WO2013139434A1

WO2013139434A1 - Method for the ultrarapid reading of photodetectors

Info

Publication number: WO2013139434A1
Application number: PCT/EP2013/000638
Authority: WO
Inventors: Rupert Huber; Olaf Schubert
Original assignee: Universität Regensburg
Priority date: 2012-03-23
Filing date: 2013-03-05
Publication date: 2013-09-26
Also published as: DE102012102521A1

Abstract

A method and a device are disclosed for the ultrarapid reading of a plurality N of photodetectors, which are spaced at a distance from one another and are lit simultaneously by an optical pulse in such a way that each of the N photodetectors generates an electrical output pulse at least approximately simultaneously as a response to the optical pulse. Said N output pulses are delayed by a delay time by means of a delay device. The N delayed output pulses are combined to generate one pulse train signal, and said pulse train signal is read by means of a read circuit.

Description

Method for ultrafast reading of photosensors

FIELD OF THE INVENTION

The present invention is in the field of photosensitivity. More particularly, it relates to a method and apparatus for ultrafast reading of at least two spaced-apart photodetectors and a spectrometer utilizing such an apparatus and method.

RELATED ART In a variety of applications, both in industry and in basic research, temporal changes or the movements of objects are detected using optical detectors. Applications include production monitoring, quality control, chemical analysis, safety engineering, biosensing and medical diagnostics.

To monitor the products moving on conveyor belts in production, for example, they are guided past a permanently installed detector line of a line scan camera. The detector line, which consists of a large number of photodetectors, is read out at short intervals at a specific sampling rate. When recording very fast moving objects so-called TDI-CCD cameras (TDI = Time Delay Integration) are used, which consist of a plurality of detector lines and have an increased sensitivity. Here, the signals of different lines are summed up at different times. Thus, the sensitivity of the recording is increased.

Another application, in which a plurality of photodetectors is read out, forms the spectroscopic examination of substances. In a spectrometer, the radiation to be analyzed is spatially separated so that different spectral radiation components run on different optical paths after separation. For this purpose, for example, a diffractive optical element can be used. The spatial intensity distribution thus provided can then be detected by optical detectors which are spaced from each other. For signal analysis, the signal generated by the detectors is further processed, for which purpose the detectors are read out.

In many cases, said spectrometric applications are used to observe time-varying processes. A prerequisite for this is the possibility of a time-resolved execution of the optical detection measurements at the spatially spaced apart detectors. If the processes to be investigated run on very short time scales, it is also essential that the signals are generated quickly and sensitively on the one hand, and that they are further processed correspondingly quickly after generation, on the other hand.

One possibility for time-resolved analysis is the use of pulsed radiation whose spectral information - after interaction with the sample to be examined - is determined. Time-sequential optical pulses characterize the sample at different times and may be captured by a photodetector as "snapshots." Further, in a spectrometer, after spatial separation of the spectral components of the optical pulse at particular photodetector positions, individual spectral slices may be received in time-resolved fashion.

When a photodetector receives optical radiation, electric charges are generated therein. For reading the photodetector in a sample and hold circuit, the charges are latched in response to a sample signal for the period of the sample signal. The signal strength of the analogue output signal corresponds to the charge temporarily stored at the respective time. Thus, the signal strength of the optical radiation is approximated by a stepped, analogue output signal, the step width corresponding to the period of the scanning signal. This output signal can then be converted by an analog-to-digital converter into a digital signal and further processed for analysis. Such reading is described in US Pat. No. 3,486,822 for a spectrometer.

Another possibility for recording and reading of spatial intensity distributions offer spatially extended CCD sensors. These form the optical detector and the readout a common, integrated device. To read such detectors, clock signals in the MHz range can be used. The sampling rate essentially results from the product of the inverse of the number of pixels and the clock signal frequency. Dead-time clock cycles in which dark pixels are output are not taken into account. The sampling rates for very fast line detectors are usually in the range of a few 10 kHz, ie they correspond to a time resolution of about 100 μ8. The use of actively switched components for signal processing thereby results in a relatively high noise level. These components have a complex circuit with its own timing, which limits the sampling rate. To prevent smearing of the data by the readout technique, CCD imaging systems have a mechanical shutter or are modulatedly illuminated. The modulation can be generated with an electro-optical modulator, with an acousto-optical modulator or with a rotating perforated disc. In this case, it is necessary to synchronize the sampling rate with the modulator. In practice, the mechanical shutter or the modulating components limit the technically achievable sampling rate.

Alternatively, spatial intensity distributions can also be measured by mechanically adjustable detection mechanisms. For this purpose, for example, a photodiode used for detection can be moved by means of a displacement table, as described in the patent application US 4,124,297. Another possibility describes the patent application US 4,732,476, in which the scanning of a spectral intensity distribution by rotating an optical grating takes place. Here, the recording duration of a spectrum is in the range of milliseconds to minutes. For a temporal analysis of spectra in the microsecond range, these recording methods are therefore unsuitable.

