discussed. At the end of the five ~icroaecond delay, a TIHOUTh signal is generated at the output of the HAND gate 870. At the and of the de-lay, the output signal VAI~h, which is the output signal of the voltage amplifier 80, i: examined. Also, the counter 868 full counter count signal FULLb is examined.
I! either o! these signals is active, the count is latched into the ADZ register by the state machine. Otherwise, the counter 868 is incremented and the five microsecond delay is ti~sd again. After the count is latched the sa-quence is repeated with the current amplifier 90.
AVSF AND ACSF AUTO-RANGE RE~T~T~ua The voltage scale register AVSF (FIG. 70) is a read-write register that is used to control operation of the voltage input ranging circuitry 84 (FIGS. 34 and 80) .
The value written into this register determine: the volt age amplifier 80 auto-ranging mode. If a zero is written to this register, the voltage amplifier 80 ~ ~ ~ ~ ~ ed' in an auto-ranging mode. A nonzero value inhibits the auto-ranging and sets the voltage amplifier 80 in a fixed gain mode. The register is not a true read-write register.
The value read from it will not necessarily be the value that was written into it. Writing a zero into the AVSF
register enables auto-ranging. However, a zero cannot be read from this register. The possible values are provided in Table 9. The value read from this register is one of , the scale factors to properly scale the 8 bit A-D output.
Five values are possible: x1, x2, x4, x8 and x16.
The AVSF register includes six flip-flops 944, 946, 948, 950, 952 and 954. The D inputs of these flip-flops are connected to the data bus DATA[5...0] operations during ranging. Any non-taro value written to the AVSF
register, is detected by a HAND gate 998 which disables autoranging. The ~ inputs of the flip-flops 944, 946, 948, 950, 952 and 954 are tied to the output o! a buffer amplifier 955. The input to the butter amplifier 955 is the signal VRCLKb. This signal will be defined in connec-tion with the auto-range state machine and is used to con-trol read and write operations of the AVSF register. The Q outputs of the flip-flops 944, 946, 948, 950 and 952 are tied to the 8 inputs of MUXes 956, 958, 960, 962 and 964, respectively. The D output of the flip-flop 954 is a sig-nal VG~rIN3Zh. This signal VGAIN 32h along with GAIN(4...0~ troy autoranging circuit which includes counter 1170 (lIG. 71) are applied to ranging circuitry 84 a~fd l4tJXes 86 0! the voltage amplifier 80 (FIGS. 34 and 80) to control the voltage gain. The counter 1170 as will be discussed below, includes the ranged value as a result of autoranging o! the voltage asplitier 80 and the current aaplilier 90. Hore specifically, th~ MUXes 956, 958, 960, 962 and 964 allow the Q outputs of the flip-flops 944, 946, 948, 950, 95Z and 954 to be connected to either to a gain bus G1~IN(4...0] or to ~r inputs of MUXes 966, 968, 970, 97Z and 974. The 8 inputs of the I4IJ7Ces 966, 968, 970. 97Z and 974 are connected to ground which allows the 161 ~1~~~~~
output signals of the flip-flops 944, 946, 948, 950, 952 and 954 to either be grounded or connected to the data bus DATA[5...0] by way of tristate devices 976, 978, 980, 982 and 984 for read operations. The tristate devices 976, 978, 980, 982 and 984 are under the control of a signal VRRDb (FIG. 71) which read operations of this AVSF regis-ter.
The output of the MUXes 966, 968, 970, 972 and 974 are also tied to one input of a plurality of AND gates 986, 988, 990, 992 and 994. The other input to the AND
gates 986, 988, 990, 992 and 994 is from an inverter 996.
The input to the inverter 996 is the VAZh signal (FIG.
69), the voltage auto-zero signal, which indicates the voltage amplifier 80 autoranging is active. The output of the AND gates 986, 988, 990, 992 and 994 are connected to the VGAIN[4...0] bus which controls the autoranging MUXes 86 (FIG. SO).
The I~JXei 956, 958, 960, 962 and 964 are under the control of a HAND gate 998 which generates a signal VRZEROh. This signal indicates that the microprocessor 30 wrote a zero to the AVSF register to initiate voltage auto-ranging. The signal VRZEROh is active high and de-termines whether the voltage amplifier 80 is placed in an auto-ranging mode or a fixed gain mode. The input to the HAND gate 998 are the Q outputs of the flip-flops 944, 946, 948, 950 and 95Z. A zero written to the AVSF regis-ter will cause the Q outputs of the flip-flops 944, 946, 948, 950 and 95Z to be high or true. This will, in turn, cause the signal VRZEROh to be active which, in turn, will cause the I~JXes 956, 958, 960, 962 and 964 to connect the Q output signals trom the flip-flops 944, 946, 948, 950, 95Z and 9S4 to the VGAIN[4...0] bus for auto-ranging.
Non-zero valuss written to the register AVSF are detected by the NAND gate 998 will place the circuitry in a fixed gain mode. This will cause the l4tJXes 956, 958, 960, 962 and 964 to connect the Q output signals from the flip-flops 944, 946, 948, 950, 952 and 954 to the I~iLTXes 966, 968, 970, 972 and 974. The IKUXes 966, 968, 970, 972 and 974 either 'ground the Q outputs of the flip-flops 944, 946, 948, 950, 952 and 954 or allow them to be tied to the AND gates 986, 988, 990, 992 and 994 which, in turn, are connected to the VGAIN(4...0] bus. The MUXes 966, 968, 970, 972 and 974 era under the control of an AND gate 1000 which inhibits autoranging of the voltage amplifier 80 when the current mode has been selected and is being autoranged. The AND gate 1000 is a three input AND gate 1000. Tha VRZEROh, indicating voltage amplifier autorang-ing signal is applied to one input. The AZHSYb signal, which represents that the auto-zero signal is busy, is ap-plied to another input. Tha output of an AND gate 1002 is a signal CURRENTh, which represents that the current mode has bean selected. The CUR.RF.NTh signal is applied to the third input of the AND gate 1000. The AZHSYb signal, used to-inhibit the auto-range state machine when the auto-zero machine is active. The AND gate 1002 inhibits auto-rang-ing when the current subsystem is selected.
A REGRESb signal is applied to the reset inputs R o! the flip-tlope 944, 946, 948, 950, 952 and 954. The REGRESb signal is available at the output of an inverter 1004 (FIG. 71). The input to the inverter 1004 is the signal RESETh from the CPCTL[3...OJ bus.
The current scale factor register ACSF is a read-write register used to control the operation of the current input auto-ranging circuitry. The value written to this register determines the current aubsystam auto ranging operating mode. The zero is written the current subeystaa is placed in an auto-ranging mode. A non-zero value inhibits the auto-ranging mode and sets the current mirror into a fixed scale value. The register is not a true read-write register. In other words, the value road troy it is not necessarily the value that was written into it. Although writing a zero into the ACSF register en-ablea the auto-ranging mode, a zero will never be read troy it.
The ACFR register (FIG. 70) includes the llip-tlops 1006, 1008, 1010, 1012, 1014. The data bus DATA[4...OJ is applied to the D inputs of these flip-flops for'write operations in a fixed gain mode. A NAND gate 1048 detects non-zero values written to this register to enable a fixed gain mode. The b inputs of the flip-flops 1006, 1008, 1010, 1012 and 1014 are tied to a buffer am-plifier 1016. The input to the buffer amplifier 1016 is a signal CRCLKb, which will be discussed in connection with the auto-range state machine, latches the gain value in this register at the completion of autoranging. The sig-nal REGERSSb is applied to the reset inputs R. The MUXes 1018, 1020, 1022 and 1024 allow the Q outputs of the flip-flops 1006, 1008, 1010 and 1012 to be connected to the CGAIN[4...0] bus, indicating that the auto-range function has been selected, or to MUXes 1026, 1028, 1030 and 1032.
The CGAIN[3...0] bus is applied to the current mirror 92 (FIG. 81) to control the divider ratio of the current mir-ror 92. The MtJXea 1026, 1028, 1030 and 1032 either allow the output signals Q from the flip-flops 1006, 1008, 1010 and 1012 to either be grounded or applied to the gain bus CGAIN[3...0] or tied to tristats devices 1034, 1036, 1038 and 1040 to allow the them to be read at the data bus DATA[3...0]. More specifically, the Q output of the flip-flop 1006 is applied to a B input of the MUX 1018. An A
input of the 14LJX 1018 is applied to the gain bus bit GAIN [ 3 ] . The Q output of the f lip-f lop 1006 is also ap-plied to inputs of OR gates 1042, 1044 and 1046. The Q
output of the flip-flop 1008 is also applied to other in-putm of the OR gates 1042, 1044 and 1046. Also, the Q
output of the flip-flop 1010 is applied to the OR gates 1044 and 1046. Lastly, the Q output of the flip-flop 1012 is also applied to an input o! the OR gate 1046.
The outputs of the OR gates 1042, 1044 and 1046 are applied to the B inputs o! the MUXas 1020, 1022 and 1024. The fixed gain bus bits GAIN[3...0] are applied to the A inputs of the MUXas 1018, 1020, 1022 and 1024. The I~JXea 1018, 1020, 10ZZ and 1024 are under the control o! a HAND gate 1048. The NAND gate 1048 generates a CRZEROh signal which indicates that the microprocessor 30 has written a zero to the ACSF register to initiate current auto-ranging. A non-zero value written to this register which places the current mirror 92 into a fixed scale value. The Q outputs of the flip-flops 1006, 1008, 1010, 1012 and 1014 are applied as inputs to the NAND gate 1048.
If a zero is written into the ACSF register the MUXes 1018, 1020, 1022 and 1024 connect the Q output of the flip-flop 1006 and the Q outputs of the flip-flops 1008, 1010 and 1012 to the gain bus GAIN[4...0]. Non-zero values written into the ACSF register causes the MUXes 1018, 1020, 1022 and 1024 to be connected to the A inputs o! the MTJXes 1026, 1028, 1030 and 1032. The B inputs of the MUXes 1026, 1028, 1030 and 1032 are grounded. The I~JXes 1026, 1028, 1030 and 1032 are under the control of an AND gate 1049 which disables the auto-ranging of the current amplifier 70 when the system is in a voltage mode.
In this condition, the output signals from the flip-flops 1006, 1008, 1010 and 1012 are grounded. There are two in-puts to the AND gate 1049. One input to the AND gate 1049 is from the HAND gate 1048. The output of a NAND gate 1048 indicates that auto-ranging has not been selected.
The ~ outputs of the flip-flops 1006, 1008, 1010, 1012 and 1014 are applied as inputs to the NAND gate 1048. The other input to the AND gate 1049 is an inverter 1050. The output o! the inverter is a VOLTAGEh signal which indi-cates that the MtJX 66 is in a voltage mode. Tha input to the inverter lOSO is the output of the AND gate 1002 which indicates that the 1KUX 66 is in a current mode. The in-puts to the AND gate 1002 is a I~OCOSEhh signal (FIG. 66) which indicates that a current mode has been selected.
The outputs o! the MtJXas 1026, 1028, 1030 and 1032 are connected to a pair of serially coupled inverter ampli-fiers 1052, 1054, 1056, 1058, 1060, 1062, 1064 and 1066.
The outputs of the inverters 1054, 1058, 1062 and 1066 are applied to the gain bw CGAIN(3...0] as wall a~ to the triatate devices 1034, 1036, 1038, 1040 either directly or by way o! AND Batas, 1068, 1070 and 1072. More spe-cifically, the output of the inverter amplifier 1054 is applied to the tristate device 1034. The output of the inverter amplifier 1058 is applied to an AND gate 1068 along with the output of the inverter amplifier 1052. The output of the inverter amplifier 1062 is applied to the AND gate 1070 along with the output of the inverter ampli-fier 1056. The output of the inverter amplifier 1066 is applied to the input of the AND gate 1072 along with the inverter amplifier 1060.
These tristate devices 1034, 1036, 1038 and 1040 10_ era also applied to the data bus DATA[3...0] to allow the ACSF register to be read. The output of the inverter am plifier 1064 is applied to a tristate device 1068. This tristate device 1068 is applied to the DATA[4] bit. The tristate devices 1034, 1036, 1038, 1040 and 1068 are con trolled by a signal CRRDb. This signal will be identified in connection with the auto-range state machine.
A signal AZST[2...0], representative of the states of the auto-zero state machine, is applied to AND
gates 1070, 1072 and 1074 (FIG. 67). This signal repre-sents the Q output signals of the auto-zero state register flip-flops 836, 838 and 840 (FIG. 69). Also applied to the AND gates 1070, 1072 and 1074 are test signals. The TEST signals are also applied to the reset input R of the flip-flop 954. The outputs o! the AND gates 1070, 1072 and 1074 are applied to tristate devices 1076, 1078 and 1080. The outputs of the tristate devices 1076, 1078 and 1080 are applied to the data bus DATA[7...5] to allow thue signals to be road by the aicroprocassor 30. The tristate devices 1076, 1078 and 1080 are under the control 0! a signal CRRDb.
The states of the auto-range state machine ARST[Z...O] are applied to the inputs of AND gates 1082, 1084 and 1086. Test signals are also applied to the in-puts of the AND gates 108Z, 1084 and 1086. The ARST[Z...O] signal represents the states of the auto-rang-ing stag register flip-flops and will be discussed in connection with the auto-ranging state machine. The out-puts o! the AND gags 1082, 1084 and 1086 are applied to tristata devices 1088, 1090 and 1092. The outputs of these tristate devices are applied to the data bus DATA(7...5]. The tristate devices 1088, 1090 and 1092 are under the control of a VRRDb signal. This signal controls the reading of the auto-range state register flip-flops states by the microprocessor 30 and will be discussed in connection with the auto-range state machine.
AUTO-RANGE STATE MACFIIN -The auto-range state machine is illustrated in FIG. 71. A state transition table, state diagram and transition state equations are provided in Appendix D.
This state machine auto-ranges the voltage 80 and current 90 amplifier gains before initiating an analog to digital conversion. During voltage autoranging, the output signals o! the voltage amplifier 80 is compared with a predetermined value by the comparator 74 (FIG. 34) to determine if the amplifier output is either too large or out of range of the A/D. At the start of the autorang-ing a gain shift register 1170 (FIG. 71) is initialized and incremented during a predetermined time period (TIHOUTh). When either the comparator 74 switches state or the tiae period expires, the value of the gain shift register will represent the gain. This gain value is stored in the AvSF register and is used to control the ranging circuitry 84.
During current autoranging, ranged currants are directed out of the IrDCO pin and applied to an external register. The voltage across the external resister is then applied to a voltage input. Ranging is then accom-plashed in a siailar manner as the voltage autoranging.
Gain values in this mode are stored in the ACSF register.
The auto-range state machine includes three state register flip-flops 1128, 1130 and 1132; NAND gates 1134, 1136, 1138, 1140, 1142, 1144, 1146; AND gates 1148, 1150, 1152, 1154, 1156, 1158, 1160 and 1162; OR gate 1164 and various output gates connected as shown in FIG. 71.
The Q outputs of the state f lip-f lops 1128, 1130 and 113 2 indicate the state variables ROh, Rih and R2h. The ~ out-2~~~~53 put of the state register flip-flops 1128, 1130 and 1132 generate the state variables ROb, Rlb and R2b. The state variable ROd is generated at the output of the HAND gate 1138. The state variable Rid is generated at the output of the HAND gate 1146. The state variable R2d is gener-ated at the output of the OR gate 1164.
Each of the state register flip-flops 1128, 1130 and 1132 is clocked by the SMCLKh signal. The reset sig-nal REGRESb, available at the output of the inverter 1004, is applied to the reset inputs R of these state register f lip-f lops.
The outputs of the state register flip-flops 1128, 1130 and 1132 define permissible output states for the auto-rangs state machine in accordance with Table D-1, Appendix D. The state register flip-flops 1128, 1130 and _1132 allow for eight states. However, only seven are necessary and are defined as follows:
SO - Idle. The state machine is in the idle state wafting for a start auto-range signal (STADCh) to be active. The state machine also stays in the idle state as long as the auto-zero state machine is busy. This inter-lock between the twv independent state machines is imple-aented by the auto-range busy signal (aRBSYh). When the state start auto-range signal is active and the auto-range busy signal is inactive the auto-zero state machine moves to state Sl.
S1 - Reset shift register. State S1 initializes a shift register 1170 for the type of conversion. For a voltage conversion the shift register 1170 is initialized to the binary value 0001 where the least significant set bit is set. This corresponds to a voltage gain o! 1. For a current conversion the shift register 1170 is set to the binary value 00000. This corresponds to a current gain of 1. Only the four least significant bits are used for set-tang the current gain. The least significant bit is set or cleared by the GRESh signal by additional decoding of the VOLTaGEh and CURRENTh signals.

1s8 S2 - Five microsecond delay. State S2 and S3 form a loop that is repeated until either the comparator 74 (FIG. 34) output switches or the shift register 1170 reaches the final gain. During S2, the time request sig-nal (TZMREQh) is active in the time-out (TIMOUTh) signal is monitored. when the time request signal goes active, the state machine enters the state S1, this triggers the five microsecond time delay. When the time delay has timed out, the time-out signal becomes active. This causes the state machine to transfer either to state S3 or S4. State S3 is entered if the RANGEh signal is inactive indicating that either the gain setting is not high enough yet or the shift register 1170 has not yet reached the final gain setting. The final gain setting is detected by decoding the VOLTh and CURRh signals with the fourth and fifth bits of the shift register 1170. For a current channel, indicated by an active CURRh signal, a fourth bit indicates that a maximum gain has been reached. Maximum gain for a voltage channel is decoded by a fifth bit of the shift register 1170 and an active VOLTh signal.
The state machine transfers to state S4 when the R~NGEh signal is active, if the auto-range function is ac-tive. It the auto-range function is disabled indicated by an inactive 11TORNGh signal, the state machine transfers to state s4 after the live microsecond time delay. This five microsecond time delay permits the amplifier 80, 9o to settle and reach a stable output value.
S3 - Clock shift register. State S3 is entered when the shift register 1170 does not yet contain a maxi mua gain and the gain is not enough to generate a suffi ciently high input signal. The clock signal is active in state S3 causing the shift register 1170 to shift one bit.
For voltage channel signals a zero is shitted into the least significant bit o! the shift register 1170. This causes the shift register to shift a one across the regis-ter generating the following sequence o! values: 00001, 00010, 00100, 01000, 10000.

