EP3888247A1

EP3888247A1 - Analogue-to-digital converter

Info

Publication number: EP3888247A1
Application number: EP19812963.7A
Authority: EP
Inventors: Ralf Beuschel
Original assignee: Ibeo Automotive Systems GmbH
Current assignee: Microvision Inc
Priority date: 2018-11-30
Filing date: 2019-11-27
Publication date: 2021-10-06
Also published as: JP2022510177A; US20220029633A1; US11984908B2; CN113169741A; CN113169741B; WO2020109378A1; DE102018220688A1; CA3122867A1; JP7365719B2; IL283479A; KR102631502B1; KR20210089735A

Abstract

The invention relates to an analogue-to-digital converter (1), comprising: an analogue input for receiving an analogue signal; a first time-to-digital converter (7); and a histogram block (10), wherein the first time-to-digital converter (7) samples the analogue signal based on a ramp signal and delivers an output (20, 25) to the histogram block (10), which creates a time-correlated histogram (21, 26, 30) on the basis thereof.

Description

Analog-to-digital converter

The present invention relates generally to an analog-to-digital converter.

Various methods for optical transit time measurement are generally known, which can be based on the so-called time-of-flight principle, in which the transit time of a light signal emitted and reflected by an object is measured in order to determine the distance to the object on the basis of the transit time.

In particular in the automotive environment, sensors are used which are based on the so-called LIDAR principle (Light Detection and Ranging), in which pulses are periodically emitted and the reflected pulses are detected to scan the surroundings. A corresponding method and a device are known for example from WO 2017/081294.

For time-of-flight measurements, especially LIDAR measurements, it is

required to monitor different analog signals and sample them at a high sampling rate. For example, the analog signals are from one

Photodiode can be emitted for light measurements, to be scanned or for that

Monitoring the current and / or voltage signals of a laser or a laser diode.

Even if solutions for the sampling of analog signals are known from the prior art, it is an object of the present invention to provide an analog-to-digital converter.

This task is solved by the analog-digital converter according to claim 1.

In a first aspect, the present invention provides an analog-to-digital converter comprising:

an analog input for receiving an analog signal;

a first time-to-digital converter; and a histogram block, the first time-to-digital converter sampling the analog signal based on a ramp signal and providing an output to the histogram block, which generates a time-correlated histogram based thereon.

Further advantageous embodiments of the invention result from the

Subclaims, the drawings and the following description of preferred exemplary embodiments of the present invention.

As mentioned, some embodiments relate to an analog-to-digital converter

(hereinafter AD converter) ready, comprehensive:

an analog input for receiving an analog signal;

a first time-to-digital converter; and

a histogram block, the first time-to-digital converter sampling the analog signal based on a ramp signal and providing an output to the histogram block, which generates a time-correlated histogram based thereon.

As mentioned at the beginning, it is particularly necessary for LIDAR measurements to monitor different analog signals and to sample them at a high sampling rate. For example, the analog signals that are emitted by a photodiode for light measurements are to be sampled or for monitoring the current and / or voltage signals of a laser or a laser diode and accordingly the AD converter is used for the AD conversion in some exemplary embodiments analog signals from a photodiode or the voltage or current from a laser (diode) or the like and it can be used in a

corresponding device may be provided, for. B. a device for LI DAR measurements or the like, which is used for example in the automotive environment, without the present invention being limited to these cases. Consequently, some exemplary embodiments also relate to a device with a detector or sensor, for example based on SPAD (Single Avalanche Photo Diode) technology, CAPD (Current Assisted Photo Diode) technology, CMOS

(Complementary Metal Oxide Semiconductor) technology or the like, for detecting light pulses which are emitted by a light source (eg laser) and reflected by an object, in which device the AD converter the invention can be used. Furthermore, such a device can be set up to determine the transit time of the emitted light pulses and, based on this, to determine, for example, the distance between the device and the object, a three-dimensional image of the object or the like. In some exemplary embodiments, the determination of the distance is based on the so-called TCSPC (time correlated single photon couting) measuring principle,

especially in embodiments based on LI DAR. The

The devices, devices or AD converters described can also be used in an autonomously operated (motor) vehicle.

Laser current pulses, in particular for LIDAR measurements, can be in the range between two to ten nanoseconds and consequently, AD converters with a frequency of 1 GHz to 5 GHz would be required, such AD converters typically being expensive and having a high consumption (e.g. .> 500 mW).