Another method uses a spatial arrangement of photodiodes, for example a photodiode array, and uses for reading each photodiode each have their own preamplifier and a separate analog-to-digital converter. The digital signals thus generated contain the optical information of the detected spectrum at the location of the photodiode at the time of detection. Very high sampling rates can be achieved with this method. However, this method is very costly, especially when additionally using lock-in amplifiers to process the photodiode signals. The previously known methods for time-resolved reading of a detector device for recording a spatial intensity distribution have the disadvantage that the time resolution is either limited to a range of about 100 μ3 or can only be achieved at considerable expense.

SUMMARY OF THE INVENTION

The present invention has for its object to provide a method that allows a large number of spatially spaced photodetectors, which are illuminated by optical pulses with repetition rates up to the MHz range, inexpensive and low noise read. Another object is to provide a device that uses this method.

This object is achieved by a method according to claim 1 and by a device according to claim 9. Advantageous developments are specified in the independent claims.

Modern ultrashort pulse lasers emit optical pulses with typical pulse durations in the nano-, pico- or femtosecond range with repetition rates in the MHz range. When a plurality N of spaced-apart photodetectors are illuminated with optical pulses, each photodetector generates an output electrical pulse approximately simultaneously in response to each of these optical pulses. The repetition rate of the electrical output pulses generated by a photodetector corresponds to the repetition rate of the optical pulses with which the photodetector is illuminated. Thus, upon detection of a pulsed spatial intensity distribution, a plurality N of pulsed electrical signals are generated. So far, it is not possible to read these signals as quickly as they are provided in a cost effective manner.

The present invention solves this problem by a method of reading out a plurality N of photodetectors, where N is an integer> 2 and the photodetectors are spaced apart and illuminated simultaneously by an optical pulse so that each of the N photodetectors respond in response to the photodetectors optical pulse generates at least approximately simultaneously an electrical output pulse. The method comprises the following steps: Delaying the N electrical output pulses by means of a delay means by a respective delay time Τγ _{, ί} , i = 1 ... N, wherein at least Nl of the delay times Τγ _{, ϊ} are different from zero,

Merging the N delayed output pulses to produce a pulse train signal superimposed on the N delayed output pulses, the delay times Ty.i being selected such that the contributions of the individual output pulses in the pulse train signal do not substantially overlap in time, and

- Reading the Pulse train signal by means of a read-out circuit, which is adapted to assign each photodetector a temporal portion of the signal corresponding to the delay time Tv _, i of the associated delay means.

The fact that the photodetectors are spaced from each other and thus located at different spatial positions, the information contained in the optical pulse can be detected location-dependent by the plurality N of photodetectors. Each photodetector thus detects optical information, for example the intensity of a specific radiation component, at a specific spatial position. The optical pulse may include, for example, visible, Near Infrared (NIR), infrared (IR), ultraviolet (UV), and / or other radiation regions. Preferably, the present invention is used in the visible and NIR range and with photon energies of preferably <100 eV, preferably <10 eV and most preferably <3 eV.

The information contained in the optical pulse is represented by the temporal and spatial spectral intensity distribution. With spaced-apart photodetectors, the spatial spectral information can be detected either without or with prior spectral separation. If the spectral sensitivity ranges of the photodetectors are selected differently from each other and narrow with respect to the spectrum of the optical pulse, only a corresponding spectral section of the spectrum of the optical pulse is detected at the location of each photodetector. In this case, a preliminary spectral separation is not necessary. On the other hand, if the spectral components of the optical pulse are spatially separated before detection, they can be identical spaced apart photodetectors that are sensitive in the entire spectrum of the optical pulse can be detected. The spectral separation can be made by means of an optical grating or a refractive optical element. The N photodetectors generate a plurality of N output electrical pulses in response to the incident optical pulse. Each of these electrical output pulses contains the information of the optical pulse at the location of the photodetector. Since the optical pulse impinges on the plurality N of photodetectors approximately simultaneously, the plurality of output electrical pulses are also generated approximately simultaneously in response to this optical pulse. Delays in the propagation of an optical pulse propagating at the speed of light or else different reaction times of the photodetectors can lead to different times of impingement of the optical pulse on the photodetectors or at different production instants of the electrical output pulses. For the present invention, however, these differences can be neglected, otherwise this influence can also be compensated by a suitable choice of the delay means. In the present description, the temporal relationship between the generation times is referred to as "approximately simultaneous." The N output pulses which are generated at least approximately simultaneously are delayed in a first method step by respective delay times TV.i, i = 1 The N delay times Ty _, i are different from each other in pairs, so that the delayed output pulses are offset in pairs in time with respect to each other It is also possible that one of the delay times is zero, ie all output pulses except one are delayed in time.