For current channels a 1 is shifted into the least -significant bit which results in the following se-quence of values: 00000, 00001, 00011, 00111, 01111.
Only the four bits of the gain are used in setting the current amplifier 90. The state machine always transfers to the state S2 on the next clock pulse.
S4 - General SOC pulse. State S4 is used to generate a start of conversion pulse to the analog digital converter. Tha state machine stays in state S4 until the SOC3b signal becomes active at which time the state machine moves to state S5. The SOC3b signal is active when the start of conversion pulse has bean active for three state machine clock periods.
S5 - wait for conversion. The atata machine is waiting for the analog end of conversion signal while in state S5. When the analog end of conversion goes high in dicating that the conversion has finished, the state machine transfers to state S6.
S6 - EOC pulse. End of Conversion EOAZh is ac tiva in state S6. This indicates to the commands/status register section that the conversion process has bean com pletad.
AUTO-RANGE INPUTS
Tha inputs to the auto-range state machine are as follows:
AZHSYh - Auto-Zaro Busy. This signal is active high wham the auto-taro atata machine is not idle. This signal is applied to the state machine by way of an in-vartar 1171.
ATORNGB - Auto-range Active. The auto-range ac-tive signal'ATORNG represents that the auto-ranging has bean activated when the sicroprocusor 30 writes a zero into a shift register 1170. More apacifically, the ATORNGh signal is available at the output of an inverter 1164 (FIG. 70) while the ATORNGb signal is available at the output of a dual input NOR gate 1166 (FIG. 70). The NOR gate 1166 is controlled by two AND gataa 1168 and 1002. The VOLTAGEh signal is applied to one input of the AND g~ 1116. The signal VRZEROh is applied to the other input. The output of the AND gate 1002 in applied to the other input of the NOR gate 1166. The output of the AND
gate 1002 indicates that the current mode has been selected and that the microprocessor has a zero to the ASCF register to initiate auto-ranging.
The ATORNGb and ATORNGh signals are applied to the state machine by way o! circuitry which includes an AND gate 1172, an inverter 1173, an OR gate 1174 and an . AND gate 1176. The OR gate 1174 is a two-input OR gate.
One input is from the AND gate 1172. The AND gate 1172 is a three-input AND gate. The ATORNGb, TIMOUTh and RANGEh signals are applied to the AND gate 1172. The other input to the OR gate 1174 is from a two input AND gate 1176.
One input to the AND gate 1176 is the ATORNGb signal. The other input is the TIMOUTh signal.
RANGEh - In Range Signal. This signal is active high to indicate that either the comparator 74 output sig-ns! COMPb has gone low or that the gain shift register 1170 has reached the maximum gain setting for the operat-ing mode selected. The auto-ranging sequencing circuitry 76 includes the gain register 1170, an OR gate 1178, AND
gates 1180 and 1182 and a Clip-flop 1184. The RANGEh sig-nal is available at the output of the HOR gate 1178.
The gain shift register 1170 is comprised of the flip-!lops 1188, 1190, 1192, 1194 and 1196. The Q outputs of these flip-!lops are tied to the D input of the next flip-tlop in succession. The Q outputs are also tied to the gain bus GAiN[4...0]. A GCLKh signal is applied to the clock C1C inputs of each of these Clip-flops. The GCI~ signal is available at the output of an AND gate 1198. The inputs to the AIdD gate 1198 are the state register signals ROb, Rih and R2b which indicate the auto range state machine is in S3. Also applied to the AND
gate 1194 is the signal SMCLRb.
The GCLIQz signal is used to shift a zero into the Clip-!lop 1188 when a voltage channel has been selected and to shilt a 1 into the Clip-Clop 1188 when a 171 ~~~~au8 curre~~!!tt channel has been selected. More specifically, 0R
gates~200, 1202, AND ate 1204 g and a HAND gate 1206 con-trol this function. A CURRENTh signal which is active high is applied to the input of the OR gate 1200. The other input to the OR gate 1200 is a signal GRESb avail able at the output of the AND gate 1204. The output of the OR gate 1200 is applied to the preset input of the flip-flop 1188 to shift a 1 into this flip-flop. This value is shifted across the shift register by the GCLKh to signal.
Similarly, a VOLTAGEh signal is applied to one input of the OR gate 1202. The other input to the OR gate 1202 is a GRESb signal. The output of the OR gate 1202 is applied to the reset input R of the flip-flop 1188 to shift a zero into this flip-flop when the voltage channel is selected.
The comparator 74 output signal COMPb is moni-tored by the flip-flop 1184. The SMCLRb signal is applied to the clock CR input of this flip-flop. The REGRESb 81g-nal is applied to the reset input R. The output of this flip-flop indicates that the comparator 74 output signal has not yet switched which means that the comparator 74 output signal is less than, !or example, 1.25 Vdc to indi-cate that one-halt the maximum has not bean reached. This signal is applied to one input o! the OR gate 1178. The other inputs to the OR gate 1206 are from the outputs of the AND gates 1180 and 1182. These AND gates indicate that the shift register 1170 has reached the maximum gain setting for the operation modes specified. More spe-3o citically, the AND gate 1180 relates to the current mode.
The CURRENTh signal is applied to one input of the AND
gate 1188. The other input is lroa the output of the shift register f lip-f lop 1194 which when active indicates the highest gain setting when the current mode is selected.
similarly, the vOLTAGEh signal is applied to an input of the AND gate 1182. The other input to the AND
gate 1182 is the output of the shift register flip-flop 1196,~ich represents the highest gain setting when the voltage mode is selected.
The outputs of the AND gates 1180 and 1182 are applied to the inputs of the OR gate 1178 along with the Q
output o! the flip-flop 1184 to generate the RANGEh sig nal. The RANGEh signal indicates that either the compara-tor 74 output signal COI~b has gone low or that the gain shift register 1170 has reached the maximum gain setting for the operating mode a~lactad.
TIHOUTh - Time Out. This signal is active high when the live microsecond timo delay has expired. This signal is available at the output of the HAND gate 870 (FIG. 68).
SOC3b - Start o! Conversion Three. This signal is active low when the start o! conversion pulse has been active for thra~ clock periods.
ANAEOCh - Analog End Of Conversion. This signal is active high when the A/D 78 has finished the conver sion. This signal is applied to the AND gate 1140 and to the AND gate 1162 by way of an inverter 1208.
RESETh - React. This signal is active high to r~s~t the stag r~gistsr flip-flops during system reset.
STADCh - Start Conversion. This signal is ac-tive high when the AI~T7C register is written. This signal is discussed blow.
PTO-RANGE OUTPUTS
The output signals of the auto-range state aachin~ ar~~ as follow:
GRESh - Gain Shift R~gist~r Rust. This signal i~ active high to rust the shift register 1170 used to generate the gain. This signal is active in stag S1.
GCL~ - Gain Shift Register Clock. This signal is active high to shift the shift r~giat~r and is active in state S3.
TII4R2:Qh - Tima Raqusst. This signal is active high to raqtiut a five microsecond delay period. This signal is active in state S2. The TII~QtEQh signal is g~rr~rat~d at the output o! an AND gate 1210. The signals 2~~~~~~

ROb,.~ and R2h are applied to the input o! the AND gate 1210 to generate the TIMREQh signal when the auto-range state machine is in state S2.
ARBSYh - Auto-Range Busy. This signal is active high to indicate when a conversion operation is active.
This signal is also used to interlock the auto-zero and auto-range state machines. The ARHSYh also inhibits the nicroprocesaor 30 from writing to the gain registers by disabling the NAND gates which decode the register select signals (AVSFh, ACSFh) and the write clock WRCLKh signal.
This signal is active in states S1, S2, S3, S4, S5 and S6 and is available at an inverting output o! a HAND gate 1212. The state variables ROb, R1H, R2b are applied to the inputs to the HAND gets 1212 to generate the ARBSYh signal in states Sl, S2, S3, S4, S5 and S6.
The signal ARBSYb is used to control write and read operations to the ASCF and AVSF registers. More specilically, ARHSYb signals inhibits the microprocessor 30 lrom reading or writing to the ASCF or AVSF registers when the auto-range state machine is active. The signal AR8SYb is generated at a non-inverting output o! the HAND
gate 1212. A non-inverting output o! this HAND gate is applied to inputs o! NAND gates 1214 and 1216. The AVSFh signal, which is generated when the microprocessor 30 places the address 50026 on the address buy ADDR[5...0], is applied to the HAND gate 1214 and an AND gate 1218. An ~rCSFh signal is applied to the inputs o! the HAND gate 1216 and a NlrND gate 1220. This ACSFh signal is generated any time the microprocessor 30 places the address 50027 on the address buy ADDR[S...OJ. The RDCLRh signal is applied to the inputs o! the NAND gates 1218 and 1220. The WRCLKh signal i~ applied to the inputs o! the HAND gates 1214 and 1216. The outputs o! the NAND gates 1214 and 1218 are the VRRDb and VRCLRb signals, which are used to control read and write operations to the AVSF register. The outputs o!
the NhND gates 1216 and 1220 are the CRRDb and CRCLKb sig-nal~ which are used to control read and write operations to the ACFS register.

2~~~~~8 z EOCh - End of Conversion. This signal is active high which sets a flip-flop in the status register to in-dicate that the conversion process has finished. It also clears the flip-flop 1246 (FIG. 62) which generates the STPIDCh signal and is active in state S6. This signal is available at an output of an AND gate 1222. The signals ROh, Rlb and R2b are applied to the inputs of this AND
gate to generate the EOCh signal only when the state machine is in state S6.
AN1~SOCh - Analog Start of Conversion. This sig-nal is active high to initiate an analog to digital con-version and is active for three clock cycle:. This signal is active in state S4. The analog of start vt conversion signal ANSOCh is generated by circuitry which includes Clip-flops 1224, 1226 and 1228, butter amplifiers 1230, 1232 and 1234 and an AND gate 1236. This signal is active high for three clock cycles and is active in state S4. A
signal representing that the auto-range state machine is in state S4 is available at the 7~ND gate 1156 and applied 2 0 to the D input of the f l ip-f lop 1224 . The state mach ins clock signal SHCLXh is applied to the clock input of the flip-flop 1224. The Q output of the flip-flop 1170 is ap-plied to the D input of the Clip-Clop 1226. Tha Q output of the flip-flop 1226 is applied to the input of a buffer aaplitier 1230. The output of the butter amplifier 1230 is applied to the D input of the flip-flop 1228. The Q
output of the flip-flop 1228 i~ applied to the butter am-plitier 1s3? and represents the analog start o! conversion J1N11SOCh. The l~t~tD gate 1236 controls resetting the flip-tlop~ 1224 and 1226 after the flip-flop 1228 is reset.
Hors specifically, the REGRESb signal is applied to one input of the J~ND gate 1226 as well as to the reset input of the f l ip-f lop 12 2 8 . The ~ output o! f lip-f lop 12 2 8 f s applied to the other input of the 1,ND gate 1236. The out-put of the 1~ND gate 1180 is applied to the R reset inputs of the tlip-flops 1224 and 1226.
The clock inputs of the flip-flops 1226 and 1228 are controlled by the microprocessor 30 interrupt signal 1~5 2~.Q~~~~
INTEh ~ More specifically, the interrupt signal INTEh, which is active high is applied to the input of an in-verter 1234. The output of inverter 1234 is applied to the clock CR inputs of the flip-flops 1226 and 1228.
AUTO-RANGE STATE ~c~uTNR n~r~~~rTny The auto-range function is initiated by the software writing to the AMUX register. More specifically, a STADCH signal (FIG. 62) is active high whenever the register AMUX is written to by the microprocessor 30.
_ This signal STADCH is available at the output of an AND
gate 1238. One input to the AND gate 1238 is a test cir-cuit. The other input to the AND gate 1238 is from cir-cuitry which includes buffer amplifiers 1240, 1242, 1244;
flip-flops 1246, 1248, 1250; a HAND gate 1252 and AND
gates 1254 and 1256. The WRCLIQi signal is applied to one input of the AND gate 1254. This signal indicates that the microprocessor 30 is writing to one of tha registers.
An AMUXh signal is applied to the other input of the AND
gate 1254. The AMUXh signal represents that the micropro-cessor 30 has written to the AMUX register by writing the address $0021 on the ADDR[5...0] bus. The output of the AND gate 1254 is a begin conversion signal HEGCONh which indicates that the AMiT7c register has been written to.
This BEGCONh signal is applied to the input of the NAND
gate 1252. The other input to the HAND gate 1252 is from the TEST[4...0] bus. The output o! the NAND gate 1254 is applied to the D input of the flip-Clop 1250 by way of the NaND gate 1252. The other input to the NAND gate 1252 is tro~a teat circuitry. The output of the flip-Clop 1250 is applied to the input of a flip-Clop 1248. The R input of the Clip-flop 1250 is connected to the output of the AND
gate 1256. The AND gate 1256 is a two input AND gate.
The RESETb signal avaf fable at the output o! the inverter aaplitier 516 is applied to one input. An AReSYh signal is applied to the other input by way of the inverter am-plifier 1244. The aN0 gate 1256 resets the Clip-flop 1250 when the auto-range state machine is busy. A Q output of the Clip-tlop 1250 i~ applied to a D input of the flip-tlop.~48. The output of the flip-flop 1248 is applied to the input of the buffer amplifier 1242. The output of the buffer amplifier 1242 is applied to the input of the flip-tlop 1246. The clock inputs of the flip-flops 1246 and 1248 are tied to the output of the inverter 527. The SMCLKb is applied to the input of the inverter 527. The reset inputs R of both of the flip-flops 1246 and 1248 are controlled by the RESETb signal. The output o! the flip-tlop 1246 is applied to the input of a buffer amplifier 1240. The output of the buffer amplifier 1240 is applied to the other input of the AND gate 1236 to generate the STPrDCh signal to indicate that the microprocessor 30 has written to the 1~MU7c register.
When the AMUX register has been written to, the control circuitry requests that the auto-range operation be performed followed by an analog digital conversion.
More specifically, the auto-range state machine performs the following functions. First, the shift register 1170 is initialized. The output of the shift register 1170 is tied to a G1~IN[4...0] bus which sets the gain for the am-plifier. The initial state of the shift register 1170 de-pends on whether the voltage or current channel has been selected for conversion. It a voltage channel has been selected, the initial value is a binary 00001. If a cur-rent channel has bean selected the initial value of the shift register 1170 will be a binary 0000. Next a five microsecond setting delay is timed. The TIMOUTh signal will be active high at the output o! the five microsecond delay. When the TIMOtITh signal is active, the output of a co~parator 74 is checked. It the comparator 74 has switched or it the maximum gain setting has been reached, a start of conversion signal is generated. otherwise the gain is increased and the delay is timed again.
Altar the conversion has been initiated, the auto-range state machine waits for an end of conversion signal ANaEOCh to inform the microprocessor 30 to generate a processor interrupt.

A/D CONTROL LOGIC
The A/D converter 78 is an eight bit successive approximation A/D converter. The ranging circuitry for the voltage 80 and current amplifier 90 provides an addi-tional four bits of dynamic range. The A/D converter 78 is described in detail in Section 7 of SZNGLE-CHIP HZCROCOntprrr~ by Motorola, copyright 1987, ANALOG CONTRO ~ T.nrTC~
The function block diagram for the analog con-trol logic is shown in FIG. 34. This figure, in connection with the block diagram shown in FIG. 72 will be used to explain the analog control logic.
These figures illustrate the control logic for the current and voltage channel IKtJXes 66 and 68, zeroing o! the voltage amplifier 80 and the current amplifier 90 and ranging of the voltage amplifier 80 and the current mirror 92. Also illustrated is an analog power supply subsystea 48 which consists of a band gap regulator sub system 1400, a shunt regulator 1402 and the power monitor portion of subsystem 47. Biasing circuitry 1404 is illus-trated for the quadcomparator subsystem 58 (FIG. 74), the band gap regulator 1400, the 8+ comparator subsystem 50, the power aonitor subsystem 47, the voltage amplifier 80 and the current asplitiar 90. Lastly, temperature moni-toring circuitry 1406 is illustrated which allows the microprocessor 30 to read the ambient temperature.
The I~tJXes 66 and 68 are illustrated in FIG. 73.
The input channels I~JXO, M1JX1, I~JXZ and lttTX3 can be used for either voltage inputs or current inputs. The input channels HO'X4 , lStJXS , IdtTX6 and 1KLTX7 can only be used f or voltage channels. The channel l~T7c8 is for temperature sensing while the Hp7I 6610 is tied to analog ground. Hore specifically, the input char~rnels are configured by l~tc,T~ces 66a-66g. The I~ttJXes 68a-68d allow the input channels HUX1, ~JXZ . l~JX3 and I~JX4 to be tied to the current channe 1 178 ~1~~~~~
IMUX.~ The MUXes 68e-68h allow the input channels to be tied to digital ground VSS.
Sample and hold MUXes 108, 110, 112 and 114 are connected between the channels MUXO and MUX1, MUX2 and IKUX3, MUX4 and MUXS and I~J7c6 and MUX7, respectively.
~ALOG POWER SUPptv The analog supply pins AVDD and AVSS are used to provide power to the analog portion of the IC 10. The analog supply pin AVDD is designed to be connected to a current source. The IC l0 contains an internal shunt regulator (FIG. 76) to regulate the voltage on the AVDD
pin to approximately 5.0 Vdc. More specifically, the ana-log power supply consists of a 2.5 Vdc reference and a shunt regulator subsystea 1402. The 2.50 Vdc reference contains a +1.25 Vdc band gap regulator reference circuit 1406 (FIG. 75) and a buffer amplifier 1412 to generate a +Z.50 Vdc reference: VREF. An adjustment pin VADJ is provided to allow the voltage to be trimmed to exactly +2.5 Vdc t0.5 Vdc. In order to trim the reference, a two resistor voltage divider 1410 which includes resistors 1414 and 1416 is connected between the VREF and AVSS pins with the aid-point connected to VADJ. The buffer ampli-fier 1412 has a source follower output such that it can only source current. This will permit multiple devices to be paralleled. Also, the regulator of the IC 10 can be slaved to smother by connecting its VADJ pin to the VREF
pin.
The band gap regulator subsystem 1406 is illua-trsted in PIG. 75. A band gap reference is a precision voltage reference. In general, the band gap reference circuit utilizes as a reference the base-to-uittsr volt-age o! a parasitic transistor which has a negative tem-perature coefficient (-TC) connected in series with a re-sistor which has a voltage developed across it with a pos-itive tuperature coetficiant (+TC). The voltage de-wlopad across the resistor is a function of a predeter-~ined current supplied to the resistor frog circuitry in-ternal to the band gap regulated reference. The differ-1~9 2~fl~~~8 once i~ temperature coefficients between the base-to-emit-ter vdltage of the parasitic transistor and the voltage across the series connected resistor provides a voltage reference signal having essentially a zero temperature coefficient. As the base-to-emitter voltages of the para-sitic transistors decreases with an increase in tempera-ture, the voltage across the current-fed series connected resistor increases generally proportionally to provide an output reference voltage which remains relatively con-stant. This reference voltage is then applied to a non inverting input of an amplifier; an inverting input of the amplifier is connected to an externally divided portion of the output of the amplifier. The output of the amplifier is a voltage proportional to the reference voltage, rela tively independent of temperature.
More specifically, the output of the band gap regulator reference circuit 1406 is nominally 1.25 vdc.
This output voltage is doubled by a buffer comparator 1412 and the external resistors 1414 and 1416 to produce a +2.5 Vdc reference at the sxternal pin vREF. The external re-sistors 1414 and 1416 are connected in series between the output o! the butter comparator 1412 and an analog ground pin AVSS. The aidpoint o! these resistors 1414 and 1416 is connected to the inverting input o! the buffer compara-for 1412 to allow the reference voltage VREF to be ad-justed. The band gap regulator circuit includes diode connected parasitic transistors 1426 and 1428, a transis-tor 1418, ra~istor~ 1420, 1422 and 1424 and a comparator 1441. Start-up circuitry 1432 is provided for the condi-tion wham power i~ tir~t applied to the IC 10. The start-up circuitry 143 includes the transistors 1434, 1436 and 1438. During this condition, voltages begin to rise from a taro level to a level that will ultimately be regulated by the band gap reference circuit 1406. Initially there is no currant fn any of the devices. In this condition, the transistor 1438 i~ biased on by a PHIAS circuit 1440 which will be discussed below. This, in turn, turns on the transistor 1434 to generate a current that flows into 180 ~;~~g~~8 a diod. connected parasitic transistor 1428. This pro-duces a~voltage at the transistor 1428 which is applied to a non-inverting input of the comparator 1441. That gener-ates a positive signal at the output of the comparator 1440 which, in turn, turns on the transistor 1418 and pro-duces currents in the transistors 1426 and 1428. Conse-quently, the band gap reference approaches a stable regu-lating point based upon the voltages across the diode con-nected transistors 1426 and 1428. When these voltages _ reach a steady state value, the transistor 1436 is turned on which turns off the transistor 1434, in which case all of the current is generated by the transistor 1418.
During regulation, the currant applied to the eaitters of the transistors 1426 and 1428 is essentially' equal. This is because the resistors 1420 and 1424 are the same value and are relatively large compared to the other voltage drops. The base emitter voltage across transistors 1426 and 1428 is dependent upon the current density through these transistors. The current density is the total current divided by the area of the transistors .
The current densities of the transistors 1426 and 1428 are different by a factor of 11 to 1, thus, their base emitter voltages will be ditterent. The ditlerence in base emit-ter voltages appears across the resistor 1422. Since the teaperature coefficient is a function of the voltage across the device, as the base emitter voltage across the transistors 1426 and 1428 goes down, their negative tem-perature coefficient increases. Due to the relative cur-rent density in the transistor 1426 relative to the cur-rant density in the transistor 1428, the series combina-tion of the resistor 14Z~ acrd transistor 1426 will have a positive temperature coefficient (+TC) and is applied to an invsrting terminal of comparator 1441. The base emit-ter voltage of the transistor 1428, which has a negative temperature coefficient (-TC), is applied to a non-invart-ing terminal of the comparator 1441. a the temperature changes voltage across the base emitter junctions of the transistors 1426 and 1428, the voltage across the resistor ~~fl~~5~

1422,11 change by a proportional amount, thus yielding an ou put signal from the comparator 1441 that is rela-tively temperature independent.
SHUNT REGULATOR
The shunt regulator 1400 (FIG. 76) provides a nominal +5.0 Vdc at the AVDD pin based on the reference voltage at VREF. The shunt regulator 1400 includes an am-plitier 1443 and resistors 1444 and 1446. More specifi-cally, the VREF from the butter comparator 1412 is applied l0 to a non-inverting input of an amplifier 1443. The AvDD
bus is the regulated 5.0 Vdc supply. The inverting termi-nal of the amplifier 1443 is connected to the AVDD bus by way of the resistor 1444. The inverting terminal of the amplifier 1443 is also connected to the AVSS bus by way of 15 the resistor 1446. The resistors 1444 and 1446 have equal value which causes the output of the amplifier 1443 to be twice VREF. Since VREF is nominally 2.5 volts, the regu-lated supply bus AVDD will nominally be 5.0 volts. A
shunt element transistor 1447 is connected between AVDD
20 and AVSS. The gate of the shunt element is controlled by output o! the amplifier 1443. When the regulated supply AVDD becomes a little too high, the negative terminal of the amplifier 1443 will be a little higher than vREF.
This will drive the output of the amplifier 1443 negative.
25 This, in turn, will cause a shunt transistor 1447 to turn on a little bit more. This draws current away from the supply bus 71VDD and brings the voltage down until the two inputs to the aaplitier 1443 are essentially identical.
The circuitry which includes the transistors 30 1448, 1450 and 1452 is part of the start-up circuitry.
Transistors 1448, 1450 and 1452 turn oft the shunt tran sistor 1447 during start-up to avoid sinking a lot o! cur rent away from AVDC.
An important aspect of the invention relates to 35 the tact that the IC 10 is current driven. This provides iamunity to voltage spikes typical in applications in the automotive industry. More specifically, the IC to is driven by an input current, developed by an external re 182 ~~~~~~ S
t sisto~1453 and an external voltage identified as VEXT ap-plied to the AVDD bus.
POWER MONITOR SUBSYSTEM
The circuitry which consists of the transistors 1454, 1456, 1458 and 1460 and the comparator 1462 (FIG.
76) performs the power on reset and loss o! +5.0 Vdc func tion. Power on reset is a delay of 8128 oscillator cycles plus an additional ims from the time the reset is removed by clearing the external control pin RESN.
The series connected transistors 1454, 1456, 1458 and 1460 form a voltage divider circuit. The drain of the transistor 1454 is applied to a non-inverting input of the comparator 1462. The output of the amplifier 1443 is applied to an inverting input of the comparator 1462.
The output of the comparator 1462 is a signal SHUNT which is an output pin which may be applied to the RESN pin of - the microprocessor 30 for the power monitor function to reset the microprocsssor 30 upon detection of an under-voltage.
The comparator 1462 monitors the conductive state or gate voltage of the shunt transistor 1447. when-ever the shunt transistor 1447 is determined to be oft, as indicated by the aaplitier 1442 output being at a more positive voltage than the divided voltage at the drain of the transistor 1454, the comparator 1462 output signal shunt will be driven negative, indicating insufficient current available to maintain the AVDD bus regulated at s.o volts.
B+ COI~ARATOR S~tHCVCT~r_ !gin The B+ comparator subsystea (FIG. 76A) is used for power supply generation and includes the following circuitry, resistors 1462, 1464, a comparator 1466 and a transistor 1468. VRE! is applied to an inverting input of the co~parator 1466 providing a +Z.5 Vdc reference. The output of the comparator 1466 is an external pin BDRIVE.
Inputs to ths' co~parator 1466 are applied to a non-invert-ing terainal of the comparator 1466 by way of an externs l pin BSENSE. The resistor 1464 and the transistor 1468 are 2~~J>J8 axamp~ y of the hysteresis mask option, available for all comparators. The resistor 1464 and the transistor 1468 are connected in series to provide feedback from the output of the comparator 1466 to the inverting terminal.