In contrast, fast time-to-digital converters are generally known and can, for example, have a time resolution of better than 500 picoseconds and the first time-to-digital converter (also called “TDC”, time-to-digital converter) can accordingly based on such a known TDC. As a result, in some exemplary embodiments, analog signals, such as the current signal or voltage signal from a light source, can be digitized inexpensively and with high time resolution.

The first TDC provides a corresponding output based on the analog signal, the output typically containing time-correlated digital values that characterize the analog signal by using the ramp signal for sampling.

The histogram block creates a time-correlated histogram, the bins of which relate to a start time and consequently represent the time interval from the start time, the respective values of the output of the first TDC being filled into each bin. Accordingly, in some embodiments, the amplitude of a periodic input signal (analog signal) can be compared to the ramp signal and the shape of the periodic input signal can be sampled sequentially in several sampling cycles.

In some embodiments, the histogram block is configured to correct the time-correlated histogram in order to reduce the effect of a time jitter. The time jitter can cause values in the output of the first TDC to be correlated with "wrong" ones (e.g. due to time or late). The histogram block can at least partially correct such temporal shifts and / or the associated too high or low values in the time-correlated histogram, wherein the correction here does not necessarily mean a complete correction in the sense that the effects of a time jitter can be completely compensated for, but rather it also includes a partial correction in which, for example, the effects of a time jitter are at least partially

are mitigated.

In some embodiments, the time jitter effect results

Time-shifted values in the output of the first time-to-digital converter and such effects can be at least partially mitigated.

In some embodiments, the histogram block combines values from different outputs. The first TDC provides sequential outputs, for example, and the histogram block can contain values from such different,

combine consecutive expenses.

For example, the values can be combined so that the maximum value of an output is filled in the time-correlated histogram. If, for example, a second output for the same bin of the time-correlated histogram contains a higher value than the one already existing and from a first output, the existing value is replaced by the higher value of the second output. However, the values can also be combined in such a way that an average value is filled into the time-correlated histogram, for example the average of a value from a previous output and the value of a current output for a specific bin of the time-correlated histogram.

In some embodiments, the values are combined based on a function, the function may be non-linear, which in principle in some embodiments results in a finer or more fine-tuned or

adjusted corrections of time jitter effects. The function can also depend on the difference in the values, so that, for example, there is a greater correction in the case of larger differences between different outputs.

In some embodiments, the values in the time-correlated histogram are shifted by at least one time instance. This can (at least partially) compensate for shifts in sample values caused by time jitter.

In some embodiments, the AD converter further includes one

Start signal input for receiving a periodic start signal. The periodic start signal can, for example, be generated by a pulse generator and delivered to the AD converter. The periodic start signal can also be supplied to the (first) TDC and / or to the histogram block, which then the

Also use the start signal to start a measuring cycle.

In some embodiments, the AD converter further includes one

Ramp generator that generates the ramp signal. The ramp generator has a ramp counter, for example, which generates a counter value based on the periodic start signal. In some exemplary embodiments, the AD converter (or the ramp generator) further comprises a digital-to-analog converter, which generates the ramp signal based on the counter value. In some embodiments, the AD converter further includes a comparator that compares the ramp signal and the analog signal and outputs a comparator signal to the first time-to-digital converter.

In some embodiments, the AD converter further includes a second time-to-digital converter, the first time-to-digital converter capturing time intervals at which the analog signal is above the ramp signal and the second time-to-digital converter capturing time intervals at which the analog signal is below the ramp signal. The corresponding outputs of the first and second TDC are delivered to the histogram block, which based on that

generates a time-correlated histogram, which can then be further evaluated, the histogram block, for example, outputting a waveform that it generates based on the time-correlated histogram, for example by using a Gaussian function, a

Adapt sine function or the like to the course of the time-correlated histogram.

The method steps explained above or herein can also be the subject of a method for operating an AD converter, in particular as disclosed herein.

Some embodiments also relate to a (computer) program that

Receives instructions that, when executed on a processor or computer, result in the method described herein being carried out.

Some embodiments also relate to a computer readable medium that receives a program or instructions that, when executed on a processor or computer, result in the program or method described herein being executed.

Exemplary embodiments of the invention will now be described by way of example and with reference to the accompanying drawing, in which: 1 illustrates a circuit diagram of an embodiment of an analog-to-digital converter; and

2 shows an output of a TDC and a TC histogram;

3 shows an output of a TDC and a TC flistogram under the influence of time jitter;

Figure 4 shows an output of a TDC and a TC histogram under the influence of time jitter, the TC histogram being corrected;

5 shows TC histograms with (center) and without (left) influence of time jitter and shows a corrected TC histogram (right); and

6 illustrates a circuit diagram of an exemplary embodiment of an analog-digital converter with a preactivation signal.