In a second method step, these N delayed output pulses are combined to form a pulse train in a pulse train signal. Thus, N individual signals are combined into a single signal, the pulse train signal. Each electrical output pulse is represented in the pulse train signal in the form of a contribution from a particular photodetector. Due to the different delay times Τγ, ί, successive contributions of a pulse train in the pulse train signal have a time interval from each other which corresponds to the difference between the associated delay times Τγ, ί. Furthermore, the contributions in the pulse train signal occupy a certain temporal section, which is determined by the corresponding delay time Τγ _, ί and a specific offset value.

After the delayed output electrical pulses have been merged into the pulse train signal, the pulse train signal is read out using a readout circuit. The read-out takes place in such a way that each photodetector is assigned a specific time segment of the pulse train signal belonging to a specific delay time Ty, i. If several optical pulses are read out, the abovementioned offset value can be determined by means of a trigger signal, so that each contribution can be assigned to the corresponding detector and the corresponding optical pulse. The trigger signal can be generated, for example, from the output signal of an additional photodetector or from a signal which has been diverted to one of the N instantaneous or delayed electrical output pulses. Alternatively, the trigger signal can also be generated by the control of the pulsed laser.

So that the individual contributions in the pulse train signal can be identified by the read-out circuit and thus assigned to the respective photodetector, it is necessary that the contributions in the pulse train signal do not substantially overlap in time. This is achieved by a suitable choice of the delay times Tv _, i. The term "substantially non-overlapping in time" in this context means that successive contributions in the pulse train signal are so far apart in time that the read-out circuit can still distinguish them from one another and thus assign the information contained to the respective photodetector In order to ensure that successive contributions of the same pulse train do not essentially overlap in time, it can vary as a function of the pulse width of the contributions The pulse width usually corresponds to the half width, ie the full width of the contribution at half the amount of the contribution The critical time interval, ie the distance at which successive contributions of the same pulse train do not substantially overlap in time, can be determined by the further processing of the pulse train signal to be dependent. Regardless of the pulse width can be ensured by suitable choice of the delay times Τγ _{, ί} that the time interval is equal to or greater than the critical time interval. The method according to the invention thus converts information from a spatial intensity distribution into a single pulse train signal. The spatial position of a photodetector is translated into the position of a temporal section in the pulse train signal. By assigning a specific time segment to each photodetector during readout, the spectral information contained in a contribution is assigned to the corresponding spatial position. In this way, with the method of the present invention, the information of the optical pulse is read out without complicated and cost-intensive electronic components and without averaging over several sampling cycles. The read speed corresponds to the speed at which the information is provided. When using an ultrashort pulse laser with repetition rates in the MHz range, the frequency at which the pulse trains are generated in the pulse train signal is also in the MHz range. The frequency of the contributions in the pulse train signal is greater by the factor N.

Delaying and merging can take place, for example, temporally and spatially separated from one another. For this purpose, the N output pulses are delayed by means of the delay means first individually and in pairs separately from each other to the N delayed output pulses. Subsequently, these N delayed output pulses can then be merged.

However, the first process step of deceleration and the second process step of merging are not necessarily performed one after another but may also be combined or combined with each other. For this purpose, for example, a certain number of delay sections are connected in series and different photodetectors are each connected to different points of this series connection. Thus, different electrical output pulses each enter the series circuit at different points and each undergo a different number of delay sections. To generate a pulse train signal by combined delaying and combining, for example, the Nth electrical output pulse is delayed by one time (TV _, N-TV, NI) and then combined with the (Nl) th instantaneous electrical output pulse to form a two-contribution signal. This two-contribution signal is delayed by the time (Τν, Ν-ι-Τν _, Ν-2) and with the (N-2) th instantaneous electrical output pulse a signal with three contributions. These steps are repeated until finally a signal with Nl amounts is delayed by the time (T _{v> 2} - T _v, i) and then merged with the first instantaneous electrical output pulse to the pulse train signal.

Delaying is preferably done using passive delay means, "passive" meaning that no active switching, such as using a clock signal, is necessary for deceleration, for example a passive delay may be caused by the propagation delay in a cable For example, the use of passive delay means allows a readout in quasi-real time, ie the signal readout is carried out at practically the same speed as the signal supply and is only delayed by transit time differences A limitation of the time resolution by active delay, such as due to a buffering or due to a clock signal frequency, is not available.

In a preferred embodiment, the N delay times Ty _, i have a value which corresponds to a multiple of a predetermined time interval At. Preferably, these result from Tv _, i = (il) At for i = 1 to N. Thus, the time interval between two consecutive contributions of the same pulse train is constant and corresponds to the value of the predetermined time interval At.

