JPS5826855B2 - How to correct conversion accuracy of digital-to-analog converter - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for correcting the conversion accuracy of a digital-to-analog converter (hereinafter referred to as a DA converter) that converts a digital input signal to an analog output signal.

The present invention provides a new and easy modification method for modifying the characteristics of a current addition type DA converter obtained in advance and obtaining high conversion accuracy.

The conversion accuracy of a DA converter is determined by the accuracy of elements such as resistors and transistors used in the circuit, and in order to obtain high conversion accuracy, it is necessary to improve the accuracy of the elements as much as possible.

However, many of the applications in which DA converters are used require higher conversion precision than is possible with conventional technology (such as element precision), so it is necessary to use pre-obtained DA converters after manufacturing. It is common practice to correct and improve accuracy using various methods.

When constructing an integrated DA converter, it is impossible to adjust the accuracy of the elements in the circuit in advance.
Modification of the above-mentioned DA converter after manufacture is a particularly important problem.

Conventionally, methods for this post-manufacturing modification include the so-called laser trimming method, in which a part of the resistor or a part of the metal wiring that short-circuits the resistor is cut with a laser;
Zener diodes are connected to various positions in the circuit, and if necessary, the Zener diodes are short-circuited and destroyed by applying a large current to switch the circuit (this is the so-called Zener zapping method that corrects accuracy). Are known.

However, the laser trimming method has drawbacks such as thermal distortion that causes instability, leaving a thermally melted solidified region on a part of the resistor or wiring, and complicating the trimming device. - The Zaraping method has the disadvantage of increasing the number of extra circuits for correction and reducing the integration density of the DA converter.

Therefore, the present invention aims to improve the above-mentioned drawbacks of the prior art, and its object is to provide a method for correcting the conversion accuracy of a DA converter that can be easily implemented with a simple device.

The present invention provides a means for improving the conversion accuracy of a so-called current addition type DA converter. A DA converter that converts a digital input signal into an analog output signal by summing on a common summing line through current switches driven by the digital input signal.

The basic idea of the present invention is to reduce the magnitude of the current appearing on the summing line by passing a reverse breakdown current between the emitter-base junction of the transistor constituting the current switch and reducing the current amplification factor of the transistor. The purpose is to modify the conversion accuracy by changing it.

A detailed explanation will be given below using diagram examples. FIG. 1 shows a detailed illustration of the invention.

In the figure, ■ is the current of the current source, Ql and Q2 are transistors forming the current switch, AB is the addition line, and I
Out indicates the current drawn from the addition line AB by the current switch, Vin indicates the digital input voltage, and Vref indicates the reference voltage.

When a digital input voltage that makes transistor Q conductive is applied, the current T out drawn by this current switch from addition line AB is expressed by the following equation, where hFE is the current amplification factor of Ql.

In order to adjust the current value that the current switch draws from the summing line AB, in the present invention, a reverse breakdown current is passed between the emitter and the base of Ql.

It is well known that when a reverse breakdown current is passed between the emitter and base of a transistor, the current amplification factor of the transistor decreases.

In FIG. 1, for example, Vin is set to be more negative in potential than Vref, and Vref y Q2- Ql s Vin
The current can be passed along the following path.

By this operation, the current amplification factor of Ql becomes hFE(1-δ)
, the current I out' drawn by the current switch from the addition line AB is expressed by the following equation when δ(1).

I out'-I out (1-δ/hFB), ie the magnitude of the current appearing on the summing line can be varied.

This amount of change is minute compared to the amount of change in the current amplification factor, allowing for highly accurate correction.

For example, if the value of the current amplification factor of the transistor is 10 before correction,
If it is 0, the output current value changes by 0.1% in response to an operation that reduces the value of the current amplification factor by 10%.

As an example of the 4-bit DA converter shown in FIG. 2, the conversion accuracy correction based on the present invention will be explained in more detail.

In the figure, I, 2I, 4I, and 8I each indicate the current of the current source with weighted magnitude.

■ is the least significant bit current, 2■ and 4■ are currents that are twice and four times as large as the least significant bit, respectively, and 8I is the current for the most significant bit that is eight times as large as the least significant bit. .

From Ql to Q8 are transistors that constitute a current switch, AB is a common addition line, ■out is the output current, and V
ref indicates a reference voltage, and bl to b4 indicate digital input terminals for each bit.

In the DA converter shown in Figure 2, if the current of the current source for the most significant bit were larger than the width that would provide acceptable conversion accuracy due to variations in the manufacturing process of the DA converter. In this case, the digital input terminal b of the relevant bit
4 is more negative in potential than Vref, ■ref, Q8
．． Q7. Conversion accuracy can be improved by passing a reverse breakdown current between the emitter and base of Q7 through the forward path of b4, reducing the current amplification factor of Q7 and reducing the current drawn from the summing line by the current switch of the bit. can.

Conversely, if the current of the current source for the most significant bit is too small, the conversion accuracy can be improved by performing the above operation on all bits other than the most significant bit.

Even if there is an error in bits other than the most significant bit, the conversion accuracy can be corrected by the same operation as above.

FIG. 3 shows an example of a current switch having another structure.

