JPH0356892A - Method and device for sample time measurement - Google Patents

Abstract

PURPOSE:To measure a sample time without reference to the aperture time, offset voltage, etc., of a DUT by inputting the output of the variable frequency signal generator to a device to be measured and connecting an amplitude measuring instrument to its output. CONSTITUTION:The sample time measuring instrument consists basically of the variable frequency generator 1, DUT 2, and an AC voltmeter 3 and includes a timing generator 4 which supplies a sample instruction to the DUT 2. In this case, even if the phase relation of the sample time to an input signal 5 varies from 6 to 8, the integer period of the input signal 5 is averaged, so an AC component is removed and the output of the DUT 2 is therefore only a DC component, which is inputted to the AC voltmeter. Consequently, the sample time is N/fz, where fz is an input signal frequency at which the output AC component disappears and N is the number of cycles of the input signal in the sample time. For example, when fz is searched fro from a low frequency and obtained, N = 1 and the sample time becomes 1/fz and is easily found.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to a measuring method and apparatus for measuring sample time of a sample hold, a track hold, and an A/D converter incorporating these.

[Prior art and its problems]

In recent years, digital signal processing technology has advanced and analog circuits have been increasingly replaced, but in this case higher performance A/D converters are required, especially the sample hold or trunk hold section (hereinafter simply referred to as sample hold). Detailed evaluation of
The analog signal is accumulated for a period of time (hereinafter referred to as sample time) and its voltage is maintained. Conventionally, the aperture time, which is the time from when a sample command is applied until an analog signal is actually captured, and the aperture jitter, which indicates its uncertainty, have been evaluated, but samples that accumulate analog signals have been evaluated. Time is not valued very much.

This will be explained in more detail with reference to FIG.

In FIG. 5, (a) is the waveform of the sample command signal Vs to the sample hold, (b) is the waveform of the human signal Vi, and (c) is the waveform of the sample command signal Vs to the sample hold.
indicates the waveform of output signal ■0.

The dashed lines common to (a), (b), and (c) in the figures represent substantially the same time.

As for the sample command signal, time L is ta, < t < tag
, ta. Commands to sample when < t < ta4.

In addition to the command of the sample command signal Vs, the human power signal V
tbi < t < tbz for i, and tb:+
A range of time L such that < t < tb4 is sampled.

Therefore, tb2 th+ -tL tb1 is the effective sample time, which is the sample time of the present invention.

More specifically, tb. and tb:l are respectively t
A fraction of the acquisition time is delayed from a and ta=,
tbz and tb4 are delayed from taz and ta4 by an averter time, respectively. Input accumulation (parallelization) time te of output signal Vo. -tc. and tcz-tc2 are equal to tb2-tbl and tb4tb:+, respectively.

The sample time does not affect the input bandwidth of the sample and hold, so it needs to be evaluated. However, when trying to measure it, it is difficult to measure the analog signal synchronized with the sample command, as in test equipment using the conventional locked histogram method. Methods in the time domain were used, such as manually holding samples and searching for points of change in the output.

When measuring sample time using such a locked histogram method, the sample time is often short compared to the period of the analog signal, and it is often difficult to maintain the time accuracy of the analog signal. It is necessary to take measures such as changing the analog signal from a sine wave to a square wave with a higher slew rate.

Furthermore, with this method, there are cases where a drift is observed in the offset voltage or averture time of the sample and hold, the output voltage resolution is limited because the sample and hold and the A/D converter are integrated, and the timing generation of the sample command is performed inside the sample and hold. Measuring the sample time is difficult if detailed external control is not possible.

[Object of the Invention] An object of the present invention is to solve the above-mentioned conventional problems and provide a measuring method and apparatus that can stably measure the sample time of a sample hold.

(Summary of the Invention) The present invention uses a signal generator that can generate a frequency band with an integer period in the sample time, and a measurement device that is protected against alternating current amplitude interference. The output does not appear due to averaging at the human signal frequency where an integer period occurs during the sample time of the integrated hold and A/D converter (hereinafter simply referred to as DOT).The sample time is determined from this human signal frequency. It is something.

[Summary of the Invention] FIG. 1 is a block diagram of a measuring device for measuring sample time of a DUT according to an embodiment of the present invention.

Basically variable frequency generator 1, DUT 2, AC voltmeter 3
In addition to this, it also includes a timing generator 4 that normally gives a sample command to the DUT 2. If the input signal frequency of the DUT 2 is f in , the frequency of the sample command is f clk , and the output signal frequency is f out , then the Nyquist frequency O < f in < f clk / 2 for turning around fclk / 2 as a boundary.
out=fin, fclk/2 <fin
<f clk then f out = f clk − f i
n. ．． f clk < f in < 3 f elk / 2
Then r out = f in - f out , . .
・The relationship is established.