FIGS. 76H and 76C illustrate exemplary circuitry for power supply generation and power supply regulation for the IC 10, respectively. FIG. 76B also illustrates the conditioning circuitry 19.

Referring first to FIG. 76H, the IC 10 is used to monitor the condition of the circuit breaker 12 (FIG.

3Z) by way of the current transformers (CT) 14, 16 and 18.

These CT's may be of the donut type which consist of a secondary winding disposed about the 1~, 8 and C phase conductors of the circuit breaker 12. During certain loading conditions, the output from the CTs may be of the order o! 100 milliamps (m~). In order to reduce this out-put current to a level suitable for the IC 10, for example, 20 microamps, the signal conditioning circuitry 19 is provided. Various conditioning circuitry may be utilized. It should be understood that the conditioning circuitry illustrated in FIG. 76B is merely exemplary.

The CT's 14, 16 and 18 may be connected to the diode bridge 1467 in various ways. For example, the CTs 14, 16 and 18 aay be connected in series with the output connected to the terminals 1469 and 1471. also, only a single CT, !or example, the B phase CT 16, may be tied to the bridge 1467 or the CTs may be paralleled.

The conditioning circuitry 19 includes a full wave diode bridge 1467 defining a pair o! alternating cur-rent terainals 1469 and 1471 and a pair o! rectified ter-ainals 1473 and 1475; the positive ter'inal identified as 1473 and the negative terminal identified as 1475. The conditioning circuitry i9 also includes a resistor 1477 and a resistor 1479. Exemplary values for the resistors 3S 1477 and 1479 are 10 ohms and 50 kilohas, respectively.

The resistor 1477 is connected between the nega-tive terainal 1475 on the bridge 1467 and ground. one side o! the resistor 1479 is also tied to the negative 184 ~~ ~t3 termi~ 1475. The other side of the resistor 1479 is than connected to one of the MUX inputs MUXO, MUX1, MUx2 or MLTX3.
In operation, the current from the current transformers 14, 16 and 18 will flow through the resistor 1477 from ground to the negative terminal 1475 of the bridge 1469 to produce a negative voltage across the re-sistor 1477. If the value of the resistor 1477 is, for example, 10 ohms, a -1.0 volt will be produced across the resistor 1477 for a CT current of about 10 mA. This will, in turn, cause a -1.0 volt drop across the resistor 1479.
If the resistor 1479 has a value of, for example, 50 kilo-ohms, this will, in turn, produce a current of 20 microamps to be applied to one of the current inputs s2 (e. g., MUXO, MUX1, MUX2 or MUX3) of the IC 10 as discussed below.
The exemplary circuitry illustrated in FIG. 76B, identified within the dashed box 1481, in conjunction with the H+ comparator system 50 (FIG. 7670 is used for power supply generation. More specifically, the power supply circuitry 1481 includes a transistor 1483, connected be-tween the positive terminal 1473 of the bridge 1469 and ground with its gate terminal connected to HDRIVE (FIG.
76a). also connected to the positive terminal 1473 is the 2 5 anode o! a diode 148 5 . The cathode o! the diode 14 8 5 i s connected to a terminal, identified in FIG. 76H as H+. A
power supply capacitor 1487 is connected between the B+
terminal and ground. ~ pair o! series connected resistors 1489 and 1491 are also connected between the B+ terminal and ground with the junction between the resistors 1489 and 1491 identified as BSENSE.
In operation, the comparator 1466 (FIG. 76~r) is used to aonitor the voltage at the HSENSE junction. A
fraction o! the voltage at the H+ junction, for example, Z.5 volts, and compare it with the voltage available at the VREF tersinal. when the HSENSB voltage is greater than the VREF voltage, the output o! the comparator 1466 goes high and turns on the trap:istor 1483 to shunt excess to ground. When the voltage at the HSENSE junc tion drops blow VREF, the comparator goes low which al lows the tran~iator 1483 to be turned 0!t to allow the ca pacitor 1487 to b charged up to the desired value, for example, 30 volts.
FIG. 76C illustrates exemplary circuitry for regulating the voltages at the VDD and AVDD pins and does not form a portion o! the present invention.
Hissing signal PHIarS 1440 !or the comparators 1412, 1440 (FIG. 75) and 1442 (FIG. 76) is illustrated in FIG. 77. 8lasing signals PHIJ~rS and NHI1~.S !or the quadcom-parator 200, ZOZ, 204 and 206 (FIG. 74), the H+ comparator 1466 (FIG. 76), the power monitor comparator 1462 (FIG.
76), the voltage amplifier 80 (FIG. 80) and the current asplilier 90 (FIG. 81) are illustrated in FIG. 78. The PHIAS and NBIaS signals from such circuitry are reference voltages that are used to set the operating current o! the ZO particular operational amplifier to which they are ap-plied. The above-mentioned biasing circuitry is in addi-tion to the auto-zeroing circuitry !or the voltage ampli-fier 80 and the current amplifier 90 are illustrated in FIG. 83 and identified as IOZ1T.
2s The circuitry illustrated in FIG. 77 is identi-rib by the function block 1440. The PBIAS circuit 1440 includes a transistor 1470 and a resistor 1499, connected im ~iu between 1~VDD and 11VSS, forming a voltage di-vider. The voltage divider produces a gate to source 30 voltage across the transistor 1470, identified as PHIaI.S.
The circuitry illustrated in FIG. 78 is used to generate the signals PHIaIS and NHIA8 !or the quadcompara-tors Z00, ZO=, Z04 (FIG. 74), the voltage amplifier 80 and the current amplifier 90. This circuitry includes its own 35 band gap regulator reference circuit connected parasitic transistors 1472, 1474, resistors 1476, 1478, a comparator 1480 and capacitors 1482 and 1484~ These signals are identified as PHI11S/I and NHIAS/I

to i~Li,cate that the signals are temperature independent since the circuitry includes the band gap reference. This band gap reference operates in a manner similar to the band gap reference 1406 described above with the exception of the additional capacitors 1482 and 1484 are used to control the biasing time of the circuitry. The output of the comparator 1480 is applied to the gates of transistors 1486, 1488 and 1490, forming current mirrors. The current airrors 1486 and 1488 are used to source the band gap 10~ regulator portion of the circuitry. The output current mirror 1490 is the NBIaS/I signal. The current mirror 1490 turns on transistors 1492 and 1494 which develop a gate to source voltage across a transistor 1496 which is the PBIJ~r.S/I relerence. The transistors 1498, 1500 and 1502 form start-up circuitry for the band gap regulator portion of the circuit.
The circuitry illustrated in FIG. 79 allows the aicroprocessor 30 to sense the ambient temperature in which the IC 10 is located. This circuitry includes a transistor 1504 and a diode connected parasitic transistor 1506. The voltage of the parasitic transistors is tem perature dependent. The resulting TEI~ signal is applied to a MtIX 66j and converted to a digital valve and read by the microprocessor 30.
The voltage amplifier 8o and ranging circuitry is illustrated in FIG. 80. This circuitry includes the voltage aaplitier 80, gain circuitry 84 and a plurality of xtTXes 86 to produce a voltage signal for ~r/D conversion that is at least halt scale. The voltage ranging may be controlled either automatically or manually to provide gains o! 1, Z, 4, 8 or 16 of the input voltage signal VNtJX
applied to a non-inverting input o! the voltage amplifier 80. The gain circuitry includes resistors 84a-84h and Htl7Ces 86a-86t. The gain circuitry is controlled by the VGJ1IN[4...0~ bus and the gain signal VG11IN 32h as pre-~~~~~~8 vious~ discussed. The resistor 84i and 1512 form test circuitry.
If the gain is one, the voltage signal is ap-plied directly to the A/D converter 78 by way of a MUX
SSa. In this condition the MUXes 86e and 88b disconnect the voltage amplifier 80 from the A/D circuitry 78 and the signal is connected directly to the A/D 78 by the MUX 88a.
During ranging, the MUXes 86a-86f connect the gain cir-cuitry 84 to an inverting terminal of the voltage ampli-~ fier 80. For gains other than one, the MUX 88b connects the output of the voltage amplifier 80 to the A/D 78. The MUXes 88a and 88b are selected by the AVSF register.
The comparator 74 is used for auto-ranging.
This comparator is referenced to a fixed voltage, for example +1.25 Vdc, developed by a pair o! serially coupled resistors 1508 and 1510, connected between VREF and AVSS.
The midpoint of these resistors 1508 and 1510 is applied to a non-inverting terminal of the comparator 74. The output of the comparator 74 is a CAMpIi signal, which is monitored by the flip-flop 1184 (FIG. 71) and forms a por-tion of the auto-range logic as discussed above. The MUX
86t is used for auto-zeroing. This MUX 86t shorts the in-verting and non-inverting terminals of the voltage ampli-fier 80 together to determine the offset correction value.
During this condition, the offset value of the voltage am-plitiar 80 is loaded into the Clip-Clop 888 (FIG. s9).
The ?~LTX 86t.is controlled by the VNULL signal available at the output o! the bullet amplifier 756 (FIG. 66).
Current amplifier 90 ranging is accomplished by the current mirror 9Z (FIG. 81) as previously discussed.
Currant inputs are applied to the current channel IMUX
(FIG. 73). This channel Il4tJX is tied to the current mir-ror 9Z and to an inverting input of the current amplifier 90. A non-inverting terminal of the amplifier 9o is tied to analog ground to maintain the current input channels MtJXO, I~JXi, 1~JXZ and lnJX3 at virtual ground. Negative currents (e.g. , currents flowing out of the 1~DC0 pin) to be ~~~~~~8 range~;re generated, for example, by connecting an exter nal ~sistor (not shown) between the MUXO pin and a nega tive voltage source. This causes ranged currents to flow out o! the 1~JX0, MUX1, MUX2 or MUX 3 pins since these pins are maintained at virtual ground.
The MUXes 96a and 96b connect the output signal IOUT/I from the currant mirror 92 either to the output pin lrDCO or to the analog ground bus AVSS. More specifically, the MUX 96a is used to connect the output signal IOUT/I of the currant mirror 92 to the MXO pin under the control of a signal IOUTONh; available at the output o! a HAND gate 759 (FIG. 66). The signal IOUTONh indicates that the in-tegrator is not in a reset mode. An INTRESh signal as well as test signals are applied to the NAND gate 759.
The MLJX 96b is used to connect the current mirror 92 out put signal IOUT/I to the analog ground bus under the con trol o! a signal DISCHh, available at the output of a bu!!er 757. The input to the bu!!er 757 is an integrator reset signal INTRESh, available at the output of the AND
gate 534 (FIG. 62).
The ML1X file is used !or auto-zeroing the cur-rent amplifier 90. Specifically, the I~tTX life connects both the inverting and non-inverting inputs o! the currant amplifier 90 to the analog ground bus AVSS under the con-trol o! a signal CSIDtTh, available at the output o! an in-verter 581 (FIG. 65). The inverter 581 is serially con-nected to the output o! the inverter 580. The input to the inverter 580 is the signal CAZh, which indicates the current aaplilier 90 is being auto-zeroed.
The MTJX lilb is used to connect the inverting input o! the amplifier 9o to the IMUX output o! the MUXes 68 (FIG. 73) whenever the amplifier 90 is not being auto-zeroed.
The ranged current from the current mirror 92 may then be dropped across an external resistor (not shown) to convert the signal to a voltage and converted by the 7~/D converter 78 as discussed above.

2.~~~~~8 The current mirror 92 is illustrated in FIG. 82.
The ~rrent mirror 92 includes current divider transistors 1512, 1514, 1516, 1518 and 1520, shunt transistors 1522, 1524, 1526 and 1528 and current mirrors 1530, 1532, 1534 and 1536. I~JXes 1538, 1540, 1542 and 1544 control current shunting while MUXes 1546, 1548, 1550 and 1552 control the gain of the circuitry. These MUXes are controlled by the CGAIN[3...0] bus discussed above.
Negative currants are directed into the current mirror 92 at IIN/I. This input current is divided into f ive parts by the current divider traps istors 1512 , 1514 , 1516, 1518 and 1520 which are all connected in parallel.
Hors specifically, the sizes of the transistors 1512 and 1514 are maintained equal at a value, for example, A. The sizes of transistors 1516, 1518 and 1520 era 2A, 4A and 8A, respactivsly. Since the transistors 1512, 1514, 1516, 1518 and 1520 are connected as current mirrors, the cur-rent through each of the transistors will be a function of the size of the transistor. Thus, the output of transis-tots 1512 and 1514 each will be 1/16 of IIN/I. The output of transistor 1516 will be 1/8 IIN/I. The output of tran-sistor 1518 will be 1/4 IIN/I. The output of the transis-tor 1520 will be 1/2 IIN/I. These fractions of the input current IIN/I art either summed together to produce the desired gain which is controlled by the MUXes 1546, 1548, 1550 and 1552 and directed to the output by way of the current mirrors 1530, 1532, 1534 and 1536 or shunted around the mirrors 1530, 1532, 1534 and 1536 by way of the transistors 15ZZ, 1524, 1526 and 1528 and the tiUXas 1538, 150, 154 and 1544.
This is an important aspect o! the invention.
Specifically, in known bipolar current ranging circuitry (for example, as disclosed in U. S. Patent No. 4, 626, 831) , the currant dividers are cascaded. However, cascading of the current dividers is not viable 'for the IC 10 because o! the relatively small operating voltage (a. g., +5.0 vdc ~ .

ENT AND VOLTAGE AMPL_TF'T~ ~FpnT,,r~
Representative circuitry for the voltage and current amplifier 80 and 90 is illustrated in FIG. 84.
These amplifiers are differential input amplifiers defin ing an internal bias current ITRIM/I. The differential inputs era identified as PLUS/I and MINUS/I. This inter-nal bias current flows through resistors 1546 and 1548 which controls the offset voltage that appears at the am-plifier 80 and 90 output. Known techniques have attempted to control the offset voltage by externally adjusting the resistor values of resistors 1546 and 1548, which requires precision variable resistors, such as digital-to-analog converter (DAC). Such DACs are relatively expansive. The auto-zeroing circuitry in accordance with the present in-vention, obviates the need for DACs and instead controls the bias current ITRIH/i to control the voltage across the resistors 1546 and 1548 to control the offset voltage.
The bias current is ranged by a current dividing circuitry illustrated in FIG. 83. The ranged bias current is then applied to the saplitier 80 or 90 to control the bias cur-rent and the offset voltage.
The bias current ranging circuitry includes the IdOXas 1600, 1602, 1604, 1606, 1608 and 1610, the current airrors -1612, 1614, 1616, 1618, 1620, 1622 and 1624 and the transistors 1626 and 1628. The I~LJXes 1600, 1602, 1604, 1606, 1608 and 1610 are controlled by the VZKRO[5...0y bus for the voltage aaplitier 80 and the CZERO[5...0] bus for the current aaplitier 90 as discussed above. These current airrors are connected in parallel to 3o allow the bias current to be divided into coaposite values and allow selected portions to be added together to gener-ate the ranged current and operate in a siailar manner as the current airror 9Z.

The INCOM eoaaunication controller (ICC) 29 pro-vides the aiaroprocusor 30 access to a two way comaunica-tion network, known as INCOlt, described in detail in U. S .
Patent No. 4,644,566.

This' stroller 29 provides modem functions, serializa-tion~i~deserialization of messages and implements the re-quired network protocol. The ICC 29 is capable of operat-ing both as a master and slave controller. Master opera-s tion is inhibited unless a permissive f lag is set in the configuration register CFR.
The microprocessor 30 communicates with the ZCC
29 through sight interlace registers located in memory ad-dress space. Four registers are used to transfer INCOM
aessages between the ICC 29 and the microprocessor 30, while the other four are used for setting the communica-tion address, speed, modulation method and control of transmit/receive operations.
The ICC 29 supports a fast status request mes sage which will reduce network response time. Since the transmit and receive registers !or the ICC 29 are indepen dent o! each other, a message, such as the last status, can be periodically updated in the transmit registers.
This allows the ICC 29 to transmit the response without the microprocessor 30 intervention when a fast status re-quest is received.
A general block diagram o! the ICG 29 is illus-trated in FIG. 85. The ICC 29 includes a processor bus interlace 1690, illustrated in FIGS. 94-101; a transceiver serial shift register 1692, illuatratad in FIGS. 102 and 103; a digital demodulator 1694, illustrated in FIGS. 86-92 and control logic circuitry 1696, illustrated in FIGS.
104-110.
PROCESSOR BL~lS INTERFACE
The microprocessor 30 communicates with the ICC
29 by way o! interlace registers IGH, IC1~L, ICM3, ICM2, IGI1, ICMO, ICSR snd ICCR located in memory address space as illustrated in Table 7. The configuration o! these registers is illustrated in FIGS. 94-101.
The registers IC~h and IC~Ii are address regis-ters, identified with the reference numeral 1700. The registers ICHO, IChl, ICM2 and ICH3 are message registers, ident~,ied with the reference numeral 1702. The registers ICCR,~nd ICSR are control and status registers, identified with the reference numeral 1704.
All of these registers are addressed by the microprocessor 30 by placing the appropriate address on the address bus ADDR(3...0]. The address is decoded by decoding circuitry, identified with the reference numeral 1706 (FIG. 94). The address decoding circuitry 1706 gen erates decoding signals DECAH, DECAL, DECM3, DECM2, DECM1, DECHO, DECSR and DECCR that are applied to D inputs of flip-flops 1708, 1710, 1712, 1714, 1716, 1718, 1720 and 1722. Register select signals SELAH, SELAL, SELM3, SELM3, SELM1, SELMO, SELSR and SELCR are available at Q outputs of these Clip-flops.
All of these registers are read write registers with the exception of the ICSR status register, which is a read only register. Read and write operations are con-trolled by RDCLK and WRCLR signals, which are generated by circuitry which includes a read write control flip-flop 1724, a NOR gate 1726 and an inverter 1728. The WRCLK
signal is available at the output of the NOR gate 1726.
The RDCLR signal is available at the output of the in-verter 1725~ A READ signal generated by the microproces-sor 30 is applied to a D input of the Clip-Clop 1724 by way of the internal control buy CPUCTL(3...0]. A Q output troa the tlip-flop 1724 is applied to one input of the dual input NOR gate 1726 to develop the WRCLR signal. The other input to the NOR gate 1726 is the phase 2 clock sig-nal PH2, available at a non-inverting output of the an in-verter 1730. A ~ output of the flip-Clop 1724 is applied to an input of the inverter 1728 to generate the RDCLK
signal.
Tiaing !or the read write control flip-Clop 1724 se well a~ for the addrue decode flip-llop~ 1708, 1710, 1712, 1714, 1716, 1718, 1720 and 1722 is provided by the P~i3 and ~' signal, available at the outputs of the in-verter 1730. Hors specifically, the PH2 signal, available at the non-inverting output of the inverter 1730, is ap-2j~~~58 t plied ~o E inputs of the flip-flops 1708, 1710, 1712, 1714, 1716, 1718, 1720, 1722 and 1724. The ~A~' signal, available at the inverting output of the inverter 1730, is applied to EN inputs of these flip-flops.
These flip-flops are all reset by the micropro-cessor 30. More specifically, a ~ signal is applied to CDN inputs of these flip-flops. The ~ signal is available at the output of an inverter 1732. The signal is generated from a RESET signal, applied to the input of the inverter 1732 by way of high gain inverters 1734 and 1736. The ~ signal is also applied to the address register 1700, the message register 1702 and the control and status register 1704. This allows the regis ters to be set to zero on system reset.
The address decoding circuitry 1706 for all of the above-identified registers is illustrated in FIG. 95.
This circuitry consists o! AND gates 1738, 1740, 1742, 1744, 1746, 1748, 1750 and 1752 and inverters 1754, 1756, 1758, 1760, 1762, 1764, 1766, 1768, 1770, 1772 and 1774.
The outputs of the AND gates 1738, 1740, 1742, 1744, 1746, 1748, 1750 and 1752 are the address decode signals DECSR, DECCR, DECM3, DECM2, DECMl, DECMO, DECAL and DECAH, re-spectively. More specifically, the address signals ADDRO, ADDR1, ADDRZ and ADDR3 from the address bus ADDR[3...0]
are applied to high gain inverter pairs 1754 and 1756;
1758 and 1760; 1762 and 1764; and 1766 and 1768. The out-put o! the inverter 1756 is applied to inputs of AND gates 1738, 1742, 1746 and 1750. The output of the inverter 1754 is applied to inputs of the AND gates 1740, 1744, 1748 and 1752. The output of the inverter 1760 is applied to the inputs of the AND gates 1738, 1740, 1746 and 1748.
The output of the inverter 1758 is applied to the inputs of the AND gates 1742, 1744, 1750 and 1752. The output of the inverter 1764 is applied to the inputs of the AND
gates 1738, 1740, 1742 and 1744. The output of the in-verter 1762 is applied to the inputs o! the AND gates 1746, 1748, 1750 and 1752. The output of the inverter ~~~~~~5 1768 ~ applied to the inputs of the AND gates 1738, 1740, 1742, 1744, 1746, 1748, 1750 and 1752.
Control signals IOOFF and ANABS from the micro processor 30 are used to enable or disable the address decoder 1706. The IOOFF signal is used to disable I/O de vices during a test mode. The ANABS signal is a register select signal from the microprocessor 30 master chip ad-dress decoder which enables address decoding on a block basis. The IOOFF signal is available at the output of the inverter 1770 and is applied to the inputs of the AND
gates 1738, 1740, 1742, 1744, 1746, 1748, 1750 and 1752.
The ANAHS signal is applied to a pair of high gain inver-ters 1772 and 1774. The output of the inverter 1774 is applied to the AND gates 1738-1752.
The ICJ~I and IC~rL address registers 1700 are shown in FIG. 96. These registers are byte wide read write registers, used to set the communication bit rate, modulation method and the 12-bit INCOH address for the ICC
29. Both of these registers are set to zero on reset and power up.
Referring first to the ICAIi register, bits ICl~IH[7,6] determine the communication bit rate of the Icc 29. Hits ICAIi[5,4] determine the modulation method used by the ICC 29. Hits ICJ,I~I[3...0] determine the higher order 4 bits o! the INCOH address.
The ICJ~H register includes the flip-flops 1776, 1778, 1780, 1782, 1784, 1786, 1788 and 1790. The data bus D11T71 [ 7 . . . 0 ] is connected to D inputs o! these f l ip-f lops to allow the microprocessor 3o to write to this register.
Q outputs of these flip-flops are connected to the data bus DATA[7...0] by way of tristate devices 1792, 1794, 1796, 1798, 1800, 1802, 1804 and 1806 for read operations.
The Q outputs of these Clip-flops are also attached to an internal control bus ICAH[7...0].
The tristate devices 1792, 1794, 1796, 1798, 1800, 180Z, 1804 and 1806 are under the control of a two input NAND gate 1808. A RDCLR signal is applied to one input along with the address decode signal SEIaH to allow the h~coprocessor 30 to read this register by initiating a re~i operation and addressing $0028.
Write operations to the ICAH register are con trolled by a two input HAND gate 1810. One input to the NAND gate 1810 is the address decode signal SELAH. The other input to the NAND gate 1810 is a WRCLK signal. The output of the NAND gate 1810 is applied to an inverter 1812. A non-inverting output from the inverter 1812 ap-plied to EN inputs of the flip-flops 1792, 1794, 1796, 1798, 1800, 1802, 1804 and 1806. An inverting output of the inverter 1812 is applied to E inputs for timing.
The ICAN register flip-flops (1776-1790) are set to zero on reset. More specifically, an ~ ~ signal, available at the output of an inverter 1814 is applied to CDN inputs to sat theca flip-flops to zero on reset.
The ICAL register is a byte wide register that datarminaa the lower order 8 bite of the ZNCOM address.
This register includes the flip-flops 1816, 1818, 1820, 1822, 1824, 1826, 1828 and 1830. The data bus DATA[7...0]
is applied to D inputs of thaw flip-flops to provide for write operations. Q outputs of thaw flip-flops are tied to the data bus DATA[7...0] for road operations by way of triatate davicaa 1832, 1834, 1836, 1840, 1842, 1844 and 1846. The Q outputs of thaw flip-flops era also tied to an internal control buy ICAL[7...0~.
Read operations of this rsgiater era controlled by a two inpiut NAND gate 1848. One input to the NAND gate 1848 is the addraaa decode signal SELAL. The read clock RDCI~1C signal is applied to the other input. The output of the NAND gate 1848 ii applied to a control tarainal OEN of each of the triatate davicaa.
write oparation~ to this register era controlled by a two input HAND gate 1850. One input to the HAND gate 1850 is the write clock wRCLR signal. The register select signal SELAL is applied to the other input. The output of the HAND gate 1850 is applied to an inverter 1852. A non-inverting output of the inverter 1852 is applied to EN in-~~~~;~5~