Furthermore, it is advantageous to detect not only one but a plurality of optical pulses which successively strike the plurality N of photodetectors at a repetition rate f _rep . If a pulsed laser source is used to scan the sample to be examined and the photodetectors are illuminated with this plurality of optical pulses interacting with the sample, a pulse train signal is generated containing a pulse train of N contributions for each optical pulse. The contributions of the same pulse train take in the pulse train signal a time range Τρχ, the magnitude of which depends on the choice of the delay times Ty _, i. To ensure that the contributions of successive pulse trains do not substantially overlap in time, the delay times are chosen so that the temporal range Τρχ is less than 1 / f _rep . With the above-mentioned delay times of T _v, i = (il) -At is thus the relationship N-At <1 / frep required.

A typical repetition rate with which the method of the present invention can be used is in the range of f _rep = 10 MHz. This corresponds to a pulse interval l / f _rep of 100 ns. The readout rate with which the plurality N of photodetectors are read out in the present invention is> 1 kHz, preferably> 10 kHz, and more preferably> 100 kHz. So that neither the contributions of the same pulse train nor the contributions of successive pulse trains essentially overlap in time, on the one hand T _PT must be less than 1 / f _rep , on the other hand the difference (Tv.i ₊ i - Tvj) must not for i = 1 ... Nl less than the critical distance that depends on the pulse width of the posts. In order to be able to read as many photodetectors as possible, the critical distance should be as small as possible and thus the pulse width of the contributions should be as narrow as possible. The pulse width of the contributions is in the range of the reaction time T _{R of} the photodetector and is limited by this. The response time T _{R of} a photodetector is usually defined by the time it takes for a signal to increase from 10% to 90% of the maximum value. The reaction time T _{R of} a photodetector is approximately the limit frequency f _c by T _R ~ 0.35 / fc, wherein the cutoff frequency f _{c is} the frequency at which the signal value has dropped to around 50% of the maximum value. In the context of the invention, therefore, fast photodetectors are preferably used which have reaction times T _R which are preferably <50 ns, preferably <5 ns, particularly preferably <0.1 ns. Accordingly, the photodetectors used have cut-off frequencies f _c which are preferably> 0.007 GHz, preferably> 0.07 GHz, particularly preferably> 3.5 GHz. The delay times are chosen so that the predetermined time interval Δt can preferably be chosen to be less than 100 ns, preferably less than 10 ns and particularly preferably less than 100 ps.

Preferably, the pulse train signal is amplified by means of a preamplifier, before the assignment of the temporal sections to the respective photodetector takes place. In a further advantageous development, the pulse train signal is digitized by means of an analog-to-digital converter before the assignment of the time segments to the respective photodetector takes place. Note that only a single preamplifier and a single analog-to-digital converter is needed to read the N photodetector signals for further processing. This shows a particular advantage of generating the analogue Pulse train signal from the plurality of delayed individual signals. If, on the other hand, one digitized the output signal of each photodetector directly, it would be necessary to use N preamplifiers and N analog-to-digital converters, which would considerably increase the design effort and the production costs. Furthermore, identical preamplifiers have different noise and not exactly the same gain. This would increase the noise when using N preamplifiers. In addition, when N analog-to-digital converters are used, N individual signals must also be synchronized with each other. For this purpose, a precise consideration of the maturity is necessary, which further increases the effort. Therefore, there are significant advantages associated with the use of only one analog-to-digital converter and only one preamplifier as discussed above.

Preferably, the method according to the invention is used in a spectrometer for determining the spectrum of an optical pulse. In the spectrometer, the spectral components of the optical pulse are spatially separated. This separation is made, for example, as previously mentioned, by means of a refractive optical element or by means of an optical grating. The spectrometer comprises a plurality of N photodetectors which are spaced apart and illuminated with different spectral components of the optical pulse. The readout of the photodetectors is carried out as described above. The contributions in the pulse train signal contain the spectral information of the optical pulse at the location of the photodetector. This information can be represented, for example, by the amplitude of the respective contribution, the width of the respective contribution, the integrated amplitude of the respective contribution or by another quantity of the respective contribution of the pulse train signal which is indicative of the intensity of the light received at the associated photodetector. By determining this quantity, the detected radiation intensity at the location of the photodetector is determined.

In addition to the method described above, the invention also includes a device for reading out the plurality N of photodetectors using this method.

The device according to the invention comprises the following elements:

Delay means for delaying at least N1 of the N electrical output pulses by a delay time Ty.i, i = 1... N, Means for merging the N delayed output pulses to produce a pulse train signal superimposed on the N delayed output pulses, the delay means being arranged to delay the electrical output pulses such that the contributions of the individual output pulses in the pulse train signal do not substantially overlap in time;

- means for reading the pulse train signal, which are adapted to each photodetector to assign a temporal portion of the signal corresponding to the delay time Tv _, i of the associated delay means.

