JPS58154928A - Method for inspecting digital-analog converter - Google Patents

Abstract

PURPOSE:To attain the discrimination with less sampling number of times, by discriminating whether or not the difference of output voltages is within a prescribed permissible error range at a comparator by taking a voltage corresponding to an LSB-bit as a center at an output side converting value of a D-A converter. CONSTITUTION:In Figure, 1 is a D-A converter to be measured, one end of sample hold switches SWa, SWb is connected to an output terminal 1b, holding capacitors C1, C2 are connected between the other end of each switch and ground, switches SWc, SWd to discharge charges of the capacitors C1, C2 are connected, and the lines are connected to input terminals of a differential amplifier 2. An output terminal of the differential amplifier 2 is connected to a window type comparator 3. The set value of the comparator 3 is set within the permissible error range in advance, allowing to discriminate whether or not the output signal is within the error range.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for testing digital-to-analog converters (hereinafter referred to as DA converters) used in computers, PCM adapters, digital recording disc players, and the like.

Conventionally, the conventional method for inspecting a DA converter is to compare the digital input of the DA converter and the analog output corresponding to the digital input. In other words, a continuous digital signal (for example, 00 in the case of a 4-bit D-A converter) is input to the digital signal input side.
00.0001゜0010,...-,1111)
The conventional method was to input the voltage and compare the output voltage obtained from the analog output side with the output of the Miller integrator circuit, but with this method, the measurement limit was determined by the accuracy of the integrator. In addition, there was a problem in that the limit of the measured distortion was also determined by the linearity of the Miller integrator.

In addition, the maximum output of the D-A converter and the maximum output of the Miller integrator must be adjusted, and all possible patterns of the D-A converter must be sampled. Many D-A
When it comes to converters, the number of samplings becomes extremely large, and the time required for inspection also becomes quite long (for example,
A 6-bit D-A converter has various problems such as 2''-1 sampling, or 65,535 samplings.

The present invention has been made in order to solve the above-mentioned problems, and it is possible to check whether the nonlinearity of the D-A converter and the error of the output signal with respect to the input signal is within the allowable error range using an extremely small number of sampling times. The present invention provides a method for testing a new -A converter, which enables the determination of whether the converter is new or not.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, a method for inspecting a D-A collet according to the present invention will be explained in detail with reference to the drawings.

FIG. 1 shows a D-A test circuit according to the present invention. In the figure, 1 indicates the D-A to be measured:
In the J converter, 1a is its input terminal and 1b is its output terminal. One end of sample/hold switches SWa + SWb is connected to the output terminal 1b of the D-A converter 1, and hold capacitors C1 and Ct are connected between the other end of each switch and the ground, respectively. Switches SWc and SWd for discharging the charges in the capacitors C, , C, are connected, respectively, and these lines are also connected to the input terminals of the differential amplifier 2, respectively. The output terminal of the differential amplifier 2 is connected to the input terminal of the window type comparator 3. Therefore, by setting the set value of the window comparator 3 in advance within the permissible error range, it is possible to determine whether the output signal is within the permissible error range.

The D-A converter testing circuit of the present invention is constructed as described above, and the testing method thereof will now be explained step by step.

Assuming that the code of the digital signal handled by the D-A converter to be measured is an offset binary signal that is symmetrical in positive and negative directions around 0, the relationship between the analog output voltage Ao and the digital signal input Di is as follows. As shown in Figure 2, there is a direct proportional relationship.

On the other hand, a system for converting a digital signal into an analog signal in a DA converter is shown in FIG. That is, for example, 16 bits of D
-A converter, 16 bits Bit
Each bit is provided with a switch SW1-SWI6, and one end of the switch SW1-5W16 is connected to a current source 81-S, respectively.
6 are connected, and the other ends of each of the switches SWI to SWI 6 are connected to the input terminal of the current-voltage converter IV. The current-to-voltage converter is, for example, OP as shown.
It consists of an amplifier etc. and an anodized voltage of t is obtained at the output of J+,'l.

When a digital signal is input to Bit1 to Bit16, the switch corresponding to the bit to which the signal is input is closed, and a current source corresponding to that bit is connected to the input side of the current-voltage converter IV. A current flows into the circuit, and by converting the current into a voltage, an analog output voltage Ao corresponding to a digital input signal Di is obtained.