puts ~ these flip-flops. An inverting output of the in-vertei 1852 is applied to E inputs.
This register is set to zero on device reset and power up. More specifically, the 3~E~ signal is applied to the CDN inputs of these flip-flops.
The registers ICM3, ICM2, ICMl and ICMO, illus-trated in FIGS. 97-100, are byte wide read write registers which include a transmit buffer and a receive butter. These registers are used to transfer INCOM
messages between the ICC 29 and the microprocessor 30.
These registers era not true read write registers since read operations access the receive butter and write operations write to the transmit butter.
The ICM1 register is an 8-bit register and con tains the INCOM message bits 10 through 3. The transmit butter for this register includes the flip-flops 1854, 1856, 1858, 1860, 1862, 1864, 1866 and 1868. The receive butter for this register includes the flip-flops 1870, 1872, 1874, 1876, 1878, 1880, 1882 and 1884.
The transmit buffer flip-!lops 1854, 1856, 1858, 1860, 1862, 1864, 1866 and 1868, can only be written to by the microprocessor 30. More :pacifically, the data bus D1~T1~ ( 7 . . . O j is applied to D inputs of these f l ip-f lops . Q
outputs of these flip-flops contain INCOM massage bits 2 5 TDAT7~1 ( 10 . . . 3 ] .
Write operations to the transmit butter era con-trolled by a two input N~ND gate 1866. The write clock WRCIdC signal is applied to one input. An address decode signal SEI8t1 is applied to the other input. The output of the NJ1ND gate 1866 is applied to an inverter 1888. 1~ non-inverting output of the inverter 1888 is applied to EN in-puts of these flip-flops. 7~n inverting output of the in-verter 1888 is applied to E inputs of these !lip-!lops.
The receive butter of the ICxi register includes the flip-flops 1870, 1872, 1874, 1876, 1878, 1880, 1882 and 1884. INCOlt messages, received on a receive data bus RDl~T7~ ( 10 . . . 3 ) are applied to D inputs of these f lip-f lops .
The contents of these flip-flops may be read by the micro proces~r 30 on the data bus DATA[7...0) by way of tri-stat~ devices 1892, 1894, 1896, 1898, 1900, 1902, 1904 and 1906. These tristate devices are under the control of a two input NAND gate 1908. The read clock RDCLK signal is applied to one input. An address decode signal SELM1 is applied to the other input.
Timing for the receive buffer and transmit buffer flip-flops is provided by a signal SRTOMR (FIG.
109) which will be discussed below. This signal SRTOMR is applied to an inverter 1910. A non-inverting output of the inverter 1910 is applied E inputs of these flip-flops.
An inverting output of the inverter 1910 is applied to EN
inputs.
The receive and transmit buffers are set to zero on reset and power up. More particularly, an 'Iii sig nal available at the output of an inverter 1890 is applied to CDN inputs of these flip-flops. The ~ signal is applied to an inverter 1889 to gsnerate an MRCLR signal, used to reset the registers ICM2, ICM3 and ICMO registers, as discussed below.
The IC142 register is an 8-bit register and con-tains the INCOK mes:age bits 18 through 11. The transmit butler o! the ICH2 register includes the flip-flops 1912, 1914, 1916, 1918, 1920, 1922, 1924 and 1926. The receive butter of the ICl~2 register includes the flip-Elope 1928, 1930, 193, 1934, 1936, 1938, 1940 and 1942.
The tran~ait buffer may only be written to by the aicroproceasor 30. More particularly, the data bus DATA[7...0] i~ applied to D inputs o! the flip-flops 1912, 1914, 1916, 1918, 1920, 1922, 1924 and 1926. Q outputs of these flip-flops are tied to a transact data bus TDATA[18...11].
writs operations to the transact butter are under the control of a two input NAIdD gate 1944. The write clock wRCLIC signal is applied to one input. Tha ad dress decode signal SEI~i2 is applied to the other input.
The output of the NAND gate 1944 is applied to an inverter 1946. A non-inverting output of the inverter 1946 is ap-~~~~?~S

plied~o EN inputs of the flip-flops 1912, 1914, 1916, 1918, 1920, 1922, 1925 and 1926. An inverting output of the inverter 1946 is applied to E inputs of these flip-f lops .
The receive buffer of the ICM2 register includes the flip-flops 1928, 1930, 1932, 1934, 1936, 1938, 1940 and 1942. INCOM massages, received on an internal bus RDATA[18...11], are applied to D inputs of these flip-flops. The message bits contained in these flip-flops may be read by the microprocessor 30 on the data bus DATA[7...0] by way of tristate devices 1950, 1952, 1954, 1956, 1958, 1960, 1962 and 1964. These tristate devices are under the control of a two input NAND gate 1966. The register decode signal SELM2 is applied to one input. The read clock RDCLR signal is applied to the other input to allow the microprocessor 30 to road the contents of this butter by initiating a read operation and placing the ad-dress $002C on the address bus P1DDR[3...0].
Timing for the receive buffer flip-flops is pro vided by the SRTOMR signal and an inverter 1968. More specifically, the SRTOMR signal is applied to an input of the inverter 1968. A non-inverting output of the inverter 1966 is applied to E inputs of these flip-flops. An in verting output o! the inverter 1968 is applied to EN in puts of these receive buffer flip-flops.
The transmit and receive butter flip-flops are set to zero by the signal available at the output of the inverter 1969. This signal is applied to CDN
inputs o! each of these flip-flops.
The ICM3 message register contains the INCOM
massage bits 26 through 19. The transmit butter for the register ICM3 includes the flip-flops 1970, 1972, 1974, 1976, 1978, 1980, 1982 and 1984. The receive buffer for this register includes the flip-flops 1986, 1988, 1990, 1992, 1994, 1996, 1998 and 2000.
The microprocessor 30 writes to the transmit butter by way of the data bus DATA[7...0], applied to D
inputs o! the flip-flops 1970, 197x, 1974, 1976, 1978, 1d 1980, ~p82 and 1984. Q outputs of these flip-flops are applied to the transmit data bus TDATA[26...19].
Writs operations to the transmit buffer are under the control of a two input HAND gate 2002. The write clock WRCLK signal is applied to one input. The register select signal SELM3 is applied to the other in-put. The output of the NAND gate 2002 is applied to an inverter 2004. A non-inverting output of the inverter 2004 is applied to the two EN inputs of these f lip-f lops .
An inverting output o! the inverter 2004 is applied to E
inputs of these flip-flops.
The receive buffer includes the flip-flops 1986, 1988, 1990, 1992, 1994, 1996, 1998 and 2000. INCOM mes-sag~ bits received from the INCOM network are transmitted by the receive data bus RDATA[26...19] and applied to D
input: of these Clip-flops. The microprocessor 30 can read the contents of these flip-flops on the data bus DATA[7...0] by way o! tristats devices 2008, 2010, 2012, 2014, 2016, 2018, 2020 and 2022. These tristate devices are under the control of a two input HAND gate 2024. The read clock RDCLR signal is applied to one input. The re-gister select signal SELM3 is applied to the other input to allow the microprocessor 30 to read the contents of this butler any time it initiates a read operation and it places the address $002D on the address bus ADDR[3...0].
Tiaing for the ICM3 receive butter is provided by the SRTOIDt signal and an inverter 2026. More spe-citically, the SRTOMR signal is applied to an input of the inverter ZOZ6. A non-inverting output of the inverter 2026 is applied to E inputs of th~s~ flip-flops. An in-verting output of the inwrtsr 2026 is applied to EN in-puts of th~s~ flip-flops.
The ICX3 transit and racsiv~ butlers are sat to zero by the signal. Thin signal is available at the output of the inverter 2006. This signal is applied to CDN inputs of th~s~ flip-flops.
The ICKO register is for trans'itting and re-c~iving a control and status bits o! the INCOM message.

The t~.nsmit buffer for this register includes the flip-flops~ 2028, 2030 and 2032 and the tristate devices 2034, 2036, 2038, 2040, 2042, 2044, 2046 and 2048. The receive buffer includes one flip-flop 2029.
Bit ICMO[7] corresponds to bit 2 of the INCOM
musage. For transmit operations, this bit is written by the microprocessor 30 over the data bus DATA[7] and ap-plied to a D input of the flip-flop 2028. write opera-tions to this flip-flop 2028 are under the control of a two input NAND gate 2050. A writs clock WRCLK signal is applied to one input. A register select signal SELMO is applied to the other input. The output of the NAND gate 2050 is applied to an inverter 2052. A non-inverting out-put of the inverter 2052 is applied to an EN input of the flip-flop 2028. An inverting output of the inverter 2052 is applied to an E input of the flip-flop 2052. A Q out-put o! the flip-flop 2028 is applied to the transmit data bua TDATA[2].
Hit 2 0! an incoming INCOH message is received along an internal rsc~iw~ data bus RDATA[2] and applied to a D input o! a flip-flop 2029. Timing for this flip-flop 2029 is provided by the SRTOI~R signal. The SRTOI~Qt signal is applied to an inverter 2031. A non-inverting output of the inverter 2031 is applied to an E input o! the flip flop 2029. An inverting output of the inverter 2031 is applied to an EN input of this flip-flop. In order to al-loy the constants o! this flip-flop to be read by the microprocessor 30, a Q output of the flip-flop 2029 is ap-plied to the data bus DATA[7] by way o! a trfstate device 2033. The tristate device 2034 is under the control of the NAND gate 2064.
IClIO bits ICHO[6...2] are reserved for test.
These bits are read by the microprocessor 30 on the data bus DATA[6...2]. Hore specifically, a read teat bus RTH[6...2] is tied to the data bus DATA[6...2] by way of tristate devices 2036, 2038, 2040, 2012 and 2044. These tristate devices are under the control of the HAND gate 2054. One input to the NAHD gate signal 2054 is the read ~~~~~58 clock ~tgnal RDCLK. The SELMO register select decode sig-nal i~ applied to the other input.
Two AND gates 2055 and 2056 are also part of the test circuitry. More specifically, the AND gate 2055 is a three input AND gate. A write clock signal wRCLK is ap plied to one input. An IMCO register decode signal SELMO
is applied to another input. A test signal TEST is ap-plied to the other input. The output o! the AND gate 2055 is applied to the AND gate 2056 along with DATA[2]. The output o! the AND gate 2056 is applied to a write test bus WTH[2...0].
The bits ICMO[1,0] are status bits. Both of these bits are true read write bits. The microprocessor 30 can write those bits by way o! the data bus DATA[1,0], which is tied to D inputs o! the flip-!lops 2030 and 2032.
Write operations to these flip-!lops are controlled by the NAND gate 2050 and the inverter 2052. More apacilically, the non-inverting output o! the inverter 2052 is applied to an EN input o! the flip-flops 2030 and 2032. The in-vetting output o! the inverter 2052 is applied to E inputs o! these flip-!lops. Q outputs o! these !lip-!lops are tied to an internal status bit read bus RSB[26,25]. The Q
outputs are also tied to the data bus DATA[1,0] by way of the tristate devices 2046 and 2048 to allow the micropro-cessor 30 to read the contents o! these flip-!lops. The tristate devices 2046 and 2048 are under the control of the read control NaND gate 2054.
The !lip-!lops 2028, 2029, 2030 and 2032 are all set to zero on systaa reset. Hore specifically, the signal is applied to CDN inputs o! these flip-!lops. This signal is available at the output o! an inverter 2057.
The ICCR register is a byte wide read write con-trol register used to control operation o! the ICC 29.
This register is illustrated in FIG. 101 and includes the flip-!lops 2058, 2060, 2062 and 2064 as well as the AND
gates 2066, 2068, 2070 and 2072. Kore particularly, the data bus D~1T7~(7...4] is applied to D inputs o! the llip-~.~ ~~~~8 !lops ~p58, 2060, 2062 and 2064. Timing !or these flip-llop~ 2058, 2060, 2062 and 2064 is provided by a write control NAND gate 2074 and an inverter 2076. More spe-cifically, an ICCR register decode signal SELCR is applied to one input o! the NAND gate 2074. A write clock signal WRCLK is applied to the other input. The output o! the NAND gate 2074 is applied to an input o! an inverter 2076.
An inverting output o! the inverter 2076 is applied to E
inputs o! these flip-!lops. A non-inverting output of the inverter 2076 is applied to EN inputs o! these flip-!lops.
Q outputs o! these flip-!lops 2058, 2060, 2062 and 2064 are applied to an internal control bus CR[7...4]
for the control bits ICCR[7...4] discussed below. These Q
outputs are also connected to the data bus DATA[ 7 . . . 4 ] by way o! tristate devices 2078, 2080, 2082 and 2084 to allow the contents o! these flip-!lops 2058, 2060, 2062 and 2064 to be read by the microprocessor 30. The tristate devices 2078, 2080, 2082 and 2084 are under the control o! a two input NaND gate 2086. One input to the NAND gate 2086 is an ICCR register decode signal SELCR. The write clock signal WRCL1C is applied to the other input.
The flip-!lops 2058, 2060, 2062 and 2064 are set to zero on systas reset. Hore particularly, an I
signal is applied to CDN inputs o! these flip-!lops. This signal is available at the output o! an inverter 2088.
The bits ICCR[7...4] are control bits. Bit ICC'R[7] enables the interrupt operation o! the ICC 29.
Hit ICCR[6] enables automatic response to a last status request message. The control bit ICCR[4] is not used.
3o The tiit ICCR[5] is a peraissive bit that allows the INCOX co~sunication controller 29 to be switched into the caster code. A zero in this bit prevents the ICC 29 !roe entering the master mode while a one permits it.
This bit is also set to zero on systas reset. This is ac-co~plished by a two input AND gate 2089. one input to the AND gate 2089 is a permissive bit !roe the CFR status register bit CFR[7]. The other input to the AND gate 2089 ~~~~~58 is tha~,~ signal. The output of the AND gate 2089 is appl~d to a CDN input of the flip-flop 2062.
The bits ICCR[3...0] are command bits and are defined below. These command bits are written by the microprocessor 30 on the data bus DATA[3...0] and applied to inputs of the AND gates 2066, 2068, 2070 and 2072. A
write control signal, available at the output of the in-verter 2086, is applied to the inputs of the AND gates 2066, 2068, 2070 and 2072 along with the data bus DATA[3...0]. The input of the inverter 2086 is connected to the output of the write control HAND gate 2074. The outputs of the AND gates 2066, 2068, 2070 and 2072 are ap-plied to the internal ICCR register bus CR[3...0].
These bits ICCR[3...0] ars not true read write bits. These bits will always read zero when read by the microprocessor 30. More specifically, the bits ICCR[3...0] are tied to the data bus DATA[3...0] by way of tristate devices 2090, 2092, 2094 and 2096. The inputs to these tristata devices are grounded. These tristate de vices are under the control of the HAND gate 2086.
The ICSR is a byte wide read only status regis-ter which contains the ICC 29 status !lags that the micro-processor 30 coamunicates to the INCOH communication con-troller 29. This register includes the tristata devices 2098, 2100, 2101, 2101, 2106, 2108, 2110 and 2112. An in-ternal ICSR control register bus SR[7...0] is applied to the inputs o! these tristate devices. The outputs of these tristate devices are applied to the data bus DATA[7...0]. These tristate devices are under the control 0! a two input NAND gate 2114. An ICSR register select decode signal SEhSR is applied to one input. A read clock signal RDCIdC is applied to the other input.
The digital demodulator 1694 demodulates incom ing INCOK messages and provides a demodulated output bits identitied as a signal DElsODAT. This demodulator 1694 is illustrated in FIGS. 86-93. A block diagram !or the digi tal demodulator 1694 is illustrated in FIG. 86. The digi tal d~pdulator 1694 includes a timing generator 2116, a raceiyer correlator 2118, demodulator control circuitry 2120 and a bit counter 2122.
The timing generator 2116 is illustrated in FIGS. 87 and 88. Timing diagrams for the various outputs of the timing generator era illustrated in FIGS. 111 and 112. The timing generator 2116 includes a master clock generator 2117, illustrated in FIG. 87 and a bit phase timing generator 2119, illustrated in FIG. 88.
The ICC 29 may be configured to operate at a variety of bit rates depending on the transmission mode selected. Exemplary bit rates era delineated in Table 19.
The master clock generator 2117 provides the various clock signals for the selectable bit rates. The ICAii register bits ICAH[7,6] determine the bit rate, while the bit ICAH[5] detersines the modulation method. As discussed above, since the ASR and FSR modulation schemes are com-patible, thus only one bit ICAH[5] is necessary to select between baseband and FSR/ASR. The selected bit rate is available at the output of a MUX 2124. A Z output of the I~iIJX 2124 is a bit phase clock signal BITPHCR which cor-responds to the selected bit rate and modulation method in accordance with Table 19. The command bit ICAH[5] is ap-plied to a select input S o! the I~iUX 2124. The inputs to the 141JX 2124 are lrom a baseband M1JX 2126 and an ASR/FSK
MOX 2128.
The various baseband bit rates in accordance with Table 19 are applied to the inputs of the l~ttTX 2126.
The various ASR/FSR bit rates, also in accordance with Table 19, are applied to the inputs of the FSR MUX 2128.
The various bit rates are selectable by the comaand bits ICAIi[7, 6] which are applied to select inputs of the XUxas 2126 and 21Z1. The various bit rates are all derived from the IC 10 crystal and the tising generator 2116 which in eludes counters 2130 and 2132.
The counter 2130 is used to generate the base-band bit rates. This counter 2130 includes the flip-flops 2132, 2134, 2136, a I~JX 2138, an inverter 2140 and an ex-clusi~ OR Bata 2142. If the ICC 29 is active, either a 7.37? I~iz or a 3.6864 I~iz crystal must be used. Depend-ing on the crystal selected, one of these frequencies will be available at the output of the exclusive OR gate 2142 based upon the phase 2 clock signal PH2 and an E02 signal applied to inputs of the exclusive OR gate 2142. The Eo2 is a signal that is 90~ out of phase with PH2.
The flip-flops 2132, 2134 and 2136 are connected as divide by two ripple counters. Kore specifically, Q
outputs of these flip-!lops are applied to D inputs. The output o! the previous flip-flop is also applied to a clock input CP o! the succeeding flip-flop. The outputs o! the flip-flops 2132, 2134 and 2136 thus era divided by two of the input. Because o! the option o! crystals, the output of the exclusive OR gate 2142 is applied to one in-put of the lKtlX 2138. The ~ output o! the !lip-flop 2132 is applied to the other input o! the MUX 2138. If a 7.3728 I~Iz crystal is used, the flip-flop 2132 divides this frequency by two to produce a 3.6864 l~iz signal at a Z output of the InJX 2138. If a 3.6864 I~iz crystal is used, this signal is applied directly to another input of the IKiJX 213 8 f or a baud rata of 15 3 . 6 Rbps . The I4tJX 213 8 is controlled by the configuration register bit ACFR(6], which is applied to an S input of the l4t~T7c 2138. The bit ACIrR[6] selects the divider ratio. The Z output of the I~tJX 2138 is s 3.6864 lrD~iz signal. This signal is applied to one input of the baseband l~itTX 2126 and also to a clock input CP o! another divide by two counter 2134 to produce a 1.8343? leis signal at its ~ output. The 1.8232 lrQiz sig-nal is applied to another input of the baseband InT7c 2126 for a baud rata of 76.8 Rbps. A ~ output of the counter 2134 is also applied to a clock input of another divide by two counter 2136. The output of this counter is applied to an inverter 2140. The output of the inverter 2140 is a 921.6 kHz signal. This signal is applied to another input of the baseband MUX 2126 for a baud rate of 38.4 Itbps sig-nal.