Fig. 1 illustrates a circuit diagram of an embodiment of an analog-digital converter 1, hereinafter referred to as AD converter 1.

In principle, as also explained above, the AD converter 1 can be used for all signals that are periodic and repeat themselves at least twice without being changed.

In the present exemplary embodiments, it is assumed without restricting the invention to the fact that the AD converter 1 is used in a LI DAR measuring system which uses the TCSPC measuring principle. The laser pulse used or the laser pulse sequence can be emitted periodically and at a high frequency, e.g. B. every two microseconds for a 300 meter range.

The basic functioning of the AD converter is based on the fact that a periodic input signal is compared with a ramp signal. The ramp signal is relatively slow and can be synchronous or asynchronous with the periodic input signal cycle. TDCs (Time-to-Digital Converter) measure the points in time at which the ramp signal is crossed for each period and the measurements are written to a time-correlated histogram block, as is also explained in more detail below:

The AD converter 1 has an analog input 2, at which the analog signal to be converted is input, and it has a start signal input 3, at which a periodic start signal or pulse signal is present, which, for example

Pulse generator comes and is also used for the generation of light pulses for the LIDAR measuring system.

Furthermore, the AD converter 1 has a comparator 4 in order to compare the analog signal which is received at the analog input 2 with an analog ramp signal.

The analog ramp signal is in the form of a rising sawtooth in this embodiment, starts at approximately zero volts, and rises to one

predetermined maximum value. Here the rise time of the ramp signal is based on a fixed multiple of the pulse generator frequency and is e.g.

framp = 1/128 * fpulse to scan with an effective resolution of 7 bits, where "framp" is the frequency of the ramp signal and "fpulse" is the frequency of the

Pulse signal, which is received via start signal input 3.

The ramp signal is generated via a ramp counter 5 which is coupled to a digital-to-analog converter 6 (DA converter hereinafter referred to).

The ramp counter 5 increments its counter by one with each received start pulse (that is, when a new sampling cycle is started) that is received via the start signal input 3. The binary value obtained is supplied to the DA converter 6, which generates a corresponding ramp signal from the binary value, which consequently also has a higher ramp voltage or a higher ramp threshold with increasing binary value.

The output of the comparator 4 is on the one hand (directly) coupled to a first TDC 7 and on the other hand (indirectly) coupled to a second TDC 8, the comparator signal first passing through an inverter 9 which inverts the comparator signal and then supplies it to the second TDC 8 .

The start pulse is also supplied from the start signal input 3 to the first TDC 7 and to the second TDC 8, so that the "start" of the signal pulse starts a measurement cycle.

The first TDC 7 measures the time instances (or time intervals) when the analog input signal coming from the analog input 2 rises above the ramp signal, that is to say it crosses upwards, with respect to the start signal which is received by the start signal input 3.

The second TDC 8 receives the inverted comparator signal and measures it

Time instances (or time intervals) when the analog input signal coming from analog input 2 falls below the ramp signal, ie crosses it down, with respect to the start signal received by start signal input 3.

The first 7 and the second TDC 8 each output their measurement results (outputs) to a histogram block 10.

This measuring process continues until the ramp counter 5 reaches a predetermined value and the time-correlated and present in the histogram block 10

Measurement result in histogram block 0 can be evaluated.

The ramp counter 5 also gives a digital counter value (corresponds to one

Sampling cycle count) corresponding to the ramp threshold to histogram block 10, with the ramp threshold initially low and increasing over time. At the beginning of each measurement cycle, all histogram values are in the histogram block

10 initialized with a "0".

The time intervals at which the analog input signal is above the ramp signal are stored in the histogram of the histogram block, with all measurements being aligned with a different ramp voltage at the time “0”, which is determined by the start signal. The analog input signals can have between 0 and N time intervals in each sampling cycle if it exceeds the ramp signal.

Whenever the analog input signal exceeds the ramp signal in a sampling cycle, the associated area of bins of the histogram is filled with the current counter value of the ramp counter 5, smaller values being overwritten with larger values in successive cycles when the analog input signal is above the ramp signal.