■ and ■ in the figure. is the current of the current source, Ql to Q4 are transistors forming a current switch, R1 and R2 are resistors, AB is an addition line, Vref is a reference voltage, ①o is a bias voltage, and Vin is a digital input voltage.

To change the current value that this current switch draws from the summing line, set vin to ■.

V □ y R2'Q2 +
Ql + Q3 v Direct current passing through Vin
The current amplification factor of Ql may be reduced by flowing a reverse breakdown current between the emitter and base of the transistor.

Based on the present invention, the method for correcting the conversion accuracy of a DA converter is applicable not only to DA converters using current switches shown in FIGS. 1 and 3, but also to DA converters having current switches of other structures.
Needless to say, it can also be applied to converters.

Next, specific examples of the present invention are shown.

FIG. 4 shows a circuit diagram of a current adding type monolithic 8-bit DA converter constructed in accordance with the present invention.

In the figure, R and 2R are reference resistances for generating weighted current, and are composed of diffusion layer resistances. 1. R=
5012R=IK,Q.

Qo to Q8 indicate transistors forming a constant current source.

The emitter area of the transistor Q1 corresponding to the smallest bit is 8μ2, and the transistor for the most significant bit has a structure in which a transistor of the same size as Ql is multiplied by the size of the bit. Q8
has a magnitude 128 times that of Ql.

The same applies to the current switch transistors from Qll to Q82.

bl to b8 are digital input terminals, OP is an output amplifier that converts current output to voltage output, R6 is a feedback resistor,
(2) E is the power supply voltage, VB is the bias voltage that determines the operating current, (2) ref is the reference voltage that determines the logic level of the digital input, and vout is the output voltage.

If the above-mentioned DA converter is manufactured with attention to process uniformity, the D/A converter has an accuracy of 8 bits depending on the chip selection.
An A converter is obtained, but at the same time a DA converter without 8-bit precision is also obtained.

According to the conventional concept, it is not possible to improve the accuracy of the latter.

However, by applying the present invention, this accuracy can be improved.

Table 1 shows an example of the linear error of the above-mentioned DA converter which does not have 8-bit accuracy, in correspondence with the digital input signal.

The linear error was calculated using the output voltage at full scale as a reference, and taking the value of 1/255th of this as the theoretical magnitude of the least significant bit (LSB).

Note that the operation was performed with the output current at full scale being 1 mA.

As shown in the figure, the size of the most significant bit of the DA converter is larger than normal, so before the conversion accuracy is corrected, the linear error of the conversion may deviate from the range of ±015LSB, resulting in an 8-bit error. did not possess the same accuracy.

Next, apply an appropriate negative voltage to the maximum bit input terminal of the DA converter, b8 shown in FIG. When the conversion accuracy correction method according to the invention was applied, the linear error was ±0.5L as shown in Table 1.
It was possible to obtain a DA converter that was corrected within SB and had an accuracy of 8 bits.

In this case, it was appropriate to apply a current of 10 to 50 mA between the emitter and the base for a short time as a reverse breakdown current.

In the above example, the accuracy correction method of the present invention is applied to a DA converter in which the current switch is separated from the constant current circuit. It is also applied to a DA converter configured as a common unit.

FIG. 5 shows an example of a DA converter belonging to the latter category.

The illustrated example has a 4-bit configuration, with Ql to Q4 representing transistors that constitute a constant current source and also performing a current switching operation, and Dl to D4 representing diodes.

R, 2R, 4R and 8R are resistors that constitute a weighted constant current source, AB is an addition line, bl to b4 are digital input terminals, E is a power supply voltage, VB is a bias voltage,
■out indicates the output current.

For example, when the DA converter operates, the conversion accuracy is low because the current value of the constant current source of the most significant bit consisting of Q4 and R is excessive, and the correction method of the present invention can be applied as follows.

That is, the potential of digital input terminal b4 is made more positive than vB, and b4. D4. Q4 through the route of Q4゜vB
The conversion accuracy can be corrected by passing a reverse breakdown current across the emitter-base junction of Q4 to lower the current amplification factor of Q4 and lowering the value of the current flowing from the summing line.

As described above, the present invention provides a new and easy correction method for correcting the conversion accuracy of a current addition type DA converter, and makes a significant contribution to improving the accuracy of integrated DA converters. This also makes a great contribution to improving the manufacturing yield of many integrated circuits that include internal DA conversion systems.

[Brief explanation of drawings]

Figure 1 is a current switching circuit for explaining the principle of the present invention, Figure 2 is an example of a 4-bit DA converter, Figure 3 is an example of another current switching circuit, and Figure 4 is a current summing type monolithic 8-pin circuit. ) Example of a DA converter, FIG. 5 shows another example of the structure of a DA converter. Q1 to Q8: transistor, ■in, b1 to b8: digital input signal, AB: addition line.

Claims (1)

[Claims] 1. Adding a plurality of magnitude-weighted currents to each other on a common summing line via current switches each driven by a digital input signal. - In a method for correcting the conversion accuracy of an analog converter, a reverse breakdown current is passed between the emitter-base junction of the transistor constituting the current switch to reduce the current amplification factor of the transistor, so that the current amplification factor of the transistor is reduced. A method for correcting conversion accuracy of a digital-to-analog converter, the method comprising changing the magnitude of a current appearing in the converter to correct the conversion accuracy.