The 19.2 Kbps baseband baud rate signal is gen-erat~ by the circuitry which includes an exclusive oR
gate 2142, and flip-flops 2144 and 2146. More spe-cifically, the 921.6 kHz signal, available at the output of the inverter 2140, is applied to a Johnson counter, which includes the flip-flops 2144 and 2146. More spe-cifically, the output of the inverter 2140 is applied to clock inputs CP of the flip-flops 2144 and 2146. A Q out-put of the flip-flop 2144 is applied to a D input of the flip-flop 2146. A ~ output of the flip-flop 2146 is ap-plied to a D input of the flip-flop 2144. A Q output of the flip-flop 2146 is a 230.4 kliz signal. This signal is applied to one input of the exclusive OR gate 2142. The other input to the exclusive OR gate 2142 is the Q output signal troa the flip-flop 2144. The output of the exclu-sive OR gate 2142 is a 460.8 kHz signal, equivalent to a bit rate of 19.2 Rbpa.
The ASR/FSR bit rates signals are developed by the counter 2132 and an OR gate 2148. More specifically, the Q output o! the flip-flop 2144 is applied to one input of the OR gate 2148. This signal represents a 230.4 kHz signal. The Q output o! the flip-flop 2146 is also ap-plied to another input of the OR gate 2148. Lastly, the output of the inverter 2140 is applied to a third input of the OR gate 2148. The output of the OR gate 2148 is ap-pli~ to the counter 2132, which includaa 'the flip-flops 2150, 2152, 2156 and 2158, all connected a: divide by two counters with their ~ outputs connected to their D inputs.
Hore specifically, the output of the OR gate 2148 is a 230.4 kHs signal which is applied to a clock input Cp of the flip-flop 2150. This flip-flop 2150 divides the input frequency by two to generate a 115.2 kHz signal at its Q
output which i~ applied to one input the ASR/FSR MUX 2128 to generate a bit rate signal of 9600 bpe. A ~ output of the counter 2150 i~ applied to a clock input of the counter 2152 to generate a 57.6 kHs signal at its ~ out-put. This signal is applied to a clock input CP of the counter 2154 to generate a 28.8 klis signal at its Q out-put. this 28.8 kHz signal is applied to a clock input CP
of t~ counter 2156 to generate a 14.4 kHz signal at its Q
output. The 14.4 kHz signal is applied to a clock input CP of the counter 2158 to generate a 7.2 kHz signal at its Q output. This 7.2 kHz signal is applied to the ASK/FSK
InJX 2128 to generate a 300 bps signal.
The counters 2130, 2132 and the Johnson counter which includes the flip-flops 2144 and 2146 are set to zero by reset and during test. More particularly, an reset signal is applied to one input of a two input OR gate 2160. A signal from the write test bus WTB2 is applied to the other input. The output of the OR gate 2160 is applied to an input of an inverter 2162. output' of the inverter 2162 is applied to CDN inputs of the tlip-flops 2132, 2134, 2136, 2144, 2146, 2150, 2152, 2154, 2156 and 2158.
The FSK modulation scheme uses two carrier fre-quencies; 115.2 kHz and 92.16 kHz. The 92.16 kHz signal is available at the output of a decade counter 2164. This decade counter includes the flip-flops 2166, 2168, 2170 and 2172, as well as the NOR gate 2174 and an AND gate 2176. The 92.16 kHz signal is available as a signal FCA, available at a Q output of the flip-flop 2172. A 921.6 kliz signal is applied to clock inputs of the flip-flops 2166 and 2170. A ~ output of the flip-flop 2166 is ap-plied to s clock input CP of the flip-flop 2168. A Q out-put o! the flip-flop 2166 is applied to one input of a two input NOR gate 2174. The other input to the NOR gate 2174 is a Q output troa the flip-tlop 2170. The output of the NOR gate 217 is applied to a 0 input of the flip-flop 2166. A ~ output of the flip-flop 2168 is applied to a D
input of this flip-flop. A Q output of the flip-flop 2168 is applied to one input of a two input l~l~tD gate 2176. The other input to the AND gate 2176 is the Q output of the flip-flop 2166. The output of the 1~ND gate 2176 is ap-plied to s D input of the flip-flop 2170. The ~ output of the flip-flop 2170 is applied to a clock input o! the flip-tlop 2172, l~ ~ output o! the flip-flop 2172 is ap-208 ~,~~:
plied r10 its D input. The Q output of the flip-flop 2172 is t?~s a divide by ten of the input signal 921.6 kHz.
Tha decade counter flip-flops 2166, 2168, 2170 and 2172 are set to zero during reset and also during the test mode. Hore specifically, the output of the inverter 2162 is applied to the CON inputs of these flip-flops.
The 92.16 kHz FSK signal FCA as well as a 115.2 kHz carrier signal FC are used for FSK modulation. The FC
signal is available at the output of the counter 2150. A
115. Z kHz baseband signal CARR is available at the output of a NOR gate 2174. One input to the NOR gate 2174 is the output of the OR gate 2148. Another input to the NOR gate 2174 is the output ot,the counter 2150. The output of the NOR gate 2174 is a 115.2 kHz signal.
Signals FDC and CR10 are used for demodulation control. The signal FDC is a 230.4 kHz signal , available at the output of an inverter 2176. J~ 230.4 kHz signal from the output of the counter 2146 is applied to the in-put of the inverter 2176. This frequency is selected as a saapling frequency at twice the carrier frequency of 115.2 kHz.
The CR10 signal is used to reset the bit phase timing generator 2119, illustrated in FIG. 88. This CKlo signal is available at the output of an OR gate 2178.
There are three inputs to the OR gate 2178. Ona input is troa the Q output of the counter 2146. The other input is troa the output o! the inverter 2140. Lastly, a ~ output o! the flip-Clop 2144 is applied to the OR gate 2178. The inputs to the OR gate 2178 are thus signals representative 0! twice the carrier frequency (230.4 kHz), tour times the carrier frequency (460.8 kHz) and eight times the carrier frequency (91.6 k~Iz). l~a illustrated in FIG. 11Z, the CiciO signal available at the output of the OR gate 2178 will thus produce a one-zero-one pattern every one half cycle at 115.1 kIiz, equivalent to Z30.4Z kHz; the sampling frequency.
The bit phase timing generator 2119 is used to count the phase coherence of the carrier frequency. The bit use timing generator 2119 produces timing control signs PHCXAD, PHCKH, PHCKHD, PHCKCD, PHCKDD, PHCKD, PHCK
and PHRST. These signals ara illustrated in the form of timing diagrams in FIGS. 111 and 112. These timing signals are developed by flip-flops 2182, 2184 and 218x;
NOR gates 2188, 2190, 2192, 2194, 2196, 2198 and 2200 and inverters 2202, 2204, 2206, 2208, 2210, 2212, 2214 and 2216.
The bit phase clock signal BITPHCR is applied to the bit phase timing generator 2119. The BITPHCK signal is available at a Z output of the MUX 2124 (FIG. 87). The HITPHCK signal corresponds to the selected baseband or ASR/FSR baud rate. This BITPHCR signal is applied to the bit phase timing generator 2119 by way of high gain in-verters 2202 and 2204. The output of the inverter 2204 is applied to clock inputs CP of the llip-flops 2182 and 2184, which are configured as a Johnson counter 2185.
More spacitically, a Q output of the flip-flop 2182 is ap-plied to a D input of the flip-f lop 2184. A ~ output of the flip-Clop 2184 is applied to a D input of the flip-tlop 2182. The Q output of the flip-flops 2182 and 2184 are used to generate the timing signals. More spe-cifically, a Q output of the flip-flop 2182 is applied to an inverter 2206. A ~ output of the tl~p-flop 2182 is ap-plied to an inverter 2208. A Q output of the Clip-flop 2184 is applied to an inverter 2210. A ~ output of the flip-flop 2184 is applied to the inverter 2212. An output of the inverter 2206 is applied to inputs of the NOR gates 2190, 2192 and 2194. An output of the inverter 2208 is applied to inputs of the OR gates 2188, 2196 and 2198. An output troy the inverter 2210 is applied to inputs of the inverters 2194, 2196 and 2198. An output of the inverter 2212 is applied to inputs of the NOR gates 2188, 2190 and Z19~. The output of the inverter 2212 is also applied to an input of an inverter 2214 to generate the PHCR signal.
The HITPHQC signal, available at the output o! the in-verter 2204, i~ also applied to inputs of the NOR gates 2188, 2192, 219~ and 2196. The outputs of the NOR gates 2188 ,190, 2192, 2194, 2196 and 2198 are the bit phase timir~ signals PHCKAD, PHCKB, PHCKHD, PHCKCD, PHCKDD, PHCKD and PHCR.
A bit phase reset signal PHRST is generated every one-sixth bit at the output of a NOR gate 2200. A
timing diagraa for this signal is illustrated in FIG. 112.
The NOR gate 2200 is a two input NOR gate. one input is !roe an output of an inverter 2216. A reset signal is applied to the input of the inverter 2216. A D input of the Clip-flop 2186 is normally grounded. This flip flop 2186 is clocked by the ~ signal, available at the output of the Johnson counter 2185. The CR10 signal is applied to the CDN input of the flip-Clop 2186 to clear this llip-flop ovary one-half carrier cycle. A Q output of the flip-flop 2186 is applied to an input of the NOR
gate 2200 to generate the pulse signal PHRST at every one-sixth bit or 64 carrier cycles at 300 baud.
The carrier input signal is applied to a RXIN
terainal on the IC 10 by way of a buffer amplifier 2218 (FIG. 85). This signal is than applied to a pair of car rier confirmation circuits 2220 and 2222, which torn a portion of the receiver correlator 2118. These circuits operate at 90~ out of phase with respect to each other.
Each of these carrier confirmation circuits 2220 and 2222 exaaines the input carrier signal to datesaine if it is within the acceptable band of frequencies centered about the carrier. This is done on a cycle by cycle basis.
Lath carrier conliraation circuit 2220 and 2222 has two outputs. One output produces a pulse if the signal is within the pass band and the saaple phase of the input signal is a logic 1. The other produces a pulse it the signal is within the pass band and the sa'ple phase of the input signal is a logic 0. The four outputs ONEA, ZEROA, ONES and ZER08 are used as inputs to a series of phase counters 2224, 2226, 2228 and 2230 (PIG. 90) which are re-set by the PI~t.ST signal, which resets these counters every one-sixth o! s bit.

The digital demodulator 1694 requires phase co-here~~,e over the short term, i.e., over one and one-half cycles for frequency detection and is able to determine continued phase coherence in the longer term, i.e., one-s sixth of a bit or 64 carrier cycles at 300 baud to dis-criminate against noise. The digital demodulator 1694 thus senses both frequency and phase of an incoming signal over a one-sixth of a bit interval. It the input fre-quency is correct and maintains phase coherence for at least three-fourths of the one-sixth bit interval, a counter 2338 is incremented. After six of these one-sixth bit intervals are processed, the contents are examined.
It the counter counts up to four or more, a demodulated output bit DEKODAT is outputted.
Each of the carrier confirmation circuits 2220 and 2222 stores its three most recent samples of the in-coaing carrier by way of two stage shift registers 2224 (with stages 2223 and 2225) and 2226 (with stages 2227 and 2229j. The incoming carrier frequency is applied to a D
input of these shift registers 2224 and 2226. The shift registers 2224 is clocked at twice the carrier frequency by the signal P'DC. The shift register 2226 is also clocked at twice the carrier frequency with a signal The signal F6~ is available at the output of an inverter 2243. The output of each stage o! the shift registers 2ZZ4 and ZZZ6 is exclusive ORed with its input by way of the exclusive OR gates 2228, 2230, 2232 and 2234. Tha outputs o! the exclusive OR gates 2228, 2230 and 2232 and ZZ34 are l~NDed by AND gates 2236 and 2238. The outputs of the AND gates ZZ36 and 2238 are applied to D inputs of a third stags ship register 2240 and 2242.
Arsu~ing a 1-0-1 pattern exists vn a D input to the first stage ZZZ3, 2227 0! the shift registers 2224 and 2226, the Q output of the first stage 2223, 2227 0! the ' shift registers 2224 and 2226 and the Q output o! the second stage ZZZS, 2229, this jeans that the past sample, which is zero, is stored in the first stage 2223, 2227 and the saaple before that, which is a one, is stored in the secon~,.stage 2225, 2229. The present sample at the input of t~ first stage 2223, 2227 is stored on the next clock pulse.
The outputs of the first stage 2223, 2227 and the third stage 2240, 2242 of the shift registers are ap-plied to NaND gates 2244, 2246, 2248 and 2250 to generate the confirmation signals aRF.~, ~, ~g and ~'B.
pulse on either the ONFJ~r or ZEROA output means that over a relatively short term of one and one-half carrier cycles, the input carrier is generally in phase with the timing signals developed through crystal oscillator. More specifically, a Q output of the first stags 2223 of the shift register 2224 is applied to one input of the NAND
gate 2244. J,nother input to the NAND gate 2244 is a Q
output of the third stage 2240. The Q output of the third stage 2240 is also applied to an input of the Nl~rNND gate 2246 along with a ~ output of the first stage 2223. The outputs o! the NJ~ND gates 2244 and 2246 are the signals and ~. These NAND gates 2244 and 2246 will pro duce a pulse on every other sample provided the three stored samples form a 1-0-1 pattern. It the most recent sample is a logical 1, output o! the HAND gate 2244 will be a logical 1. I! the most recent sample is a 0, the output o! the NI,ND gate 2246 will be a logical 1. The N7~rI~lD gates ZZ48 and 2250 of the carrier confirmation cir-cuit 22Z? operate in a similar manner to produce to aR'EB
acrd ~E1~38 signals.
The spilt registers 2224, 2226, 2240 and 2242 are set to zero on reset. Hore specifically a isig nal available at the output o! an inverter 2Z5Z is applied to CDN inputs o! these shift registers.
The phase counters 2224, ZZZ6, 2228 and 2230 are used to separately count the number of pulses developed on the four outputs (e.g., NaND gates 2244, 2246, 2248 and ZZ50) of the confirmation circuits 2ZZ0 and ZZ2Z during a time interval equal to one-sixth of a bit. If any of these counters reaches and count of 4S during 64 carrier cycles which occur during one-sixth bit intervals at 300 baud d~ 12 out of 16 at 1200 baud, it is assumed that a valid carrier signal existed for that one-sixth bit inter-val.
The output signals 3R'~~, , tag and signals from the carrier confirmation circuits 2220 and 2222 are applied to the inputs of the phase counters 2224, 2226, 2228 and 2230. These phase counters illustrated in FIGS. 90 and 91 are used to separately count the number of pulses developed at the outputs of the phase confirmation circuits 2220 and 2222 during a time interval of one-sixth of a bit at the various baud rates of 300 baud, 1200 baud and 4800 baud. At 9600 baud, the counters count the num-ber of pulses on the confirmation circuit 2220 and 2222 during a one-third bit interval. Each of the counters 2224, 2226, 2228 and 2230 includes six stages: 2254a-2254d, 2256x-2256d, 2258x-2258d, 2260x-2260d, 2262x-2262d and 2264x-2264d. The output signals ~, ~, ~g and ~3B are applied to inputs of exclusive OR gates 2266, 2268, 2270 and 2272. These carrier confirmation output signals are also applied to inverters 2274, 2276, 2278 and 2280. The outputs of the inverters 2274, 2276, 2278 and 2280 are applied to three input AND gates 2282, 2284, 2286 and 2288 along with output signals from the first two stages 2254 and 2256. The outputs of these AND gates 2282, 2284, ZZ86 and 2288 are applied to MUXas 2290, 2292, 2294 and Z~96. These AND gates are used to generate the 1Z00 baud signals at the 1KLJXes 2290, 2292, 2294 and 2296.
The third and fourth stages 2 2 58 and 2 2 60 of each o f the counters Z~~4, ZZ26, 2228 and 2230 are applied to AND
gates 2298, 2300, 2302 and 2304 along with the outputs of the AND gates ZZBZ, 2284, 2286 and 2288 to develop the 300 baud signal available at the I~iUXss 2290, 2292, 2294 and 2296. The output of the inverters 2274, 2276, 2278 and 2280 is applied directly to the t~JXes 2290, 2292, 2294 and 2296 !or the 4800 baud and 9600 baud signals.
As previously mentioned, the baud rate is selected by the bits ICAH ( 7 , 6 ] . The compluents of these signals are available at the outputs of inverters 2306 and 214 ~
2308. These complemented signals are applied to select input, S0, S1 of the MtJXes 2290, 2292, 2294 and 2296 to select the proper baud rate and generate signals ONEACRY, ZEROACRY, ONFBCRY and ZEROBCRY at Z outputs of the MUXes 2290, 2292, 2294 and 2296. These output signals are ap-plied to the filth stages 2262a-2262d of the phase coun-ters 2224, 2226, 2228 and 2230 by way of NAND gates 2310, 2312, 2314 and 2316 and exclusive OR gates 2318, 2320, 2322 and 2324. The outputs of the fifth and sixth stages 2262 and 2264 are applied to AND gates 2326, 2328, 2330 and 2332. These NAND gates 2326, 2328, 2330 and 2332 gen-erate a pulse every one-sixth of a bit at baud rates 300, 1200 and 4800 and at one-third bit intervals at 9600 baud.
The output of these HAND gates are applied to a HAND gate 2334 which, in turn, is applied to an input of a I4TJX 2336 along with a baseband signal available at the RX pin o!
the IC 10.
The ICAHS bit selects between ASR/FSR and base-band. The baseband carrier signal CARR, available at the output of the NOR gate 2174, is applied to clock inputs CP
of the lirst stage 2254 of each of the phase counters 2224, 2226, 2228 and 2230.
Each o! the counter stages 2254, 2256, 2258, 2260, 2262 and 2264 are reset by the phase reset signal PHRST by way o! inverters 2338, 2340, 2342, 2344 and 2346.
The outputs o! the inverters 2340, 2342, 2344 and 2346 are the signals ON?J~CLR, ZEROACLR, ONE8CLR and ZEROHCLR.
These signals are applied to reset inputs of stages live and six 2Z6? and 2264.
The output of the phase counters 2224, 2226, 2228 and ZZ30 is a signal DPH, which indicates correlation of the carrier signal, available at the output of the MUX
2336. This signal DPH is applied to a demodulator counter 2338 by way of a Clip-flop 2340. The strobe signal PHCK
i~ also applied to the flip-flop 2340. The strobe signal PHC1C is six times the bit rate except at 9600 baud where it is only three times the bit rate. The output of the Clip-Clop 2340 is applied to the demodulator counter 2338 ~'r and al~,lip-flop 2342 to generate a demodulator reset sig-nal ~EMODRST. The demodulator counter 2338 counts the number of outputs from the phase counters 2224, 2226, 2228 and 2230. The demodulator counter 2338 includes three stage flip-flops 2346, 2348 and 2350. The demodulated output signals DEMODAT from the demodulator counter 2338 are applied to an OR gate 2352 which, in turn, is applied to a message shift register and BCH computer discussed be-low.
In order to account for the various selectable baud rates, circuitry is provided which includes an exclu-sive OR gate 2354, a flip-flop 2356, an OR gate 2358 and a l4tJX 2360 is used. This circuitry is applied to the de-modulator counter 2338 by way of an exclusive OR gate 2362 along with a stage one output signal from the flip-flop 2346. The output of the M1JX 2360 provides a strobe signal which is either three or six times the bit rate. More specifically, a Q output of the flip-flop 2340 is applied to one input o! the exclusive OR gate 2354. A ~ output of the flip-flop 2356 is applied to the other input of the exclusive or gate 2354. The output of the exclusive OR
gate 2354 is applied to an input o! the flip-llop 2356.
This tlip-llop 2356 is clocked by a strobe signal PHCHAD
at six tiaea the bit rate. The output of the flip-flop 2356 along with the output o! the flip-flop 2340 are ap-pli~d to inputs of the OR gate 2358. The output of the OR
gate 2358 i~ applied to one input of the MUX 2360 to gen-erate a signal that is three times the bit rate when 9s00 baud is selected. The output of the flip-flop 2340 is also applied directly to another input o! the Mtt7c 2360 to gsnerate a signal that is six times the bit rate.
The bit rates are selected by the comaand bits IGH[7,6,5] as well as circuitry which includes a NJ~ND
gate 2364 and an inverter:2366. The NI~ND gate 2364 and the inverter 2366 decode the baud rate and modulation aethod comaand bite ICaFI[ 7, 6, 5 ] . The output of the NPrNND
gate 2364 is applied to a select input S o! the 1~TX 2360.