As mentioned, when the ramp counter 5 reaches the maximum value, the

Ramp generation is ended and the data in the histogram of the histogram block 10 are ready for the evaluation, being after the evaluation

corresponding waveform output 11 can take place.

The measurement cycle is restarted by resetting the ramp counter 5 and filling the histogram of the histogram block 10 with “0” values.

Such a measurement cycle, which can be carried out with the AD converter 1, is now also described with reference to FIG. 2, which illustrates an output 20 of the TDC 7 of the AD converter 1 above and a TC histogram (TC = time correlated , ie time correlated) 21, as is generated in the histogram block 10.

In this embodiment there are six ramp values, the output 20 of the first TDC 7 on the ordinate having entries for the six ramp values and on the abscissa shows the time. The sampling of the analog input signal by the first TDC 7 results in twelve entries for the output 20.

The TC histogram 21 shows the state after a complete measurement cycle. Again there are six values on the ordinate corresponding to the six

different ramp values and on the abscissa the bins are arranged according to the past time since the start "0".

The bins of the TC histogram 21 are each time the value of

Ramp signal filled when the analog input signal is above the ramp signal, whereas the bins remain unaffected at the time the analog input signal is below the ramp signal. The course of the TC histogram 21 roughly corresponds to a sine or Gaussian curve.

As will be explained in the following with reference to FIG. 3, the sequential scanning can react sensitively to time jitter and with some

1, the histogram block 10 of the AD converter 1 of FIG. 1 is set up accordingly, so that the time-jitter performance is improved.

The time jitter effect is illustrated in FIG. 3, which illustrates an output 25 of the first TDC 7 at the top and the resulting TC histogram 26 at the bottom, as is shown in the histogram block 10 under the influence of an (uncorrected)

Time tremors arise.

For a sample "3" of the output 25, as shown at reference number 25a, a negative sample time jitter occurs, which leads to the sample "3" being sampled too early at 25a and therefore the beginning and end of the associated one TDC output signals are shifted one time instance to the left (earlier).

For a sample "4" of the output 25, as shown at reference number 25b, a positive sample time jitter occurs, which leads to the sample "4" being sampled too late at 25b and therefore the start and end of the associated sample TDC output signals are shifted one time instance to the right (later). By proceeding that the maximum values are entered at the end in the TC histogram 26, the value 26a in the third bin is changed from the value “1” (see FIG. 2, there the third bin has the value “1”) The value "3" is increased, whereas the value 26b in the fifth bin is reduced from the value "4" (FIG. 2) to the value 2 and the value 26c of the third to last bin is changed from the value "1" (FIG. 2) to the value "4" increased.

If one compares the curve of the TC histogram 21 from FIG. 2 with the curve of the TC histogram 26 from FIG. 3, it is noticeable that the TC histogram 26 is less similar to a sinusoidal curve than the TC histogram 21 from FIG Fig. 2.

An exemplary embodiment in which the sensitivity to time jitter is improved is explained below with reference to FIG. 4 by the following measures that the histogram block 10 of the AD converter 1 can provide.

It is assumed that a time instance T is sampled at a level N (ramp threshold), then a new value in the TC histogram (memory) for the time instance T is Fh (T, N).

The following rule (1) is now provided:

If (if) the previous histogram value Fh (T, N-1) = N-1, then set

Fh (T, N) = N, otherwise (ise) set Fh (T, N) = combination (Fh (T, N-1), N).

Furthermore, for the method which was explained with reference to FIG. 2, the values were combined to the maximum, that is to say combination (A, B) = B, in other words, the bin value was always set to the largest value, what can be prone to trembling accordingly.

In the present exemplary embodiment, however, a rule (2) applies, namely

Combination (A, B) = (A + B) / 2, that is, the new value corresponds to the mean of the values A and B, which is also referred to as the “50% combination method”. Alternatively, in other exemplary embodiments, combination (A, B) = truncated ((A + B) / 2) is provided, which corresponds to a shift to the right by one bin, the fraction being discarded.

Furthermore, it can also be provided as an alternative exemplary embodiment that a non-linear mapping function is used, which depends on the difference between B-A, and that the histogram values are corrected accordingly in order to reduce the effect of the time jitter.

FIG. 4 shows again the output 25 of the first TDC 7 with a

corresponding time jitter, as has already been explained in connection with FIG. 3.

The resulting TC histogram 30, which, by applying the above-mentioned measures, namely rules (1) and (2), is shown in FIG. 4 below, the effect of the time jitter being compared to the TC histogram 26 of FIG. 3 is reduced.