The delay means can be realized, for example, by cables of different lengths, so that the electrical output pulses generated by the photodetectors travel different distances through the cables. Due to different lengths of maturity, the electrical output pulses are delayed by different delay times Tv _, i and thus offset in time. Another possibility is to use commercially available passive delay elements that have a rise time of no more than 1 ns and delay times in the range of about 100 ps to nanoseconds. If necessary, several of these elements can be connected in series to achieve higher delay times. A further embodiment provides a propagation delay directly in the semiconductor material of the photodetector. Due to a relatively low propagation speed of the electrical output pulses in semiconductors, therefore, only a relatively small distance is required.

The means for combining the delayed output pulses may be, for example, commercially available combination elements. These include, for example, N input cables which are connected together at their outputs and terminate in a single output cable.

In an advantageous development, the means for reading comprise an analog-to-digital converter. With this an analog pulse train signal can be converted into a digital signal and then further processed with a suitable data processing system. Preferably, the means for reading comprises a preamplifier. With this, the contributions of the pulse train signal can be amplified, whereby the further processing is simplified.

In a further development, the device comprises a plurality N of photodetectors for detecting the optical pulse and for generating the electrical output pulses. For use with visible light, for example, silicon photodiodes or photodiode cells can be used.

An advantageous development of the invention relates to a spectrometer for determining the spectrum of an optical pulse using a plurality N of photodetectors, where N is an integer> 2 and the photodetectors are spaced from each other and arranged so that they each light of an associated spectral range of the optical pulse receive. The spectrometer includes the aforementioned delay means, means for merging and means for reading. Furthermore, the spectrometer comprises means for determining a size of the pulse train signal which is indicative of the intensity of the light received at the associated photodetector. The determined variable can be assigned to the spatial position of the corresponding photodetector and thus the spectrum of the optical pulse can be determined.

BRIEF DESCRIPTION OF THE FIGURES

Further advantages and features of the invention will become apparent from the following description in which the invention with reference to preferred embodiments with reference to the accompanying drawings is explained in more detail. Show:

Figure 1 is a schematic representation of a preferred embodiment of

readout device according to the invention,

FIG. 2 shows a schematic representation of a pulse train signal,

Figure 3 is a schematic representation of an alternative embodiment of the

readout device according to the invention, FIG. 4 shows a schematic representation of the structure for measuring a pulse train signal,

FIG. 5 shows a measured pulse train signal using two photodetectors.

DESCRIPTION OF THE PREFERRED EMBODIMENT

Figure 1 shows a schematic representation of a preferred embodiment of the readout device 8 according to the invention, with which a pulse train signal 10, as shown in Figure 2, can be generated. More specifically, FIG. 1 shows on the left side an optical pulse 12 and, to the right, a plurality N of photodetectors 14. The photodetectors 14 are connected to delay means 18. These are followed by means 22 for combining delayed output pulses 20, to which an analog-to-digital converter 24 is connected, followed by a data processing system 26. In each photodetector 14, an electrical output pulse 16 is generated in response to the optical pulse 12, which is shown schematically in Figure 1. In the direction of passage behind the delay means 18 also delayed output pulses 20 and behind the means 22 for merging a pulse train signal 10 are shown schematically.

The function of the read-out device 8 will be described below. On the left side in FIG. 1 it is shown how the optical pulse 12 impinges on a plurality N of photodetectors 14. The optical pulse 12 is represented by a plurality of arrows. In this case, an arrow corresponds to a spatial section of the optical pulse 12, which impinges on a specific photodetector of the plurality N of photodetectors 14. In response to the optical pulse 12, each of the plurality N of photodetectors 14 generates an electrical output pulse 16 approximately simultaneously. The delay means 18 causes the output electrical pulses 16 to be delayed by a delay time T _V; j = (il) At, for i = 1 ... N, delayed, where At is a predetermined time interval, such that for each output electrical pulse 16 there is a delay

Output pulse 20 is generated. In the language of the present disclosure, the output pulse 16 of the first photodetector 14 would be a "delayed" output pulse delayed by a delay time of zero to standardize the description Output pulses 20 of the (i + l) -th and (i-l) -th photodetector 14 temporally offset by the predetermined time interval At. By the means 22 for merging the delayed output pulses 20 are combined to form a pulse train signal 10. The analog pulse train signal 10 is digitized by means of an analog-to-digital converter 24 and the thus digitized signal is forwarded to a data processing system 26.