A bit framing counter 2344 is used to count 12 one-~xth bit intervals to provide a frame of reference to determine whether the incoming signal comprises two start bits; both logic one's. Should the phase counters 2224, 2226, 2228 and 2230 count to eight during the 2-bit inter-vsl, a valid start bit is assumed and a signal BRCK is generated. The bit framing counter 2344 is a four stage counter and includes the flip-flops 2368, 2370, 2372 and 2374 as wall as AND gates 2376, 2378 and 2380. This 10' counter 2344 is clocked by the strobe signal PHCKAD. The first, second, third and fourth stages outputs are applied to the AND gate 2376 to provide a divide by twelve signal DIV12. The first, second and third stage outputs are ap-plied to the AND gate 2378 to provide a divide by six DIVE
signal. The first and second stage outputs are applied to the AND gate 2380 to provide a divide by three DIV3 sig-nal. The DIV12, DIVE and DIV3 signals era applied to a lslJX 2382. Thus signals era selected by inputs applied to SO and S1 inputs on the 1KUX 2382.
A command bit decode signal from the HAND gate 2364 is applied to an SO input. The other input S1 is under the control of a HAND gate 2384. The HAND gate 2384 is a two input NAND gata.~ The NAND gate 2384 allows the bit freeing counter 2344 to be adjusted after a valid start bit has been detected. Hors particularly, and TRa'R signals are applied to inputs o! the HAND gate 2384. The signal at the output o! an inverter 2377, which is serially connected to a receive detect latch 2379. It during the first two bit interval, the de-modulator counter 2338 has counted to eight indicating that eight o! the twelve one sixth bit intervals cor-responding to the two start bits o! a received massage have bean received, the ~'ET latch 2379 is sat. when this latch 2379 is sat, the signal 1~~6ET is low for the remainder of the massage. Since the bit !taming counter 2344 is also used for transmitting massages, the other in-put to NAND gate 2384 is a 3'R~A. This signal is active low wham the ICC 29 is transmitting a massage.

217 54, 2113-I-1 The output of the HAND gate 2384 is then applied to t~ select input S1 of the MUX 2382 to select one of the four inputs of the MUX from the bit counter 2344. The output of the MUX 2382 is applied to a D input of the flip-flop 2386. This flip-flop 2386 is clocked by the strobe signal PHCKAD. The output of the flip-flop 2386 is applied to an input of the inverter 2388. The output of the inverter 2388 is a frame signal FRAME which produces a pulse after each bit is detected.
' Circuitry is also provided to reset the system each time the demodulator counter 2338 counts to eight during the 2 start interval. This circuitry includes the NAND gates 2397 and 2398 and AND gates 2400 and 2402. The output of the AND gate 2402 is a reset word signal RSTWORD
that is applied to a CDN input of the flip-flop 2340. The RSTWORD signal is active at the end of a message as will be discussed below. This RSTWORD signal is also applied to a CON input of the flip-flop 2342 used to generate the deaodulator counter reset signal DEMODRST which, in turn, resets the desodulator counter 2338. The DEMODRST signal is generated by circuitry which includes inverters 2381 and 2385, err OR gate 2383 and an AND gate 2387 is used to generate the ~, This signal is available at an output of the inverter 2381. An output of the 0R gate 2383 is tied to the input of the inverter 2381. The OR
gate 2383 is a three input OR gate. An IRESET signal is applied to one input, available at the output of high gain inverters 2385 and 2405, to allow the circuitry to be re-set on systea reset. An output fro' the AND gate 2400 is applied to another input o! the OR gate 2383. The FRAME
signal is JINDed with the strobe signal PHCICCD to produce a pulse at the start of each bit. Lastly, an output o! the AND gate 2387 is applied to the OR gate 2383. The AND
gste 2387 is a three input AND gate. l~ ~R~A signal, indi-eating that the transmitter is oft is applied to one in-put. The R~ signal, indicating that the start bits have been detected, is applied to another input. Thirdly, a ~ output troe the flip-flop 2342 is applied to an input.

~~fl~~

fop-flop 2342 is used to release the reset on the de-modu~ptor counter 2338 near the end of each one-sixth bit interval.
The AND gate 2402 is a three input AND gate. An signal is applied to one input. This signal is available at the output of an inverter 2404. The other inputs are from the HAND gates 2396 and 2398. The output of the NAND gate 2398 represents an end of message. More specifically, the ENDMSG signal is applied to an end of message latch 2404, which includes the NOR gates 2406 and 2408. The other input to the end of massage latch 2404 is a strobe signal PHCKDD. The other input to the HAND gate 2398 is a strobe signal PHCRD.
The HAND gate 2396 is a two input NAND gate.
one input is the receive detection signal RCVDET, which is active when a signal is being received by the ICC 29 as discussed above. Ths other signal is from the AND gate 2400. The AND gate 2400 is a two input AND gate. one in-put is the strobe signal PHCKCD. The other signal is the frame signal FRAME available at the output of the inverter 2388.
Circuitry which includes a NAND gate 2410 and an inverter 2412 is used to develop a bit clock signal BRCK.
This signal is used to clock a bit counter 2414. The HAND
gate 2410 it a three input HAND gate. One input is a strobe signal PHCRBD. A bit frame signal FR1~14E is applied to another,input. The output of the HAND gate 2384 is ap-pli~d to the third input. The NAND gate 2384 represents that a message is being received. The output of the NAND
gate 2410 is applied to the input of the inverter 2412.
The output of the inverter 2412 is the signal BRCR.
~r DRC1C signal is generated at the output of an AND gate 2416. This signal is used in conjunction with the ICC 29 trans~ittar as will be discussed below. The AND gate 2416 is a two input AND gate. One input is the strobe signal PHCRCD. The other input is the FR1~14E sig-nal.

Tha bit counter 2414 is illustrated in FIG. 93.
This fit counter 2414 is a six stage counter and includes the flip-flops 2430, 2432, 2434, 2436, 2438 and 2440. The bit rate clock signal BRCK is applied to the input of the first stage 2430. The output of the last stage of the bit counter 2414 is the end of message signal FxIDMSG.
The output of the outputs of the first five stages 2430, 2432, 2434, 2436 and 2438 are applied to a six input HAND gate 2442 along with a 8~'R signal, avail-able at the output of an inverter 2444. The output of the NAND gate 2442 is applied to a latch 2446, which includes the NAND gates 2448 and 2450. The other input to the latch 2446 is an output from an inverter 2452. The output of the latch 2446 is a signal 026, which is applied to a BCH computer to be discussed below. The 026 signal latches when the bit counter counts to 26.
Since this counter 2414 is used both for receiv-ing massages and transmitting messages to determine the end o! a word, after a word is counted, the counter 2414 as well as the latch 2446, is reset by circuitry which in-cludes the inverter 2452 and inverter 2454 and HAND gates 2456 and 2458. When messages are being received by the ICC 29, a receive detection signal RCVDET and 3'Ra'A, ap-plied to the inputs o! the HAND gate 2456, are active.
The output o! this NAND gate 2456 indicates that a message i~ being received by the ICC 29. During conditions when the ICC 29 is transmitting messages an INITX signal, ap-plied to another input o! the NAND gate 2458, is active.
The INITX signal indicates that a massage transmission has been initiated. Lastly, a reset signal Ice, available at the output of the inverter 2054, is applied to the third input. The output o! the NAND gate 2458 indicates that either a message is being received or that the ICC 29 is transmitting a massage. The output of the NAND gate 2458 is applied to an input of the inverter 2452. The output of the inverter 2452 i~ used to reset the bit counter 2414 as well as the latch 2446. More spe-citically, the output of the inverter 2452 is applied to CDN inputs of all six stages of the bit counter 2414. The outp~ of the inverter 2414 is also applied to one input of the 026 signal latch 2446. The output of the HAND gate 2458 is also used to reset the BCH computer which will be discussed later.
Circuitry which includes a flip-flop 2460 and a NOR gate 2463 is used to develop a transmitter signal identified as . This signal is used in conjunction with the transmitter control circuits discussed below.
More particularly, a ~ output of the sixth stage of the 2440 of the bit counter 2414 is applied to one input of the two input NOR gate 2462. The '~A signal is applied to the other input. The '~RaR signal is active low indi-cates that the transmitter is on. The output of the NOR
gate 2462 is applied to a D input of the flip-flop 2460.
The flip-flop 2460 is clocked by the frame signal FRAME.
The 3'RaFF signal is available at a QN output of the flip-flop 2460. This flip-flop is reset by an INITX signal, which is applied to a CDN input of the flip-flop 2460.
INCOM SHIFT REGISTER
32 stage serial shift register 2462 is used for INCOH messages in both receive and transmit opera-tions. This shift register 2462 includes the flip-flops 2464-2514, illustrated in FIGS. 102 and 103. Each stage of the shift register 2462 is adapted to receive two inputs which are selected by a LOAD signal used for loading messages into the shifted register 2462. More specifically, as discussed previously, a transmit data bus TDATa[Z6...0] which allows the microprocessor 30 to coa~unicate with the ICC 29, is applied to DA inputs of each stage of the shift register 2462. The demodulated output bits DF.MOO~rT for received messages are applied to DB inputs o! the first stage 2464 of the shift register Z46Z for received messages. Received messages are then sbilted through the shift register 2462. The receive data bus RDATJ~[Z6:..0] is applied to Q outputs of each o! the stages. The RD~T~I[Z6...0] bus allows demodulated incoming messages to be communicated to the microprocessor 30.

A ~ signal is applied to the select inputs SA
of e~h stage to allow the shift register 2462 to select between received messages and transmitted messages. The LOAD signal is available at the outputs of parallel con-s nected inverters 2516 and 2518. The output of the in-verter 2520 is connected to the input of the parallel con-nected inverters 2516 and 2518. The LOAD signal, dis-cussed in conjunction with FIG. 109 below, is applied to the input of the inverter 2520. The ~ signal is avail-able at the output of the parallel connected inverters 2516 and 2518.
A shift register clock signal SRCR, discussed in conjunction with FIG. 110, is applied to the clock inputs of each stage of the shift register 2462. The SRCK signal is available at the outputs of the parallel connected in-verters 2520 and 2522. The input to the parallel con-nected inverters 2520 and 2522 is an ~ signal.
Except for stage one 2464 and stags two 2466, which are the status bits for the reply massage, the bal ance of the stages 2468-2514 are set to zero on system re set. More particularly, a signal available at the output of the parallel connected inverters 2524 and 2526 is applied to CDN reset inputs of the third through thirty-second stage. The inputs to these inverters 2524 and 2526 are tied together and to the output of an in-verter Z5Z8. The input of the inverter 2528 is connected to the output o! an inverter 2 5 3 0 . The ~ s igna 1 i s applied to the input of the inverter 2530.
The status bite for the reply message are avail able on an internal reply statue bit bus RSH[26,25].
These status bits are available at the Q output: of the flip-flops 2030 and 2032 which tore a portion of the iciKo mea:age register. Delinitiona for the status bits are provided in Table Z1. These status bits RSB[Z6,25] are applied to either the set SDN or reset CDN inputs of the first and second stage flip-flops 2464 and 2466. More particularly, these bits RSH[26,25] are applied to inputs of dual input NAND gates 2532 and 2534 along with a STSLD

signal which indicates that a reply message is being for-mulat~d. The outputs of the NAND gates 2532 and 2534 are applied to the select inputs SDN of the stage one and stage two shift register flip-flops 2464 and 2466 to shift ones into these flip-flops when selected. These reply status bits RSH(26,25] are also applied to inputs of in-verters 2536 and 2538. The output o! these inverters are applied to two input HAND gates 2540 and 2542 along with the STSLD signal. The outputs of the NAND gates 2540 and 2542 are applied to the reset inputs CDN of the flip-flops 2464 and 2466 to set these flip-flops to zero when selected.
ICC CONTROL Lnc_=TC_~
The control logic for the ICC 29 is illustrated in block diagram form in FIG. 104. The control logic in cludes transmitter control logic circuitry 2516, illus trated in FIG. 105, a HCH computer 2518 illustrated in FIG. 106, addrus and command decoder logic 2520, illustrated fn FIGS. 107 and 108 and control and status logic 2522 illustrated in FIGS. 109 and 110.
Referring first to the transmitter control logic 2516, the ICC 29 transmitter output is a signal TXOUT.
This signal is applied to a tristata device 2524 (FIG. 85) whose output is tied to the external pin TX. The TXOUT
pin is available at the output o! a NOR gate 2526 (FIG.
105). Whsrr the ICC 29 is operated in the master mode, it can transact any time. when the ICC 29 is operated in the slave node, it can only transmit c! a reply is requested by the initisting controller. In the slave mode, the ICC
29 is under the control o! an AND gate 2528 (FiG. 110).
The AND gate 2528 is a two input J,ND gate. One input is an interlace enable signal EAxIPT. This signal is avail-able lroa the address/command decoder 2520, which will be discussed below. The other input is a slave mode signal available at the output o! an inverter 2530. The input to the inverter 2530 is the command bit ICCR(5). When this comsand bit is a logical zero, the ICC 29 will be in the slave mode. The output o! the AND gate 2528 is applied to the il~,~ut of an interface enable latch 2532, which in-clud~ the HAND gates 2534 and 2536. The output of the latch 2532 is the control bit ICSR[6] which indicates that the communication controller 29 interface is enabled. The latch 2532 is also used to generate an ENAH signal, avail-able at an output of a flip-flop 2533. More particularly, the output of the latch 2532 is applied to a D input of the flip-flop 2533. The signal FNAB is available at a Q
output of this flip-flop. The strobe signal PHCxAD, available at the output of the inverter 2590, is applied to a clock input CP of this flip-flop.
The interface enable latch 2532 may be disabled by a two input AND gate 2538. One input to the AND gate 2538 is a reset signal ~. The other input is a dis-able interface signal DISINT, available at the output of the address/command decoder 2520. The DISINT signal dis-ables the interface enable latch 2532 when a reply is not necessary.
In the master mode, the ICC 29 can transmit any time. Initiation of transmissions is under the control of a three input NJ~ND gate 2540 (FIG. 109) . The command bit ICCR[0], which indicates a transmission command, is ap-plied to one input. The control bite ICSR[6,5] era ap-plied to the other input:. The control bit ICSR[6] indi-catea that the interface is enabled.. The control bit ICSR[5] i~ active low when the trane'itter is inactive to allow the ausage start bits (e.g., two 1's) to be gensr-ated. More particularly, the output of the NAND gate 2540 is applied to a start bit latch 2542, which includes the N1~ND gates ZS44 and 2546. The output of the start bit latch 2542 is applied to a pair o! flip-flops 2548 and 2550 to generate the start bits. A ~ output of the flip-flop 2550 is applied to an input o! a massage transmit AND
gate 2552. The other input to the AND gate 2552 is from a F1~ST STATUS J~rIdD gate 2572, which will be discussed below.
The output of the AND gate 2552 i~ a transact message sig-nal 3~. The signal is applied to a SDN input of the flip-flop 2574 (FIG. 105) to set this flip-flop to one durinc~ha start bit interval. A D input of the flip-flop 2574 His grounded. The flip-flop 2574 is strobed by a transmitter control strobe signal DRCK, available at the output of the AND gate 2416 (FIG. 92).
Atter the two start bits are generated, an AND
gate 2554 resets the flip-flops 2548 and 2550 and the start bit latch 2542. The AND gate 2554 is a two input AND gate. One input is from the output of the flip-flop 2550. A signal is applied to the other input to allow the flip-!lops to be set to zero on system reset.
The ~ signal is available at the output of an in-verter 2556. The flip-flops 2548 and 2550 art strobed by PHCR signal which strobes the flip-!lops every bit.
In both the master mode and the slave mode, a reply latch 2558, which includes the NAND gates 2560 and 2562, is set. The reply latch 2558 is under the control o! the two input NOR gate 2526. One input to the NOR gate 2526 is the ~Y signal, which indicates that a reply is required. The command bit ISCR(5] is applied to the other input. The output o! the latch 2558 is applied to one in-put o! a two input NAND gate 2564. The other input to the NAND gate 2564 is a last status latch 2566 which includes the NAND gates 2568 and 2570. The output o! the NAND gate 2564 is applied to one input o! a two input HAND gate 2572 along with a strobe signal PHCRDD. The output o! the HAND
gate 2572 is applied to the NAND gate 2552 along with the output o! the !lip-flop 2550 to generate a transmit mes-sage signal .
Signals IAITX and ~X~'N are also generated by the 3o transmitter control logic circuitry 2516. More particu larly, a Q output of the flip-flop 2574 is applied to the inverter 2576. The output o! the inverter 2576 is an ini tiate transmit signal IAZTn. This signal ZRITR is inter locked with the 8CH computer 2518 as will be discussed be 3s low.
A ~ output o! the flip-!lop 2574 is applied to a transmitter latch 2578 which includes the NAND gates 2580 and 2582 to generate a ~R~R signal. The transmitter latch 2578 ~ reset by the fiR3~' signal. This signal is avail-able,~t a ~ output of the flip-flop 2460 (FIG. 93) and in-dicates that the message counter has counted 32 bits. The output of the transmitter latch 2578 is applied to an in-s put of an inverter 2584. The output of the inverter 2584 is the '~R~A signal, active low, indicates that the trans-mitter is on. This '~Ra'A is used to develop the control bit ICSR[5~, which indicates that the transmitter is ac-tive. More particularly, the fiRb'A signal is applied to an inverter 2586 (FIG. 110). The output of the inverter 2586 is applied to a D input of a flip-flop 25'88. A Q output of the flip-flop 2588 is the control bit ICSR[5]. Flip tlop 2588 is strobed by a PHCKAD signal, available at the output of an inverter 2590. The input to the inverter 2590 is the strobe signal A ~ output of the flip-flop 2588 is also used to generate a BUSY signal as well as the ~5 signal, which is active low. More specifically, a ~ output of the flip-tlop 2588 is applied to one input of a two input NAND gate 2591. The other input is the RCVDET signal which indi-cates that a message is being received. The output of the HAND gate 2591 is the BUSY signal.
TR~n1S14T_TTER CONTROh Various modulation methods are available: ASK, FSR and baseband. The start bits are shitted into a flip flop 2592 by tying a ~ output of the Clip-flop 2574 to a SDN input o! the flip-flop 2592 as discussed above. The signal MODIN, available Eros the BCH computer 2518, which will be discussed below, is applied to a D input of the Clip-tlop 2592. The Clip-flop 2592 is strobad by the BRCK
signal available at the output of the inverter 2412 (FIG.
92j. The baseband signal, available at a ~ output of the Clip-Clop 2592, is applied to an input of a MUX 2594. It is also applied to an FSR modulator MUX 2596. More par-ticularly, 115.2 kHz (FC) and 92.16 kIiZ (FCAj signals are applied to the inputs of the MUX 2596. The ~ output of the Clip-Clop 2592 is applied to the select S input of the MUX 2596 to shitt between 115.2 kIiz and 92.16 kHz for FSK

modul~ion. The output of the MUX 2596 is applied to an inverter 2598 which, in turn, is applied as an FSK input of the MUX 2594. A Q output of the flip-flop 2592 is ap-plied to one input of a two input HAND gate 2600, used for ASK modulation along with the carrier signal FC. The out-put of the ASK modulator 2600 is applied to another input of the MUX 2594. The control bits ICAFI[5,4) are applied to the select inputs S1 and SO of the MUX 2594 to select between ASR, FSR or baseband. The output of the MUX 2594 10' is applied to the OR gate 2526 along with an intermessage spacing generator latch 2602 output. The intermessage spacing latch 2602 includes the HAND gates 2604 and 2606 and provides zeros between massages. The output of the intermessage spacing latch 2602 is applied to another in-put of the OR gate 2526. The output of the 0R gate 2526 is the TXOUT signal.
The intermessage spacing latch 2602 is under the control of a two input NAND gate 2604. The HAND gate 2604 is a two input HAND gate. One input is from the counter signal BRCR. The other input is the Q output of the flip-flop 2574. Thus, whenever a transmission is initiated, the HRCR signal times the transmission and provides zeros at the end of a aessage.
COL AND STAT1 nrI
Various control and status signals are generated by the control and status logic circuitry 2522. The con-trol bits ICSR[7...5] have previously been discussed.
Status bits ICSR[4...Oj are derived from the circuitry il-lustrated in FIG. 109.
The status bit ICSR[0] indicates a receiver (RX) overrun. This status bit is developed by the circuitry which includes a flip-flop 2620, NAND gates 2622 and 2624 and inverters 2626, 2628, 2630 and 2632. A receive signal available from the instruction decoder 2520, is ap plied to a clock input CP o! the flip-flop 2620 by way of inverters 2626 and 2628. A ~ output o! the flip-flop 2620 is NANDed with the statue bit ICSR[ 2 J , by way of the HAND
gate 2622 and applied to a D input o! the flip-flop 2620.