The sample value at 30a (see value at 26a in FIG. 3) is reduced from the value “3” (26a in FIG. 3) to the value “2” (30a in FIG. 4) (original is “1”).

The sample value, which corresponds to the value at 26b in FIG. 2, remains at the value “3” and is therefore not changed (original is “4”).

The sample at 30b is reduced from "4" to "3" (original is "4").

The sample value at 30c is reduced from "3.5" (see 26c in Fig. 3) to "2.5" (original is "1").

5 now shows the result of the application of rules (1) and (2) (far right, TC histogram 30) in comparison with the normal (maximum)

Combination procedure (center, TC histogram 26) if there is time jitter and on the far left without time jitter (TC histogram 21). 5 it can be seen that the proposed “50%

Combination method ”(rule (2)) generates a TC histogram 30 which comes closer to the original course of the TC histogram 21 without time jitter than the application of the explained combination method, in which simply the maximum value is taken and which in the case of time jitter is in the middle 5

shown TC flistogram 26 generated.

The maximum combination method shown in the middle of FIG. 5 tends to broaden the shape of the histogram and to produce larger steps in the values between the individual bins, since it represents both the time jitter and the shape of the signal.

The 50% combination method tends to increase the rise and fall times according to the time jitter statistics.

6 illustrates a circuit diagram of a further exemplary embodiment of an analog-digital converter 40, hereinafter referred to as AD converter 40 for short.

In principle, the AD converter 40, as also explained above, can be used with all

Use signals that are periodic and at least twice

repeat without being changed and it essentially corresponds to the AD converter 1 described in detail above.

In some embodiments, it is important to sample the rising signal edge of the input signal. In a LI DAR measuring system, for example, it is important to scan the leading signal edge of the optical laser.

Therefore, in this embodiment, a preactivate signal is provided to initialize the AD conversion before the analog (input) signal is converted. This preactivation signal is sent, for example, 3 ns before the start signal, without restricting the present invention to this example. The AD converter 40 has an analog input 41, at which the analog signal to be converted is input, and it has a start signal input 51, at which a periodic start signal or pulse signal is present, which, for example

In addition, the AD converter 40 has a preactivation signal input 42, to which a preactivation signal is present, which, as mentioned, is sent, for example, 3 ns before the start signal.

Furthermore, the AD converter 40 has a comparator 43 to convert the analog signal received at the analog input 41 to an analog signal

Compare ramp signal.

specified maximum value. Here the rise time of the ramp signal is based on a fixed multiple of the pulse generator frequency and is e.g.

Pulse signal, which is received via the pre-activation signal input 42.

The ramp signal is generated via a ramp counter 44 which is connected to a

Digital-to-analog converter 45 (DA converter referred to below) is coupled.

The ramp counter 44 increments its counter by one with each received preactivation signal received via the preactivation signal input 42. The binary value obtained is supplied to the DA converter 45, which generates a corresponding ramp signal from the binary value, which consequently also produces a higher ramp voltage or a higher ramp value as the binary value increases

Has ramp threshold. The output of the comparator 43 is on the one hand (directly) coupled to a first TDC 46 and on the other hand (indirectly) coupled to a second TDC 47, the comparator signal previously passing through an inverter 48 which inverts the comparator signal and then supplies it to the second TDC 47 .

The pre-activation signal is also supplied from the pre-activation signal input 42 to the first TDC 46 and to the second TDC 47, so that the

Preactivation signal initialized a scan.

The first TDC 46 measures the time instances (or time intervals) when the analog input signal coming from the analog input 41 rises above the ramp signal, ie crosses it up, with respect to the preactivation signal which is received by the preactivation signal input.

The second TDC 47 receives the inverted comparator signal and measures it

Time instances (or time intervals) when the analog input signal coming from the analog input 41 falls below the ramp signal, that is to say crosses it down, with respect to the preactivation signal which is received by the preactivation signal input 42.

The first 46 and the second TDC 47 each output their measurement results (outputs) to a synchronizer 50, which synchronizes the outputs of the TDCs 46 and 47 with the start signal 51.

For each measuring cycle, the synchronizer 50 determines the time interval between the preactivation signal and the start signal by means of

t_diff (SC) = t_Start - t_Voraktivierung, where t_d iff (SC) the time interval between the start signal and preactivation signal measured at the synchronizer 50, t_Start the time of the start signal and t_Voraktivierung the time of the

Pre-activation signal is.