Figure 2 is a timing diagram when the input signal has an integer period during the sample time. Even if the phase relationship of the sampling time with respect to the human input signal 5 changes from 6 to 8, the AC component is removed because the integer period of the input signal 5 is averaged, and as a result, the output of the DUT 2 becomes only the DC component, and the AC voltmeter is man-powered.

In this way, the input signal frequency at which the output AC precept disappears is fz, and the number of cycles of the human signal included in the sample time is N.
Then, the sample time is equal to N/fz. For example, if fz is obtained only by searching from a lower frequency, then N is E, so the sample time is 1/fz, and the sample time can be easily determined.

FIG. 3 is a block diagram of a measurement device for measuring sample time according to another embodiment of the invention. In this block diagram, although the human input signal 5 for the DtJT 2 in FIG. By examining the frequency of the input signal (fz is determined by scanning the frequency of the input signal.

FIG. 4 shows an example of the distribution of the frequency of the signal in FIG. 3. White noise generator 9 and bandpass filter 10
The broadband human input signal 15 created by O ~ f clk/
2, f clk/ 2 ~ f clk, f clk
~3f clk /2 . In Figure 4, f in is f c
lk to 3fclk/2.

If fz exists in the band of the manually generated noise in this way, only fz will be removed from the human signal as explained in FIG.
In the distribution shown in , there appears a part where some part of the output signal like fzl6 disappears, and this f z' and the above f inXf
clk. ．． fz is obtained from the relationship 2 of f out, and the sample time can be calculated in the same manner as in the previous embodiment.

In the above explanation, the signal sources are not limited to sine waves and ft sounds, but any suitable signal including fz, such as multi-tone, may be used. In that case, the frequency of fz can be changed to It may be determined from a human signal source as in the first embodiment, or it may be determined by measuring the output signal of the DUT as in the second embodiment.

In addition, although we have only mentioned an AC voltmeter and a spectrum analyzer as devices for detecting the DUT output, it is sufficient to detect a certain amount of change in AC, so for example, a selective level meter may also be used (if the DUT includes an A/I converter) The output format is digital, and a DSP using FFT or the like may be used.

Furthermore, even in the case of a simple chomba where the OUT does not have a hold function, this device can be used by simply adding a low-pass filter etc. that averages the sample time; for example, if the DUT is an optical chomba. But just by changing the signal to light, you can measure the shirt speed.

[Effects of the Invention] As explained above, according to the present invention, the sample time can be measured without depending on the averter time or offset voltage of the DUT, and therefore, even if other parameters are unknown or not constant. Even if the DUT is like this, it will not be affected. In addition, the procedure for guiding the measurement is simple, so different DU
It is easy to apply to T, measurement time can be shortened, etc.
It has great effects.

[Brief explanation of drawings]

FIG. 1 is a block diagram of a test device according to a basic embodiment of the present invention, FIG. 2 is an example of tie approval in FIG. 1, and FIG. 3 is a block diagram of a test device according to another embodiment of the present invention. , FIG. 4 is an example of the signal distribution in FIG. FIG. 5 is a waveform diagram showing the sample and hold operation. 1: Variable frequency signal generator, 2, 11: DUT, 3: AC voltmeter, 4, 13: Tying generator, 5: DUT2
Human signal, 6, 7, 8: Example of phase relationship between sample time and human signal, 9: White noise generation L 1 0: Bandpass filter, 12: Spectrum analyzer, 14:
Example of frequency component of output signal of DUT11, 15: DUT
Example of frequency distribution of human signal given to 11

Claims (4)

[Claims] (1) A sample time measuring device in which the output of a variable frequency signal generator is input to a device under test, and an amplitude measuring device is connected to the output. (2) The sample time measuring device according to claim 1, wherein the variable frequency signal generator is a multi-frequency signal generator, and the amplitude measuring device is a frequency component analyzer. (3) A variable frequency signal is input to a sampling device under test, the output amplitude of the sampling device under test is observed, and the frequency of the variable frequency signal at which the gain of the sampling device under test becomes minimum is changed from the frequency of the variable frequency signal to the sampling device under test. A sample time measurement method that determines the sample time of a sampling device to be measured. (4) The sample time measuring device according to claim 1 or 2, further comprising an averaging device added to the output of the device under test.