The ~t ICSR[O) is set if the message register has not been,~eleased (ICSR[2] = 1) when a new message is ready to be loaded into the message register. This status bit is cleared by system reset or writing to ICCR[1] = 1. More specifically, reset is under the control of the AND gate 2624. The NAND gate 2624 is a two input NAND gate. One input is the bit ICCR[1] which is applied by way of the inverter 2630. The other input is the ~ signal, available at the output of an inverter 2636. The output 0! the AND gate 2624 is applied to a reset input CDN of the flip-flop 2620 by way of the inverter 2632.
The ~D signal is also used to generate a mes-sage register receive buffer strobe signal SRTOI~t. This signal is applied to an AND gate 2638 along with a ~ out-put of the flip-flop 2634, indicating that the receive operation is not complete.
The bit ICSR[1] indicates a BCIi error which is set after a message containing an error is received. This bit is generated by circuitry which includes a flip-flop 2640, an OR gate 2642 and an inverter 2644. A B~ sig-nal, indicating an error, is available at the output of the inverter 2644. This signal is applied to one input of the OR gate 2642 along with a Q output of the flip-flop 2640. The flip-Clop 2640 is clocked by the RZV signal.
The flip-flop 2640 is reset in the acme manner as the Clip-flop 2620.
The bit ICSR[2] when set, indicates that a re-ceived message has been loaded into receive message butter. This bit is available at the output of the flip-flop 2634. A D input of this flip-llop is grounded. The flip-Clop 2634 is clocked by the l~ signal and is reset in the same manner as the flip-flops 2620 and 2640.
The bit ICSR[3] when set indicates the comple tion of a muaage transmission. Circuitry for generating this bit includes the flip-flop 2644, a transmitter latch 2646, which includes NAND gates 2648 and 2650 and an AND
gate 2652. The transmitter latch 2646 output is applied to a D input of the flip-flop 2644. Thi: latch 2646 is latch while the transmitter is active. More spe-cifir~lly, a TXOFF signal, available at an output of an inverter 2654 is applied to a clock input of the flip-flop 2644. The TXOFF signal is active high and indicates that the transmitter is off. A ~ output of the flip-flop 2660, which is low after the message start bits are generated.
Thus, the bit will be set at the end of message after the TXOFF signal becomes high, indicating that the transmitter is off.
The bit ICSR[3] is cleared on reset by the AND
gate 2652. The ~ signal is applied to one input.
The bit can also be reset by writing to ICCR[ 2 ] = 1. The ICCR[2) signal is available at an output o! an inverter 2654. An output from the NAND gate 2540 is also applied to another input o! the AND gate 2652 to reset the flip-flop 2644 when a new transmission is initiated.
The bit ICSR[4] is set alter completion of a fast status message transmission. Circuitry for generat-ing this bit includes a Clip-flop 2656, a latch 2658 which includes the NAND gates 2660 and 2662 and an AND gate 2664. The latch 2658 is set by the fast status enable latch 2566. The fast status enable latch 2566 is under the control of a three input OR gate 2665. The command bit ICCR[5] indicating master mode is applied to one in-put. The bit ICCR[6], which indicates a fast status en-able, available at an output of an inverter 2567 is ap-plied to another input. Lastly, a signal F~ is applied to the third input. The FAT signal indicates that a fast status reply message has been requested in an incoming message. The last status enable latch is strobed by the strobe signal ~, available at an output o! an inverter 2669.
An output o! the latch 2658 is applied to a D
input o! the Clip-llop 2656. The TXOFF signal is applied to the clock input CP to set this bit when the transmitter is o!! alter the fast status enable latch 2566 is set. A
~ output o! the flip-flop 2656 is applied to an input of the 1~ ch 2658 to reset it after the status bit ICSR[4] is sat . ~
Reset of this bit is under the control of the two input AND gate 2664. The ~S signal is applied to one input. The bit may also be reset by writing to ICCR[3]. The bit ICCR[3] is available at an output of an inverter 2666.
An interrupt signal INT is generated at an out put o! an AND gate 2670. When sat, this bit will generate an interrupt on receive and transmit operations. More specifically, the bit ICCR[7], which indicates interrupt enable is applied to one input of the AND gate 2670. An output of a NAND gate 2672 is applied to another input.
The NAND gate 2672 is a three input NAND gate ~ outputs from the flip-flops 2634, 2644 and 2656 are applied to the inputs to generate interrupts during receive and transmit operations.
The LOAD signal, used to load messages into the shift register 2462 is generated by a NAND gate 2574. The HAND gate 2574 is a two input HAND gate. One input is from the fast status enable latch 2566. The other input is from the flip-flop 2550, which indicates the message start bits have been generated.
The signal STSLD, is used to enable the AND
gates 2532 and 2534 (FIG. 102). This signal is a status bit load signal and is used to allow the status bits to be loaded into the first two stages 2464, 2466 of the shift register 2462. This signal STSLD is available at an out put o! an inverter 2572. The output of the NOR gate 2526, previously discussed is applied to the input of the in-verter 2672.
Lastly, the ~ signal, used to strobe the shift register 2462, is generated by circuitry which in-cludes an aIND gate 2674, NAND gates 2676 and 2678 and an 3 5 inverter 2 680 ( FIG . 110 ) . Outputs o! the NJ~ND gates 2 67 6 and 2678 are applied to inputs of the ANO gate 2674. The signal b'i'd, available at an output o! the inverter 2680, indicates that the bit counter 2414 has not yet counted 2s bits,~s applied to one input of the NAND gate 2676 along with the bit rate clock signal BRCK and the ZR'ITX signal .
The strobe signal PHCKCD and the LOAD signal are applied to inputs o! the NAND gate 2678.
BCH COMPL1't' .R
The BCH computer 2518 is illustrated in FIG. 106 and computes a five bit error code based upon the first 27 message bits. The HCH computer 2518 is implemented as a five stage shift register 2674, which includes flip-flops 2676, 2678, 2680, 2682 and 2684 and an exclusive OR gate 2686.
In the message receive mode, as the demodulated bits DEMODAT are loaded into the shift register 2462, they are simultaneously applied to a receive/transmit MUX 2686.
The input signals to the MUX 2686 are selected by the fi~3A
signal, applied to a select input S of the MUX 2686.
Before the 26th message bit is received, the de-modulated messags bits DEMODAT are applied to an exclusive OR gate 2688 by circuitry, which includes an AND gate 2690, an inverter 2692 and a NOR gate 2694. More spe cifically, an 026 signal is applied to the inverter 2692, whose output is applied to one input of the AND gate 2690.
The deaodulated message bits DElIODAT are applied to the other input of the AND gate 2690. The output of the AND
gate 2690 is applied to an input of the exclusive OR gate 2688 along with an output of the NOR gate 2694 to allow the first 26 »ssage bits to be shitted into the shift register 2674 to generate the error code. Attar the first 26 aessage bits are received, the error code stored in the shift register is shitted out under the control of the NoR
gate 2694. The NOR gate 2694 is a two input NOR gate.
The 026 signal is applied to one input. A ~ signal from the last stage 2684 of the shift register 2674 is applied to the other input.
The error code is compared with the error code in the received message by way o! an exclusive OR gate 2696. The output of the exclusive OR gate 2696 is applied to a two input NAND gate 2698. The output of the HAND

231 ~~.~~f gate X98 is applied to a DA input of a flip-flop 2700 which selects the DA input after the 26 message bits are received to generate a BCHOK signal if the error codes match. The BCHOK signal is fed back as an input to the AND gate 2698. While the error code is being computed, the previous BCKOK signal is latched by tying a Q output o! the flip-flop 2700 to a DB input.
The BCH shift register 2674 is clocked by a BCH
clock signal HCHCLK, generated by circuitry which includes a NAND gate 2702 and an inverter 2704. More particularly, the bit rate clock signal BRCK along with the 3~33~ are applied to the inputs of the NAND gate 2702. The output of the HAND gate 2702 is applied to an input to the in verter 2704. The HCHCLK signal is available at the output 0! the inverter 2704.
The BCH shift register 2674 is cleared by a signal, available at an output o! an inverter 2706.
The input to this inverter 2706 is a BCHCLR signal, which is available at the output o! the NAND gate 2458 (FIG.
93), which indicates that a massage has bean received.
Clearing o! the flip-flop 2700 is under the control of a two input AND gate 2708. One input to the AND gate 2708 is the 3'~8R signal indicating that the transmitter is off.
The other input is the BCHCLR signal.
It is also necessary to compute a 8CH error code !or reply ~esaages. Thus, the TXD output o! the message shift register 2462 is applied to the MUX 2686. This in-put ie selected by the TRb'P' signal during transmission of a aeasage. The message bits are shifted into the shift register 2674 to generate a HCH error code in the same wanner as before. In the meantime, the message bits are also applied to a MLTX 2708 to develop a MODIN signal which i~ applied to the transmitter (FIG. 105). Alter 26 message bits, the signal 026 selects another input o! the MUX 2708 to allow the error code to be added at the end of the sessage.

232 ~' ~r ADDRESS COMrLAtdD DECODER
The address command decoder 2520 is illustrated in FIG 107. The INCOM address, stored in the ICAFi[ 3 . . . o register and the ICAL[7...0), is compared with the address received on the receive data bus RDATA(22...11] by way of exclusive OR gates 2710, 2712, 2714, 2716, 2718, 2720, 2722, 2724, 27265, 2728, 2730 and 2732; NOR gates 2734, 2736 and 2738 and HAND gates 2740, 2472 and 2744 to gener ate an ADDROR signal it the received address matches the address in the ICAH and ICAL registers.
The address bit comparisons of the bits B22-B19 are applied to the NOR gate 2734. The address bit compar isons of the bits B18-B15 are applied to the NOR gate 2736. Lastly, the address bit comparisons of the bits H14-B11 are applied to the NOR gate 2738.
The output: of the NOR gates 2734 and 2736 for bit comparisons are applied to the NAND gate 2240 along with a BLOCK signal, available at an output of an inverter 2746. when block instructions are employed, the lower tour order bits Hii-814 are ignored. The output of the NAND gate 2740 is applied to one input of the NAND gate 2744.
when no BLOCK instructions are employed, the NaND gate 2744 is under the control of the NAHD gate 2742.
The NAND gate 2742 receives inputs troy the NOR gates 2,734, 2736 and 2738 and provides address comparisons for all o! the address bits H22-B11. The HAND gate 2744 is also controlled by the UNIV address signal. This signal allows messages to be received by all devices on the net work.
The comaand field B10-B7 is used to define the cosmand in enable intertace control messages. These bits Bi0-H7 are decoded by a tour input NOR gate 2748. A sub-comaand field I delines the subcommand in enable interface control sessages. These bits are decoded by a NOR gate 2750. The outputs of the NOR gates 2748 and 2750 are ap-plied to an AND gate 2762 whose output is a command signal ~7~ID0 .

A receive message strobe signal RCVMSGSTB, used to g~erata the receive message signal RCV, is developed by the circuitry which includes a flip-flop 2754, on AND
gate 2756, a NOR gate 2758 and an inverter 2760. A
and end of message signal ENDMSG are ANDed by the AND gate 276 and applied to a D input of the flip-flop to indicate that a massage has bean received and the transmitter is not on. This signal is shifted into the flip-flop on a 81~K signal available at an output of the inverter 2760.
The flip-flop 2754 is reset by the NOR gate 2758. The strobe signal PHCKCD is applied to one input of the NOR gate 2758 while the RESET signal is applied to the other input to reset the RCVMSGSTB signal on system reset and by the strobe signal PHCKCD.
INSTRUCTION D'- ODER
Instruction decoder circuitry 2760 is illus-trated in FIG. 108. The instruction field consists of the message bits H6-H3. Message bit 82 defines the meaning of the message bits. These bits are applied to decoder cir-cuitry which includes inverters 2762-2780; NAND gates 2782-28101 1,ND gates 2812, 2814 and 2816 and exclusive OR
gates 2818 and 2820.
The instruction field B6-H2 definitions are de fined in Table 20 along with the control bit 82. These bits are applied to the inverters 2762, 2766, 2770, 2774 and 2778 which, in turn, are connected to serially coupled inverters 2764, 2768, 2772 and 2776, respectively. The output signals frog the inverters 2762-2780 are connected to the various N1,ND gates, aND gates and exclusive 0R
gates as shown to generate the signals ~, FA3T, SLY, B'~ and UHlO available at outputs of the J,ND gate 2812 and the NAND gates 2782, 2783, 2786, 2788, 2804 and 2810, respectively. These signals are all active low.
Hore particularly, the OHiO addressin t 9 YPe is available at an output of the four input NaND gate 2810.
8lts 86, Hs and HZ are applied to the NAND gate 2810 along with an output of the exclusive OR gate 2820. Bits 83 and ~~fl~~~~

H4 ar~~compared by the exclusive 0R gate 2820. B2 is a cont~l bit and will both be a logical 1 for instruction decoding. Hits B5 and H6 will both be logical l~s for the instructions $C, $D, $E and $F (Table 20). The output of the exclusive OR gate 2820 will produce a logical 1 for instructions $1, $2, S5, $6, $9, $A, $D and $E. Thus, the output of the NANO gate 2810 will be low for instructions S6, $D and $E.
The ~R addressing type instruction is avail able at the output of the NAND gate 2804. Hits B5, ~4 and H2 are applied to the inputs o! the NAND Bata 2804. Since thaw bits must be a logical 1 in order to produce an ac tive low signal at the output of the HAND gate 2804, this portion of the circuitry will decode instructions $4, $5, $C and SD as long as the output of the HAND gate 2808 is a logical 1. Hit H3 is applied to one input of the NAND
gate 2808. An output of the HAND gate 2806 is applied to the other input. Hits B'~ and H3 are applied to the inputs of the NAND gate 2806 to produce a logical 1 at the output of the HAND gate 2806 to produce a logical 1 at the output of the NAND gate 2808 for the instructions $4, $5, SH and $C to generate the B'I~R signal.
The REPLY, ~3PT, , ~ and l~P signals are all enabled by a three input AND gate 2814. The RCVKSGSTB, HC~iOIC and control bit H2 signals are ANDed to generate a permissive signal, identified as DECODE, indi-cating that s correct massage has been received. The out-put o! the AND gate 2814 is applied to inputs of the NAND
gates 2782, 2784, 2786, 2788 and 2792.
3o The ~' signal is available at an output of the NAND gate 2788. Also applied to the HAND gate 2788 are the control bit ~5, available at an output of an in-verter 2828, which indicates whether the ICC 29 is in the master mode, the ADDROR signal, bit H6 and the output of the HAND gate 2802. It the correct address has been de-coded and the ICC 29 is in the slave mode, an active low SLY signal will be generated when bit H6 ~ 1, which oc-curs for instructions $8, $9, $A, $8, $C, $D, $E and $F, and t~ output of the HAND gate 2802 = 1. The inputs to N~ agate 2802 are outputs from the NAND gates 2709 and 2800. The circuitry which includes the HAND gates 2798, 2800 and 2802 and the exclusive OR gate 2818 will produce a 1 at the input of the HAND gate 2788 for the instruc tions $8, $9, S7~ and $F.
The EA~iRT signal is available at the output of the NJ,ND gate 2786. In addition to the DECODE signal, An ADDROR and bits B3, B4, $3 and ~ are applied to its in puts to generate the signal for instruction $3.
The signal is available at an output of the NaNO gate 2784 and decodes the instructions $2. One input to the NAND gate 2784 is the DECODE signal. The output of the NAND gate 2794 is applied to the other in-put. The N11ND gate 2794 is a two input N1,ND gate. The 1~DDROR signal is applied to one input. The output of the NAND gate 2796 is applied to the other input. The CC'S, 7sDDROR signals as wall as bits B5, 84 and B3 are applied to the inputs o! the Nl~rND gate 2796.
The F~3'!~ signal is available at an output of the Nl,ND gate 2782. It the control bit H2 is set, an instruc-tion field of $3 is decoded and a command field CMDO of o exists, the FAT signal is generated. The C1KD0 and DECODE
signals as well as bits B2, B3, ~5 and 83 are applied to the input of the NaND gate 2782 to generate the F~TT sig-nal.
The 1~ signal is available at an output of the 1~ND gate 2812. The outputs of the N1~1ND gates 2790 and Z79Z are applied to the inputs. The NJ~ND gate 2790 is a two input NaND gate. The signals RCVlsSGTH and ENJ~rB are applied to its inputs to generate this active low signal anytime a message with the interface enable instruction.
During other conditions, the signal ~V is under the con-trol of the two input N7~ND gate 2792. In order to gener-ate an active low ~ under these conditions, the output of the input' of the NaND gate 2782 to generate the FA~fi signal.

The 1~ signal is available at an output of the AND date 2812. The outputs of the NAND gates 2790 and 2792 are applied to the inputs. The NAND gate 2790 is a two input NAND gate. The signals RCVMSGTH and ENAB are applied to its inputs to generate this active low signal anytime a message with the interface enable instruction.
During other conditions, the signal 1~ is under the con-trol of the two input NAND gate 2792. In order to gener-ate an active low 1~V under these conditions, the output of the NAND gate 2792 must be low which requires its in-puts to both be high. The inputs to the NAND gate 2792 era the outputs from the AND gates 2814 and 2816. The output of the AND gate 2814 is the DECODE signal which in-dicates that a correct massage has been received and the control bit H2 is set. The AND gate 2816 is a two input AND gate. The ENAH and ADDROK signals era applied to its inputs. The AND gate 2792 produces a high output when the ICC 29 has been addressed and a massage has been received with an enable interface instruction.
2D L DESCRIPTION OF THE IC ld In the illustrated embodiment, the IC l0 is housed in an 80 pin quad plastic flat package (QPFp), gull wing, surface mount package. The IC 10 is a hybrid device fabricated utilizing CMOS technology and implemented in a way to ovarcoaa the shortcomings of utilizing CMOS for analog functions. Various ratings, operating conditions arid do characteristics are provided in Appendix A.
A detailed pin assignment for the IC 10 is shown in FIG. 113. So'a pins have a dual function. For axaapla, a pin gay have one function in one configuration and a different function in another configuration as will be discussed in more detail below.
The following is a brut description of the sig nal definitions for each of the pins illustrated in FIG.
113. A pin summary is provided in Table 22.
PA7...PAO: Port A - These eight bidirectional port pins can be individually programmed to be inputs or outputs by the software.