FIG. 2 shows the time profile of the signal strength of a pulse train signal 10, as generated by the read-out device 8 shown in FIG. The illustrated pulse train signal 10 includes three pulse trains 48, each containing a number N of contributions 28. The totality of the contributions 28 of the same pulse train 48 increases in the pulse train signal

10 a time range 30 of duration Τρχ a. The contributions 28 each have a pulse width 44 which corresponds to the full width of a contribution 28 at half the contribution level. The time interval of successive contributions 28 of the same pulse train 48 is At and is shown by way of example for the distance between the first contribution 28 and the second contribution 28 of the right-hand pulse train 48 in FIG. This distance results from the delay times of T _v, i = (il) -At.

It can be seen in FIG. 2 that the time interval Δt is sufficiently greater than the pulse width 44, so that the pulses of the same pulse train 48 do not overlap in time, or at least substantially do not overlap. In this context, "substantially not overlapping in time" means that successive contributions 28 in the pulse train signal 10 are so far apart in time that they can still be separated from one another during further processing of the pulse train signal 10, so that the information contained in each contribution 28 corresponds to the corresponding one The special case that the pulses do not overlap in time is explicitly included The temporal range 30 has a duration of Τρτ = (Nl) Δt. Furthermore, in FIG. 2 the pulse spacing 32 is between the contributions 28 of the same photodetector The pulse train spacing 32 corresponds to the spacing of successive optical pulses 12. In the pulse train signal 10 from FIG. 2, the pulse train spacing 32 is greater than Τ χ + At = N-At, so that the contributions

28 different pulse trains 48 do not substantially overlap in time.

FIG. 3 shows a schematic representation of an alternative embodiment. As in FIG. 1, an optical pulse 12 impinges on a plurality N of photodetectors 14. In accordance with the numbering of the N photodetectors 14 and the associated signals used hereafter, the proportions of the optical pulse 12 in FIG. 3 are occupied from bottom to top with the numbers "1" to "N". The N photodetectors 14 are connected via N connections - represented by lines - to means 34 for merging and delaying. The means 34 for merging and delaying include Nl individual delay sections 35, of which only four are drawn. The Nl delay sections 35 are connected in series, so that the output (bottom) of a delay section 35 is connected to the input (top) of the subsequent delay section 35, and from bottom to top - in the opposite direction of the signals through the series connection - numbered from "2" to "N". For i = 2 to N, the i-th of the N photodetectors 14 is connected to the input of the i-th delay section 35. The first of the N photodetectors 14 is connected to the output of the second delay section 35 and to an analog-to-digital converter 24. Furthermore, a data processing system 26 is shown, which is connected downstream of the analog-to-digital converter 24.

In contrast to the read-out device of Figure 1, in which the delay means 18 and the means 22 for merging consist of separate components, the function of these two components in the alternative read-out device of Figure 3 in a component, namely the means 34 for Zusammenfuhr and Delay united. The Nth electrical output pulse 16 (not shown) entering the means 34 for merging and decelerating from the Nth of the N photodetectors 14 passes through the Nth delay section 35 and is thereby delayed by the time Δt. Subsequently, the N-th output pulse delayed by Δt with the (N-1) -th instantaneous output electric pulse 16 (not shown) becomes a signal of two

Contributions merged, which passes through the (N-l) -th delay section 35 and thereby in turn is delayed by the time At. These steps are repeated until finally a signal of N-1 amounts passes through the second delay section 35, thereby being delayed by the time Δt, and then brought to the pulse train signal 10 with the first instantaneous electrical output pulse 16. As everyone

Contribution 28 is delayed by the time At in a pass through a delay section 35, results in the pulse train signal 10 for the i-th contribution 28 - as in Figure 1 - a delay time of T _Vj i = (il) At. Figure 4 is a schematic representation of an experimental setup intended to demonstrate the operability of the present invention. At the top is shown a laser system 36, which provides pulsed radiation with a wavelength of 1.55 μπι, with a repetition rate of 4 MHz and with a pulse duration of 100 femtoseconds. Below the laser system 36 is a frequency doubling

Element 38 is shown for doubling the frequency of the emitted laser radiation. Below this is shown a plurality of photodetectors 14 for detecting an optical pulse and for generating electrical output signals. To the plurality of photodetectors 14, a readout circuit 40 is connected, which in turn is connected to an oscilloscope 42. Not shown in FIG. 4 is a sample to be examined with which the optical pulse 12 can interact. Since only the functionality of the present invention is to be shown by the illustrated construction of Fig. 4, the pulsed spatial laser radiation is direct, i. without prior reflection from a sample to be examined or prior transmission through a sample to be examined, directed to the photodetectors 14.