Now, based on the above configuration, each bit Bitl to Bi
Considering the weighting of t16, the most weighted bit MSB (Most 51gn1-
If the full scale value of the analog output AO is Fs, then Bit 1, which is the ficant Bit), is 1/2
Fs/2 of LSB
(Least Significant Bit) Bit 16 is 1/1 of the full scale of analog output AO.
2". Looking at the upper bits from the LSB side, as is clear from the proportional relationship between the digital input and analog output, the output is always exactly twice the analog output voltage. and D
- The A converter responds to the digital input signal from Bitl to
Among Bit16, any output voltage AO can be synthesized by adding the analog output voltages of the corresponding bits (that is, in 16 bits, 2"@-1 type of voltage can be synthesized. Next, the inspection method of the present invention will be explained.

In the circuit of FIG. 1, the switch SWc is installed in advance.

After closing SWd and discharging the charges in the capacitors C, C2, the switches SWc and SWd are opened again.

In this state, switch SWb is opened and switch S
After closing Wa and supplying digital input signals 1000...00 to the D-A converter 1, the switch SWa was opened, and the analog output voltage at that time was held by the capacitor C2, and the switch SWa was opened. In this state, switch SWb is closed and D-A converter 1
After supplying digital input signals 0111...11 to
The switch SWb is opened and the analog output voltage at that time is held by the capacitor C2. As shown in Figure 4, when the D-A cosiparter exhibits ideal characteristics, the output voltage is always exactly 1/2 when looking at the bits from the MSB side to the LSB side. explained with. Therefore, each digital power 1000 charged in capacitors C, C2...
A comparison of the analog output voltage relationships corresponding to 00, 0111, . . . , 11 is as shown in FIG.

In other words, the 1000 charged in capacitor C7
...The analog voltage AO corresponding to the digital input signal of 00 is the voltage Fs/2 which is 1/2 of the analog output voltage full scale value R, and the analog output voltage Ao corresponding to the digital human input signal of 0111...11 is as described above. 1000...
・From the case of a digital input signal of 00, B becomes the LSB
It is reduced by the voltage corresponding to it16.

Each of the analog output voltages described above is input to the differential amplifier 2 with a gain of A, multiplied by the gain A, and output. That is, L.S.
A voltage that is A times the voltage corresponding to Bit 16, which is B, is obtained. This output voltage is then input to the window comparator 3, and it is determined whether the voltage is within a predetermined tolerance range. That is, if the voltage within the allowable error range is V, the characteristics of the window comparator 3 are as follows: LSf3-/4- AV≦Vi≦LSB-A + ,
al/ΔV: Tolerance vi: When the input voltage of the window comparator is, the output of the window comparator 3 is positive;
VB > LSB-A + AV"IA < LSB-
A-'V! In this case, it is only necessary to make the output of the window comparator 3 negative, and the output of the differential amplifier 2 is within ±ΔV, which is within the tolerance range for the voltage of LSB-A, which is the gain of LSB multiplied by A. If it is, the output of the window comparator 3 becomes positive, indicating that the above-mentioned comparison result of each digital signal is a good product.

Further, the allowable error range ΔV can be considered to represent a selection range, and when converted to the input side of the differential amplifier 2, it becomes a voltage of ΔV/A.

This relationship is illustrated in FIG. 6.

In the figure, the vertical axis represents the analog output voltage Ao of the DA converter, and the horizontal axis represents the digital input signal.

In the figure, the straight line drawn as a solid line indicates the characteristic when the input/output characteristics of the D-A converter are ideal, and the dotted lines drawn in parallel above and below the straight line indicate the allowable error limit. That is,
If the output of the D-A converter falls within the area narrowed by this dotted line, it indicates that it is within the allowable error range of △V/A, and the D-A converter is determined to be a good product. be able to.

Conversely, if the output of the D-A converter contains an error of more than ±ΔV/A and takes a value outside the area surrounded by the dotted lines in Figure 6, it is determined that the D-A converter is defective. It is something that can be done.

That is, if the output of the window comparator 3 is positive, the error of the D-A converter is within ΔV/A and it is a good product, and if the output of the window comparator 3 is negative, it is ΔΔ
It contains an error greater than V/A, and it can be determined that the D-A converter is defective.

The above is the inspection method of the present invention, but this alone is not complete. That is, in FIG. 5, for example, in the output voltage between Bit 2 and Bit 16 below MSB, Bit 2 is Bit 2 + ΔX, Bit 3 is Bit 3
-ΔX (ΔX is a constant), even if it includes an error of ΔX, when they are added together, ΔX cancels out and the error ΔX
may be overlooked.