PB7...PBO: Port B - These eight bidirectional Port~pins have multiple functions depending on the operat-ing mode of the IC 10. In the single-chip mode, these port pins can be individually programmed as inputs or out-s puts by the software. In the expanded, emulation, or test modes, these eight port pins contain the high-order ad-dress bus.
PC7...PCO: Port C - These eight bidirectional port pins can be individually programmed to be inputs or ~ outputs by the software. The low-order four pins can also be configured to be the logical OR of the outputs of the four comparators.
PD7...PDO: Port D - These eight bidirectional port pins have multiple functions depending on the operat ing mode of the IC 10. In the single-chip mode, these port pins can be individually programmed as inputs or out-puts by the software. Zn the expanded, emulation, or test modes, these eight bidirectional port pins form a multi-plexed data and address bus. when PH2 is asserted, these pins are outputs and contain the least-significant 8-bits of the address. When PH2 is negated, these pins are bi-directional and contain read or write data.
EXPN - This low-true signal enables the expanded mode of operation. The single-chip mode is enabled by connecting EXpZ1 to VDD. This input is sampled when RESN
changes frog an electrical low la e1 to an electrical high lwel. The operating mode of the IC 10 is determined when the device leaves the reset state. Table 1 defines the pin input levels for various operating modes.
PHZ - The function of this output pin depends on the operating mode of the IC 10. In single-chip and self-tast modes it will remain low. In all others, it is the processor's phase 2 clock. Phase 2 is the oscillator out-put divided by two, and changes on the falling edge of OSCZ. Table 2 defines the output of PH2 for various oper-ating modes.
REN - The function of this output pin depends on the operating mode of the IC 10.

~~~~~~8 s In single-chip and salt-test modes it is used as a diagnostic pin. It will remain high except during internal read operations between $4000-S~FFF.
In expanded mode, it is used as the write strobe for external memory devices mapped in the address range o! $4000 to $7FFF. When low, the memory device can strobe data from the Port D
pins.
~ In emulation and test modes, it becomes the processor's internal E clock signal. E clock is PH2 delayed by 90~.
Table 2 defines the output o! REN for various operating modes.
WEN - The function o! this output pin depends on the operating mode o! the IC 10.
In single-chip and sell-test modes it is used a~ a diagnostic pin. It will remain high except during internal read operations between S40oo $7FFF.
In expanded mode, it is used as the write strobe !or external memory devicaa mapped in the address range of $4000 to $7FFF. When low, the meaory device can strobe data from the Port D
piny.
In emulation and test modes, it becomes the processor's internal E clock signal. E clock is PHZ delayed by 90~.
Table 2 delinea the output o! WEN !or various operating 3 0 mode .
P9LN - The function o! this output pin depends on the operating code o! the IC 10.
In single-chip and sell-test modes it is used a~ a diagnostic pin. It will remain high except during internal read operations between $8000 SEEFF.
In expanded mode, it is used as the read strobe for external read-only-memory devices 239 ~ (~'~
mapped in the address range of $8000 to $FFFF.
When low, the memory device should place read data on the Port D pins.
s In emulation and test modes, it becomes the processor s internal LIR signal . This pin then indicates when the processor is reading an in-struction from the external data bus. A high indicates that the instruction register is being loaded.
Table 2 defines the output of PSEN for various operating modes.
ALE - Tha function of this output pin depends on the operating mode of the IC 10.
s In single-chip and self-test modes it will remain low.
In all other modes, it is used to latch the least-significant a bits of the address present on Port A.
Table 2 defines the output of ALE for various operating modes.
TX - This digital output is the transmitter out-put from the ICC subsystem.

RX - This digital Schmitt input is the receiver input to the ICC subsystaa.

HUSYN - This low-true digital output is the busy output troy the ICC subsystem.

SC1C - This bidirectional pin is the serial clock for the SPI subsystem.

NISO - This bidiractional pin is ~mastar in, the serial outs for the SpI subsystem.

MOSI - This bidirectional pin is the master out, serial ins for the SPI subsystu.

SSN - This low-true input pin is the slave selects input for the SPI subsystem.

PWM - This digital output is the pulse-width aodulated output from the PwI~I subsystem.

TCI~ - This digital output is the timer primary output compare.

TCAP - This digital input is the timer input capt~e signal.
IRQN - This low-true digital input is the asyn chronous external input to the microcontroller. A mask programmable option permits selection of two triggering methods: 1) negative edge-sensitive triggering only, or 2) both negative edge-sensitive and low level-sensitive triggering. In the latter case, either type of input to the IRQN pin will produce an interrupt. The interrupt re-quest must be present at laast 125 ns in edge-triggered mode.
It the level-sensitive mask option is selected, the IRAN pin requires an external resistor to Vpp for 'wire-OR' operation.
The IRQN pin also puts the IC 10 in a test mode when placed at +g V during reset. This mode is for test only and should not be used during normal operation.
RESN - This low-true input provides an external method of initializing the IC 10. i~lhen using the external reset, RESN aunt stay low for a minimum of 1.5 processor phase 2 cycle. RESN is received by a Schmitt receiver.
HSENSE - This analog input is the non-inverting input to the H+ comparator.
BDRIVE - This analog output i~ the output of the H+ coaparator.
1,p08, ANEG - These analog inputs are the invert-ing and non-inverting inputs of the comparator.
J~OtJT - This analog output is the comparator out put pin. In many applications, this comparator is used as the input receiver for the ICC subsystem and is connected to RX.
1~TX3...1KUX0 - These tour analog input pins are one halt o! the J~/D subsystem inputs. They can be indi-vidually programmed to operate in either the voltage or current mode. In the voltage mode, they are high impedance inputs.
In the current mode, an active current source maintains a virtual ground level for currents ~ of the 2~~~~~~3 davic~pin. When unselected in the current mode, each pin is c~nectad to digital ground.
MUX7...MUX4 - These four analog input pins are the other half of the A/D subsystem inputs. They can operate only in the voltage input mode. They are always high impedance inputs.
1~DC0 - This analog output is used by the A/D sub-system in the currant mode of operation. An external re-sistor or capacitor between this pin and analog ground converts the mirrored and ratioad current from the selected input into a voltage for A/D conversion. If an external capacitor is usad,,the internal amplifier is con-figured as an integrator and currant autoranging must be disabled.
CP3...CPO - Thasa four high-impedance analog in-puts era the inverting inputs of four comparatora.
Those pins are also used during testing to select various test modes.
VADJ - This analog input is used to adjust the analog reference voltage: VREF.
VREF - This analog output is the internal +2.5 v reference. It is the output of the reference buffer am-plitier and must ba connected to the external reference trig resistor network.
AVDD - This pin is the +5 V analog supply volt-age. An external resistor is used to create a current source for the shunt-regulated power supply. AVDD will be regulated to approximately 2 ~ vREF.
AVSS - This pin is the analog ground reference.
OSCl - This is the input of the oscillator circuit.
OSZ - this pin is the output of the crystal os-cillator circuit. It is the inversion of the osl input.
VDD - these pins era the digital +5 volt DC sup-ply.
v88 - these pins era the digital negative sup-ply. They should be connected externally to the AVSS pin.

~ SHUNT This output pin is high when the -power supp~Y is shuntingcurrent from AVDD.

NC - connected.
Not TABLE ZZ
SIGNAL DEFINIT ONS

SIGNAL ~ DIRECTION TYPE

BS~ 1 Output CMOS

PD7...pDO 2...9 CMOS
Hi-directional S~ 10 Bi-directional Schmitt MISO 11 Hi-directional CMOS

HOSI 12 Hi-directional CMOS

SSN 13 Input CMOS

P82 14 Output CMOS

PH7...PBO 15... 22 Bi-directional CMOS

PA7...PAO 23... 30 8i-directional CMOS

31 Supply Digital + Supply vSS 32 Supply Digital + Supply PC7...PCO 33... 40 Bi-Directional CMOS

CPO...CP3 41... 44 Input Analog BDRIVE 45 Output Analog BSENSE 46 Input Analog APOS 47 Input Analog 48 Input Analog AOUT 49 Output Analog AVDD 50 Supply Analog + Supply AVSS 51 Supply Analog + Supply l4tJX7...1nJX052... 59 Input Analog 1"Df0 60 Output Analog NC (not connected) 61 63 output Analog Reference 63 Input Anal RDI 64 Output ~Og 63 Output ~Og 66 Output mpg Output ~Og T~ 68 Output ~Og SHUNT 69 Output ~Og VSS 70 Supply Digital Supply 71 Supply Digital + Supply TCAP 7Z Input CM

OS
PSEN 73 Output CMOS

OSCZ 74 output Analog OSC1 75 Input Analog RESN 76 Input Schaitt IRAN 77 Input qKOg ~~.'~~~~8 APPENDIX A
The following specifications apply to the EEPROM 40 memory under the conditions that VDD - 5.0 Vdc ~10~ and the ambient temperature TA is between -40°C and +85°C.
The specifications are indicated in Table A1.
EEPB_4~ SPECTFT~'ATTON~
X UNIT
EEPROH Erase Time ERA 10 EEPROH Program PROG 10 ms Time 2 I~iz osc.
EEPROH Program PROG 20 ms Time 1 - 2 I~iz osc.
Wrfte/Ersse 10,000 cycles Endurance APPENDIX B

B+ COMP RAT('11~ SpF~'TFTrs~ rTn~~

SYM B. V
VIO Input offset voltage 2~0 mV

max VH Hysteresis 20 mV
i m IIO Input offset current n 1 nA max IIB Input bias current 20 Na max VIN Input voltage range +SS
to V

V
REF

A~ Voltage amplification 80 dB min VOL Output voltage eIOL = 3 ma 700 mV max _ VOH Output voltage ~IOH = -3 mA VpD -700 my mm tR Response time VDIFF = 100 mV 1 acs max 2 Tpg-QUADCOMPARATOR SPECIF A TIONS

~ TY~ ~ UNITS

VIO Input offset voltage t20 mV

IIO Input o!lset current 0 - tl nA

IIB Input bias current 20 VIN Input voltage range VSS - VDp V
V Ne ti ~- ga 1.3 V
ve threshold 1.Z 1.25 (output high) 1~ Voltage amplification 80 dB

VOL Output voltage 0 - 700 mV

!IOH'3m~1 V Out t OH pu - VDD mV
3 voltage VDD-700 ! I0~3a~r tR Response time 1 acs VpIF X100 aV
t H
~
i R ys 8o my eres e 20 -~ TABLE B3 COMP~TOR SDRPT11Tr1TT nnQ
VIO Input o!lset voltage t20 mV max IIO Input o!lset current 1 nA max IIH Input bias current 20 Nir max VIN Input voltage range +Sg V VREF

1~p Voltage amplification 80 dB min V0L Output voltage eIOL'3m~r 700 mV max V

0H Output voltage lIOH'-3mA V p-70o mV
m~

n tR Response time VDIgF=100 mV i~cs max APPENDIX C
a AUTO- ERO STATE l~ar'ltTrtF
STATE DIAGR~~t FIG. C-1 is a state diagram for the auto-Zero func-tion. Each bubble represents an operating state. The ar-rows between state show permissible transitions and the conditions required for the transition. Table C-1 shows the transitions from state to state in table form.

AUTO-ZERO STATE ASS NM uT
State RO R1 R2 3TT~AReSYh SO
IDLE
M U h STAZhsA
Yh S
uSEC
TiMCUTh~

10~ DELAY ~vAMPbFUL~b) TIMOUTh~ - ~
~ S3 VAMPb= CLOCK LATCH

fULLb ADZ

~s S2 C L K CL OCK
crock CN T R

CLR

2o CNTR S~

TIMoUTh LA T CH

CAMPb= CLOCK A M Z

FUILb is CLK S TIMOUThs uSEC

CNTR DELAY ~ '~'' b u~' CLOCK

STATE T ,NSITION T BrF
aFIG. C-Z shows the state assignment on the 3-bit Karnaugh map along with the allowable state transitions.

0 SD S4 S3 S1~
to R 0 1 S ~ ~~5 ~ S6 I S2~
FIG. C-Z. Auto-Zero Karnaugh Mapping 15 TB~NSITION TAH TATS EO~ItATIONS
The logic equations !or Ro, R1 and RZ are given in the following sections. The equations are written for the states in the transition table where the resulting state is a logic 1. For example, in the first line o! the tran-20 sition table, the only equation that needs to be written is the equation !or R1 resulting from the transition from state SO to Sl. The equations blow are referenced to the transition table by the originating state.
The following notation is used:
Z5 ~' not operator ' arid operator + or operator h su!!ix high true signal b ~u!!ix low true signal 30 d sutlix flip-flop D input signal The ~ operator has precedence over the + operator.

~~.~~~~8 SO -__,___, SO
I 000 -__,-__, ,___~ S1 I
---> 010 S1 ___,___, S1 010 ----_, 010 I

I
.___~ S2 '"-> 110 I

I
.___, S3 .___, 011 S~ -______~ S1 110 -_-___ -> 010 S3 ______, SI 011 -______, ~1 s~ -______, ss ool _______, lol ss _-_.___> ss lol ---.___, lol ---> se ---> 111 I

.___, r ,___, loo Sa --_-___, SS 111 -______, 101 S7 _______, SO 100 -_.____, 000 FIG.
C-3.
Auto-2~ro State Transition Table TIONS FOR ROd state SO No t~r~a n~~d~d for this stag.

state Sl ROd ~ !ROh * Rlh * !R2h * TIMOUTh * VAMPb stag SZ No t~raa nudad for this stag.

stag S3 No terse n~~d~d for this stag.

stag S4 ROd ! ROh * ! Rih * R2h state S5 ROd !ROh * !Rih * RZh stag S6 ROd ROh * Rih * RZh stag S7 No t~rsa nudad for this stag EDUCED STAT - O tATIONS FOR R~

Reducing the terse which only involve ROh, Rih and RZh, using Karnaugh map techniques, and coabining with 2iQ~~5~

terms which cannot be reduced yields the following equa-tion afor ROd.
ROd = !ROh * Rlh * !R2h * TIMOUTh * VAMPb & FUhLb +
!Rlh * R2h ROH * R2h state SO Rid = !ROh * !Rih * !RZh * STAZh !ARHSYh *

state Sl Rid = !ROh * Rlh * !R2h state S2 Rid = ROh * Rih * lR2h state S3 No terms needed for this state.

state S4 No terms needed for this state.

state S5 Rid = ROh i /!Rih * R2h * TIMOUTh CAMPb FULLb *

state S6 No terms needed for this state.

state S7 No terms needed for this state.
REDUCED ST

ATE .nlTlTTnwt c sne El Rld !ROh * !Rlh * !R2h * STAZh !ARHSYh = * +

ROh b +
*
!Rlh *
R2h *
TIMOUTh *
CAI~b is FULL

Rl !R2 *

state SO No terms needed for this state.

state S1 R2d = lROh * Rih * lR2h (!VAMPb * TIMOUTh * +
! FUhLb ) state S2 No terms needed for this state.

state S3 RZd = lROh * Rlh * R2h state S4 R2d = lROh * !Rlh * R2h state S5 Rid = ROh * !Rih * R2h (lTIMOUTh TIMOUTh . * +

Cl~lPb * FULLb ) state S6 R2d = ROh i Rlh * R2h state S7 No terms needed for this state.
REDUCED

STATE EOL1ATTONS FOR R?

R2d lROh * Rlh * !R2h * TIMOUTh (lV~b + FULLb) = * ! +

ROh lRlh * R2h * !TIMOUTh * +

ROh lRlh * R2h * TIMOUTh * b +
* Cab * FULL

!ROh * R2h +

Rlh 'R2h *

OUTPUT EQUATTnu~
!ZER~ESb = !ROh * !Rlh ZERCLKh = ROh * Rih TII~tEQh = !ROh * Rlh * !R2h + ROh * !Rlh * R2h PIZHSYh = !(!ROh * !Rlh * !R2h EOJ~Zh = ROh * ! Rlh * ! R2h CJ~rZh = ROh * R2h + ! Rlh * R2h VJ~Zh = Rlh * ! RZh CZCLKh = !CAZh VZCLRh = lVAZh APPENDIX D
a AUTO-RANGE STATE ~ta~urar~
STATE DIAG ant FIG. D-1 is the state diagram for the auto-range function. Each bubble represents an operating state. The arrows between state show permissible transitions and the conditions required for the transition. Table D-1 shows the transitions from state to state in table form.

STATE ASSIGNMENT
AUTO-ZERO STATE AS'.S'TC't~t~r~r State RO R1 R2 ~r a ~'AZ9SYh SO
iD~E
CLOCK

STADChA~ S

_ RESE PULSE
T

SHIP
T

REG

ANAEOCh CLOCK

2o M a S5 h S USEC
~TIMOUTh~ATORNGh=

DELAY ANA
RANGEh)~ Q

_ TIMQUThsA
Q G

TIMpuTh~ SOC3b ATORNGhs SHIP T PULSE

_ CL SOC3b K

SATE TRANSITION TAB
aFIG D-2 shows the state assignment on the 3-bit Karnaugh map along with the allowable state transitions.
00 Ol 11 10 0 S~ S1 ''S2 S3 1~ R 0 1 S6 ~s ~v FIG. D-2. Auto-Range Karnaugh Mapping TRANSITION TABLE ST1T1' EpUATIONc The logic equations !or RO R1 and R2 are given in the following sections.
STATE EQUATIONS FOR RO
state SO No terms needed !or this state.
state S1 No terms needed !or this state.
state SZ ROd ~ !ROh * Rlh * R2h * (TIMOUTh * ATORNGh RANGEh + TIMOUTh * !ATORNGh) Z5 state S3 No terms needed !or this state.
state S4 ROd = ROh * Rlh * R2h state S5 ROd ~ ROh * !Rlh * R2h state S6 No terms needed !or this state REDUCED STATE EQUATIONS FOR RO
Reducing the terms which only involve ROh, Rlh and RZh, using Karnaugh map techniques, and combining with terns which cannot be reduced yields the following equa-tion !or ROd.
ROd ~ !ROh * Rlh * R2h * TIMOUTh * ATORNGh * RANGEh +
!ROh * Rlh * Rah * TZMOUTh * !ATORNGh +
ROM * Rah stag No terms needed for thisstate.
SO

state S1 R1d = !ROh * !R1h * R2h state S2 Rld = !ROh * Rlh * !R2h state S3 Rid = !ROh * Rlh * !R2h state S4 Rld = ROh * Rih * R2h SOC3b *

state S5 No terms needed for thisstate.

state S6 No terms needed for thisstate.
REDUCED STA
nIl~

TTntJ e FnR R' Rld = ROh TE E
* Rlh * R2h *SOC3b +

!ROh * Rlh +

ROh *

STATE EQUATIONS ~n ~ R2 state SO R2d = !ROh * !Rih * lR2h* !AZBSYh * STADCh stag S1 R2d = !ROh * !Rih * R2h state S2 R2d = !ROh * Rih * R2h (!TIMOUTh + TIMOUTh *

ATORNGh * RANGEh) state S3 R2d = !ROh * Rih * !R2h state S~ R2d = ROh * Rih * R2h stag S5 R2d = ROh * !Rih * R2h !ANAEOCh *

state S6 No terms needed for thisstate.
REDUCED STAT

Reducing involve ROh, Rlh and the terms which only R2h, sing u ICarnaugh map techniques, and combining with teraa which cannot be reduced yieldsthe following equa-tion or f Rsd.

RZd = !ROh ~ !Rlh * !R2h * !AZBSYh STADCh +
*

lROh * lRlh * R2h +

lROh * Rlh * R2h * !TIMOUTh lROh * Rlh * R2h * TIMOUTh * ATORNGh * RANGEh +

lROh ! Rlh * R2h * TIMOUTh ATORNGh +
* !

lROh * Rlh * !R2h +

ROh Rih * R2h +
*

ROM lRlh * R2h * !ANAEOCh *

3 Q~pUT EQUATIONS

GRESh ~* lRlh * R2h * !SMCLRh = IROh GCLICti * Rlh * ! R2h * ! SMCLRh = !
ROh TIMREQh = !ROh * Rlh * R2h ~~~~2~~

ARBSYI~s !(!ROh * !Rlh * !R2h) EOCh ~ ROh * !Rlh * !R2h ANASOCh = ROh * Rlh * R2h 2~.~~~~8 APPENDIX E

a Stress ratings for the IC 10 are provided in Table E1. Stresses above ratings provided in Table A1 c an cause permanent damage to the device.

TAHLE_E1 Supply Voltage Range -0.3V to +7.0V

Temperature Under Hias -55C to +125C

Storage Temperature -65C to +15oC

10. Input Diode Currant 1-mA

Input Voltage (not ZRQN or EXPN) VSS - 03V to VDD + 0.3V

Input Voltage IRAN ~ EXPN VSS -03V to 2 x VDD + 0.3V

Continuous Output Current 25 mA

Continuous Supply Current l00 mA

Dissipation 0.5w Normal operating conditions for the IC to are provided in Table E2. These limits apply for no l rma ations of the IC 10. oper-TAH-, EZ
T~ Prmbient Temperature -40 +85 °C
VDD Supply Voltage 4.5 5.0 5.5 V
VIN Input Voltage 0~0 VDD V
fOSC Oscillator o.0 7.3728 8.0 ~z The DC characteristics of the IC 10 are provided in Table A3. The characteristics listed in Table a3 are valid over the operating range of temperature and voltage as delined in Table Al unless otherwise specified.

258 ~~~~~~8 TAB-DC CHARACT ERISTICS

TEST ,~ ~~8 ~ ~ UNITS

1 VIL Input Low Voltage 0.0 0.2 V

I H Input high voltage 0.7 X VDp VDp DD
V

3 V+ + Schmitt 2.7 4.i V

4 V- - Schmitt 1.1 2.1 V

5 VH Hysterasis 0.6 _ V

6 IIN Input Current - 1 7 IOZ Tri-state Leakage - 10 uA

8 VCOL Output Voltage - 0.10 V

1 = -10~A

VCOH Output Voltage ~p-0.10 - V

1 = lO~tA

10 VOL Output Voltage - 0.4 V

1 = l.6mA

11 VOH Output Voltage VDp-0.8 - V

1Z CO~ Capacitance, - 12 pF

output 13 CIN Capacitance, - 8 pF

input 14 Ipp Supply Current o 10 4 IrD~iz ~~~1~~~~

APPENDIX F

CURRENT S raSYSTEM A1~L

IFIER SPFC TFTC~s~rTnv '=VALUE

VO S Offset Voltage 0 to -1 .
my max IIO Input offset current 1 A

n UIg Input bias current max 1 nA max VIA Input common mode VSS to 1.5v A~ Voltage Amplification 80 dB
i m IO output Current n 100~A max Unity gain bandwidth 80 kl~iz min TAB

~!D CONVERTER SC~3SYSTEM SPE~TFIyATIONS

~ ~, VOg Offset Voltage 0 to -l.0 mV

max IIO Input offset current 1 nA ma x IIg Input bias current 1 nA max VIA Input common mode range to V
V

RgF
SS

-0.5 V

VIp Input range VSS to VREF

+0.5 V

Resolution 8 bits i m Linearity n tl LSe min Differential non-linearity 0 5 LSB m .
Tpg Conversion Time ax 24~s max

Claims

1. An electrical circuit interrupting device including an overcurrent trip unit comprising current sensing means for sensing an electrical current flowing through said electrical circuit interrupting device, trip means responsive to said current sensing means for generating a trip signal as an adjustable predetermined function of the time and the magnitude of the electrical current flowing through said electrical circuit interrupting device, a user interface panel provided to present a visual representation of said adjustable predetermined function, input means for selectively adjusting parameters of said adjustable predetermined function, and a plurality of indicator means associated on said user interface panel with said visual representation, each indicator means indicating a parameter of said adjustable predetermined function, characterised in that each said indicator means can take a first state in response to a trip signal generated by the trip means and a second, different state in response to adjusting of the corresponding parameter of said adjustable predetermined function through said input means.

2. An electrical device as claimed in claim 1, characterized in that said input means comprises means for adjusting selected parameters of said adjustable predetermined function and means visually presenting values of said selected parameters, said input means having a viewing mode in which said values of said selected parameters are viewed, and adjust mode in which said values of said selected parameters are adjusted, said indicator means having a first operating condition in said second state when said input means is in said viewing mode, and a second operating condition in said second state when said input means is in the adjust mode.

3. An electrical device as claimed in claim 2, characterized by means switching said indicator means out of said second state if no input is made through said input means during a predetermined time interval.

4. An electrical device as claimed in any one of claims 1 to 3, characterized in that said indicator means comprises light means having a first light output in said first state and a second light output in said second state.

5. An electrical device as claimed in claim 4, characterized in that said first light output is a first color, and said second light output is a second color, and said light means is a two color light emitting diode.

6. An electrical device as claimed claim 5, characterized by display means for visually presenting values of said selected parameters, said input means having a viewing mode in which said values of said selected parameters are viewed, and a program mode in which said values of said selected parameters are adjusted, with said second color light output arranged to flash in one mode of said input means, and to be steady in the other mode of said input means.

7. An electrical device as claimed in claim 6, characterized in that said one mode is said program mode.

8. An electrical device as claimed in claim 6, characterized by means switching said input means out of said viewing mode and program mode and turning off said second color light when no input is made through said input means during a predetermined time interval.

9. An electrical device as claimed in any one of claims 1 to 8, characterized in that a user interface panel is provided having first switch means thereon for adjusting said parameters and second switch means for testing said trip means, said first switch means being a first identifiable color and said second switch means being a second, different identifying color.