The radiation emitted by the laser system 36 is first passed through the frequency doubling element 38, which consists, for example, of a beta barium borate crystal, to halve the central wavelength of the optical pulses 12 to a wavelength of approximately 775 nm. This is necessary in the

Sensitivity range of the silicon photodetectors 14 used in the illustrated embodiment to arrive. The individual silicon photodetectors 14 are now illuminated with optical pulses 12 having a wavelength of about 775 nm at a repetition rate of 4 MHz and in response generate output electrical pulses 16, which are passed to the readout circuit 40. The readout circuit 40 includes the

Delay means 18 and the merge means of Figure 1 or means 34 for merging and decelerating of Figure 3 (not shown in Figure 4) to generate from the output electrical pulses 16 a pulse train signal 10 (both not shown in Figure 4); which is passed to the oscilloscope 42 and graphically represented by this. Note that in this construction, in which mainly the

Functionality is to be demonstrated, no sample is provided in the beam path.

FIG. 5 shows a pulse train signal 10 represented by the oscilloscope 42 of FIG. 4. The time segment shown here shows seven pulse trains 48 each having two contributions 28. The pulse train spacing 32 between the contributions 28 of the same photodetector 14 from successive optical pulses 12 is about 250 ns. The time range 30 which the contributions 28 of the same pulse train 48 occupy has a duration of approximately Τ χ = 75 ns. The pulse width 44 of the contributions 28 is about 15 ns. Furthermore, FIG. 5 shows reflections 46 whose signal strength amounts to approximately 20% of the signal strength of the posts 28. These reflections can arise, for example, at connection points of the cables.

The experimental setup of Figure 4 and the pulse train signal 10 of Figure 5 demonstrate the operability of the present invention. At the present pulse train interval 32 of about 250 ns and at the present pulse width 44 of the contributions 28 of about 15 ns, the number of illuminated photodetectors 14, which in the experimental setup of FIG. 4 is only 2, can already be significantly increased without any other modifications. without the contributions 28 in the pulse train signal 10 substantially overlapping in time. Depending on the data processing system, the number of photodetectors 14 could already be increased to 10 or more in the described construction. By means of a narrower pulse width 44 and / or a larger pulse train spacing 32, the number of detectors can even be increased considerably. This can be achieved, for example, by using photodetectors 14 with shorter response times and / or by using a slightly lower laser repetition rate. In this case, the pulse train signal 10 can optionally be optimized with respect to the required temporal and spatial resolution. The lower the requirement for the temporal resolution, the lower the laser frequency can be selected. This has resulted in a larger pulse spacing 32 so that a pulse train 48 may comprise more temporally substantially non-overlapping contributions 28. Thus, the number of photodetectors 14 and thus the spatial resolution can be increased. Conversely, the temporal resolution can also be increased by reducing the spatial resolution.

Although in the drawings and in the foregoing description preferred embodiments are shown and described in detail, this should be considered as purely exemplary and not restricting the invention. It should be noted that only the preferred embodiments are shown and described and changes and modifications that are currently and in the future within the scope of Invention are to be protected. The features shown may be of importance in any combination.

LIST OF REFERENCE NUMBERS

8 readout device

10 pulse train signal

12 optical pulse

14 plurality N of photodetectors

16 Electric output pulse

18 delay means

20 Delayed output pulse

22 Means of merging

24 analog-to-digital converters

26 data processing system

28 contribution

30 Time range

32 pulse train interval

34 Means of merging and delaying

35 delay section

36 laser system

38 frequency doubling element

40 readout circuit

42 oscilloscope

44 pulse width

46 reflections

48 pulse train

Claims

claims

A method of reading out a plurality N of photodetectors (14), where N is an integer> 2 and the photodetectors (14) are spaced apart and illuminated simultaneously by an optical pulse (12) such that each of the N photodetectors (14) generates an electrical output pulse (16) in response to the optical pulse (12) at least approximately simultaneously, the method comprising the steps of:

Delaying the N output electrical pulses (16) using a

Delay means (18, 34) for a respective delay time Tv _, i, i = 1 ... N, wherein at least Nl of the delay times Ty _, i are different from zero,

Merging the N delayed output pulses (20) to produce a pulse train signal (10) superimposed on the N delayed output pulses (20), wherein the delay times Tv _, i are selected so that the contributions (28) of the individual output pulses (16) in Pulse train signal (10) do not substantially overlap in time, and

Reading out the pulse train signal (10) by means of a readout circuit (40) which is suitable for assigning to each photodetector (14) a time segment of the pulse train signal (10) corresponding to the delay time Τγ _{, ί of} the associated delay means (18).

The method of claim 1, wherein the step of delaying the N electrical output pulses (16) by means of a delay means (18) occurs prior to the step of merging the N delayed output pulses (20).

The method of claim 1, wherein the step of delaying the N electrical output pulses (16) comprises the step of combining the N delayed output pulses (20) by means of a merging means (34) and

Delaying is combined, wherein the means (34) for merging and Delay a plurality of cascaded

Delay sections (35) includes and various electrical

Output pulses (16) each have a different number of

Go through delay sections (35).