Therefore, in the present invention, not only the comparison between Bitl and 13it2 to Bit16, which are the MSB bits, as described above, but also the comparison of Bit2 and Bit3 to Bit16 as shown in FIG.
Comparison of t16, and furthermore, as shown in Figure 8, Bit3 and B
By comparing and measuring 15 times, sequentially going to the lower bits, such as comparing bits 4 to 16, and comparing bits 15 and 6 (LSB),
This prevents measurement errors such as those mentioned above. And in any measurement, ideally the LSB bit 13it1
Since an output voltage with a difference of 6 minutes should be obtained from the differential amplifier 2, it is possible to measure it equally in all cases. l According to the above method, if a 16-pin D-A converter is used, the circuit shown in Fig. 1 will perform 30 samplings and 1
All bits can be tested with only 5 comparisons. In addition, in general, for an n-bit D-A converter, sampling is performed 2(n-1) times.
- This can be done by comparing the output voltages once.

Next, the above-mentioned measuring method is illustrated in sequence as shown in FIG. 9. That is, if the vertical axis is the output voltage AO of the DA converter and the horizontal axis is the digital human input signal Di, then the first
The digital input signal of 1000...00 and 011
If the result of comparing 1...11 digital input signals is good, then 1111 which is the sum of both digital signals
...It can be predicted that the analog output for the 11 digital input signals is also correct. That is, in the first measurement, it can be determined that the relationship between the person and the output is good at the position indicated by a in the figure. The analog output voltages at this time are O and FS, respectively.

If the result of the second comparison test of each analog output signal with respect to the digital input signals 0100...00 and 0011...11 is good, then the digital input signal 1100...00 The analog output voltage for can also be predicted to be correct. If this is shown in the figure, as shown by b, a digital input is made.

Furthermore, the third time 0010...oo and 0001...
Each analog output signal 4 for 11 digit input signals
If the result of the Fs comparison test is good and there is one, then the first
1010...oo due to balance with 000...00
It can be predicted that the analog output voltages Fs FS 3 for 1, 1, 2 00, 1110, . This point is indicated by C in the figure.

By sequentially comparing each bit in this manner, in 16 bits, all bits can be checked in 15 comparisons.

As described above, according to the first method of inspecting a D-A converter of the present invention, if the conventional D-A converter is n bits, To check the analog output voltage, 2n-
Only 2 samples were required compared to 1.
Since the test can be performed with (n-1) times of sampling and (n-1) times of voltage comparison, the test circuit and test process can be significantly simplified, and the test time can also be significantly shortened. The measurement accuracy can be improved since the test can be carried out after the temperature has reached a steady state. According to the inspection method of the present invention, variations in the output voltage of the D-A converter are essentially irrelevant, so that the accuracy of the Miller integrator that occurs when comparing the output voltage of the Miller integrator as in the conventional method is Constraints that measurement limits are determined and D-A
This has many advantages, such as eliminating the need to adjust the output voltage of the converter and the gain of the Miller integrator every time a test is performed.

[Brief explanation of the drawing]

Fig. 1 shows a test circuit for the D-A converter of the present invention, Fig. 2 is a characteristic diagram showing the relationship between input and output of the D-A converter, Fig. 3 is a diagram of the operating principle of the D-A converter, and Fig. Figure 4 is an explanatory diagram showing the weighting of each bit of the D-A converter,
FIG. 5 is an explanatory diagram showing the first comparative measurement in the D-A converter testing method of the present invention, FIG. 6 is an explanatory diagram showing the allowable error range in the input/output characteristics of the D-A converter, and FIG. , FIG. 8 is an explanatory diagram showing the second and third comparative measurements, respectively, and FIG. 9 is an explanatory diagram showing a state in which each comparative measurement is performed sequentially in the inspection method of the present invention. Explanation of symbols 1... D-A converter to be measured 2...
・Differential amplifier 3, ...window comparator Bitl~Bi
t16... Name of patent applicant for each pit of D-A converter Aiwa Co., Ltd. Figure 4 0

Claims (1)

[Claims] At least L of an n (n is an integer) bit D-A converter
For each n-1 bit except the SB bit, sample the case where an input signal is given only to that bit and the case where an input signal is given to all bits lower than that bit, and calculate the two cases. The output voltages of the DA converter are compared using a differential amplifier, and it is determined whether the difference between the output voltages is within a predetermined tolerance range centered on the voltage corresponding to the LSB bit in the output side conversion value of the DA converter. A method for testing a DA converter, in which the quality of the DA converter is determined by making a determination using a comparator.