The times t start and t_pre-activation are determined by means of the time t_nominal, which is at least as long as the maximum measured time Distance between the start signal and the pre-activation signal, ie tjiominal> max (t_diff (SC)). The value t_nominal is used as a constant time value in each

Measurement cycle stored in the histogram block 49

The synchronizer 50 synchronizes the measurement values of each measurement cycle by adding a constant time value t_offset (SC) = tjiominal - t diff (SC) to the outputs of the TDCs 46 and 47.

This measuring process continues until the ramp counter 44 reaches a predetermined value and the time-correlated and present in the histogram block 49

Measurement result in histogram block 49 can be evaluated.

The ramp counter 44 also gives a digital counter value corresponding to the

Corresponds to the ramp threshold value to the histogram block 49, the ramp threshold value being low at the beginning and increasing in the course.

At the beginning of each measurement cycle, all histogram values in histogram block 49 are initialized with a “0”.

The time intervals at which the analog input signal is above the ramp signal are stored in the histogram of histogram block 49, all measurements being aligned with a different ramp voltage at the point in time “0”, which is determined by the preactivation signal.

Whenever the analog input signal exceeds the ramp signal, the associated area of bins of the histogram is filled with the current counter value of the ramp counter 5, smaller values with larger values

successive cycles are overwritten if the analog

Input signal is above the ramp signal.

As mentioned, when the ramp counter 44 reaches the maximum value, the

Ramp generation ended and the data in the histogram of the histogram Blocks 49 are ready for the evaluation, and a corresponding waveform output 52 can take place after the evaluation.

The measurement cycle is restarted by resetting the ramp counter 44 and filling the histogram of the histogram block 49 with “0” values.

Reference symbol analog-digital converter

Analog input

Start signal input

Comparator

Ramp counter

DA converter

First TDC

Second TDC

Inverter

Histogram block

Waveform output

Issue of the first TDC 7

TC histogram

Output of the first TDC 7 with time jitter TC histogram

TC histogram

Analog-to-digital converter

Analog input

Preactivation signal input

Comparator

Ramp counter

Digital-to-analog converter 45

First TDC

Second TDC

Inverter

Histogram block

Synchronizer

Start signal input

Waveform output

Claims

1. Analog-digital converter (1), comprising:

an analog input for receiving an analog signal;

a first time-to-digital converter (7); and

a histogram block (10), the first time-to-digital converter (7) sampling the analog signal based on a ramp signal and delivering an output (20, 25) to the histogram block (10), which based thereon a time-correlated one

Histogram (21, 26, 30) created.

2. Analog-to-digital converter according to claim 1, wherein the histogram block (10) is arranged to correct the time-correlated histogram (21, 26, 30) in order to reduce the effect of a time jitter.

3. Analog-digital converter according to claim 2, wherein the effect of the time jitter leads to time-shifted values in the output (25) of the first time-digital converter (7).

4. Analog-digital converter according to one of claims 2 or 3, wherein the histogram block (10) combines values from different outputs (20, 25).

5. Analog-digital converter according to claim 4, wherein the values are combined such that the maximum value of an output is filled in the time-correlated histogram (21, 26, 30).

6. Analog-digital converter according to claim 4, wherein the values are combined such that an average value is filled in the time-correlated histogram (21, 26, 30).

7. Analog-to-digital converter according to one of claims 4 to 6, wherein the values are combined on the basis of a function.

8. Analog-to-digital converter according to claim 7, wherein the function is non-linear.

9. Analog-digital converter according to claim 8, wherein the function of the

Difference of values depends.

10. Analog-digital converter according to one of claims 4 to 9, wherein values in the time-correlated histogram (21, 26, 30) are shifted by at least one time instance.

11. Analog-digital converter according to one of the preceding claims, further comprising a start signal input (3) for receiving a periodic

Start signal.

12. Analog-digital converter according to claim 11, further a ramp counter (5) which is based on the periodic start signal generates a counter value.

13. Analog-digital converter according to claim 12, further comprising a digital-analog converter (6) which generates the ramp signal based on the counter value.

14. Analog-digital converter according to one of the preceding claims, further comprising a comparator (4) which compares the ramp signal and the analog signal and outputs a comparator signal to the first time-digital converter (7).

15. Analog-digital converter according to one of the preceding claims, further comprising a second time-digital converter (8), wherein the first time-digital converter (7) detects time intervals at which the analog signal is above the ramp signal and the second time-digital converter (8) detects time intervals at which the analog signal is below the ramp signal.