4. The method of claim 1 or 3, wherein the step of delaying by means of combining and delaying means is combined with the step of merging such that for i = 1 ... Nl an ith signal around the Time (Tv, N + ii -Tv, Ni) is delayed and then combined with the (Ni) -th N output pulses (16) to an (i + l) -tem signal, wherein the first signal through the N- of the N electrical output pulses (16) and the pulse train signal (10) by the time Τγ _, ι delayed N-th signal is formed.

5. The method according to any one of the preceding claims, wherein the

Delaying by means of a passive delay means (18, 34), in particular without the use of a clock signal, takes place.

6. The method according to any one of the preceding claims, wherein the

Delay times Ty _, i are given by Tv, i = (il) -At, where At is a predetermined time interval.

7. A method according to any one of the preceding claims, wherein a plurality of optical pulses (12) successively impinge upon the plurality N of photodetectors (14) at a repetition rate f _rep , the delay times Tv _, i being chosen such that the contributions ( 28) of the same optical pulse (12) in said pulse train signal (10) occupy a time range (30) of duration Τρχ which is less than 1 / f _rep .

8. The method of claim 6 or 7, wherein N-At <1 / f _rep , wherein At

predetermined time interval and f _rep the repetition rate of the N plurality of photodetectors (14) incident optical pulses (12).

9. The method according to any one of claims 6 to 8, wherein Ät <100 ns, preferably At <10 ns, particularly preferably At <100 ps.

10. The method according to any one of the preceding claims, wherein the

Pulse train signal (10) using an analog-to-digital converter (24) is digitized.

11. The method according to any one of the preceding claims, wherein the

Pulse train signal (10) is amplified by means of a preamplifier. 12. The method according to any one of the preceding claims, wherein the readout of the plurality N of photodetectors is carried out at a read rate which is> 1 kHz, preferably> 10 kHz, and more preferably> 100 kHz.

13. A method for determining the spectrum of an optical pulse (12) by means of a spectrometric device, the N spaced apart photodetectors

(14), wherein N is an integer> 2, and the photodetectors (14) are arranged to respectively receive light of an associated spectral region of the optical pulse (12), comprising the steps of: reading the plurality N of photodetectors (14) 14) according to one of the claims

1 to 12,

Determining a magnitude of the pulse train signal (10) in said time sections associated with the photodetectors (14) to determine the spectral components corresponding to the photodetectors (14), the magnitude of the pulse train signal (10) in said temporal sections being indicative of Intensity of the light received at the associated photodetector (14). 14. Device (8) for reading a plurality N of photodetectors (14), wherein

N is an integer> 2 and the photodetectors (14) are spaced apart and illuminated simultaneously by an optical pulse (12) such that each of the N photodetectors (14) respond at least approximately simultaneously in response to the optical pulse (12) generates electrical output pulse (16), the device comprising the following elements:

Delay means (18, 34) for delaying at least Nl of the N electrical output pulses (16) by a delay time Tv _, i, i - 1 ... N,

Means (22, 34) for combining the N delayed output pulses (20) to generate a pulse train signal (10) superimposed on the N delayed output pulses (20), the delay means (18, 34) being adapted to control the output electrical pulses (20); 16) in such a way that the contributions (28) of the individual output pulses (16) in the pulse train signal (10) do not substantially overlap in time, and

Means for reading the pulse train signal (10) arranged to associate with each photodetector (14) a temporal portion of the pulse train signal (10) corresponding to the delay time T _v, i of the associated one

Delay means (18, 34) corresponds.

15. Device (8) according to claim 14, wherein the means for reading comprise an analog-to-digital converter (24).

16. The apparatus of claim 14 or 15, wherein the means for reading comprise a preamplifier.

17. Device (8) according to one of claims 14 to 16, wherein the plurality N of photodetectors (14) is formed by silicon photodiodes.

18. Device (8) according to any one of claims 14 to 17, wherein the

Delay means (18, 34) include cables and different ones

Delay times are generated by different cable lengths.

19. Device according to one of claims 14 to 18, in which the photodetectors (14) have reaction times T _R which are <50 ns, preferably <5 ns, particularly preferably <0.1 ns, and / or have limit frequencies f _c , the preferably> 0.007 GHz, preferably> 0.07 GHz, particularly preferably> 3.5 GHz.

20. A spectrometer for determining the spectrum of an optical pulse (12) by means of a plurality N of photodetectors (14), where N is an integer> 2 and the photodetectors (14) spaced from each other and arranged so that they each light of an associated Spectral range of the optical pulse (12), the spectrometer comprising the following elements: a device (8) for reading the plurality N of photodetectors (14) according to one of claims 14 to 19,

Means for determining a size of the pulse train signal (10) in the said, the photodetectors (14) associated with temporal sections to the the

Photodetectors (14) corresponding spectral components to determine, wherein the size of the pulse train signal (10) in said temporal portions is indicative of the intensity of the light received at the associated photodetector (14).