SU1339887A1 - Method of functional tuning of d-a converters - Google Patents

Abstract

The invention relates to the field of information-measuring equipment and microelectronics and can be used in the manufacture of precision digital-to-analogue converters (DACs) in a hybrid-film version. The purpose of the invention is to increase the fit speed. Matching the resistors in each bit of the converter, while monitoring the output voltage, until the control voltage value determined by measuring the output DAC voltage results in a reduction in the number of operations for fitting each bit, which accelerates the fitting process. 5 il. (/) OS OS OS OS “h

Description

 one

The invention relates to the field of information technology and microelectronics and to be used in the manufacture of precision digital-to-analogue converters (| AP) in a hybrid-film version,

The purpose of the invention is to increase the fit speed.

FIG. 1 shows the equivalent digital-to-analog converter circuit at the moment of fitting the i-ro bit; Fig. 2 shows the dependences of the currents flowing through the adjustable resistor and through the two-pole device relative to the magnitude of their imbalance (conversion of these currents into voltage takes place linearly, the current voltage of the converter is similarly dependent - the voltage is indicated when HOI is turned on DAC); in fig. 3 is the equivalent circuit of a quadrupole, where R, s gy are the resistance of an equivalent triangle of the adopted quadrupole model; in fig. 4 - equivalent to the resistor decoding matrix for fitting the second bit of the PAP (the first bit is the highest bit); in fig. 5 is a block diagram of the device, realizing the method of functional fitting of digital-to-analog converters.

The digital-to-analog converter contains a quadrupole 1 formed by an unsupported part of the resistor decoding matrix R-2R, a two-terminal 2 formed by the fitted part of the resistor decoding matrix, the resistor 3 is adjustable by i-ro, the converter 4 is the voltage, the input resistance of which , the source 5 of the reference voltage, the source 8 of the reference voltage, adjusted by the CAC, which includes the decoding resistor matrix 9.1e when the switch 9.2, the scaling resistor 9.3, the operating device 9.4d code sensor 10, the reference DAC 11, the comparator 12, a digital voltmeter 13, a microcomputer 14, a laser control unit 15.

The method of functional adjustment of the D / A converter is implemented as follows.

The adjustment starts with the low-order bits of the DAC Before fitting the i-th discharge to the input of the converter 4 current - the voltage is connected at the beginning one end of the elementary divider,

five

25

98872

called by the discharge resistor 3 and the two-pole 2, then the other, find the currents 1; , Ijj, is left connected to the input of the converter 4 that arm in which a greater current flows. In this case, it is considered that the resistance of the resistor changes only towards its increase.

10 The current flowing through the discharge resistor 3 is greater than the current flowing through the two-port 2 (Fig. 2, points A and A). In this case, it is necessary to adjust (increase the resistance). Resistor 3. The resistance of the two-terminal 2 will be equivalent to the fitting standard for the resistance of the resistor 3, Considering that the resistance of the two-terminal 2, as the reference standard of resistor 3,

,(one)

2;

where C. is the nodal voltage of this

discharge or voltage drop across the resistor 3 and the two-port 2,

find resistance

R ,, Y-- 2R-iR, (2)

- - i

where AR is the absolute difference between the resistance of the two-pole 2 and the resistance of the resistor 3. From the above expressions we find the relative magnitude of the imbalance D;% 35 of the inclusion of the i-th bit and all the previous ones together with the additional

A, |,

(3)

where and. - voltage at the output of the converter 4 when turning on the i-ro discharge;

and 2 is the output voltage of the converter 4 from the inclusion of all the previous low-order bits together with the additional one.

These ratios correspond to FIG. 1 and 2 and determine the case of fitting the discharge resistor 3. If the current flowing through the two-port 2 is greater, then it is necessary to adjust (increase) its resistance by acting on the resistor by a fitting tool.

In this case, formulas (1) - (3) remain valid, and the standard of adjustment (2R) for this digit is resistor 3. Accordingly, the current

thirty

400%

0

1 j flows already through the resistor 3, and the voltage U ,; formed by the inclusion of i-ro bit. From the formula (1) - (3) we can express the resistance R; as

R; (1-f5o) 2R.

The parameters of the quadrupole 1 (FIG. 3) depend on the number of the bit to be adjusted:

.

72 (1-0 +1

“,.

 -n) 2 -1

The obtained relations (5) - (7) make it possible to find the transfer coefficient of the source 5 of the reference voltage to the nodal point (Fig. 1), and therefore, taking into account formulas (1) - (4), we can obtain the following ratio of currents

. Wso.

{J P9 ..

R 2 100- /; four,

and the 100th;

(9)

 R 2 100-p ui where ft; - coefficient depending on

customizable bit numbers

o2i-, 1

  (ten)

Although the quadrupole 1 of the proposed equivalent circuit (Fig. 1) excludes the highest bit, the validity of formulas (8) and (9) is not lost.

Thus, for all without the bits of a DAC, the control value of the current (voltage) I k; (UK;) can be obtained from formulas (8) and (9) in accordance with the schedule (Fig. 2) at.

After that, taking in relation to the control current (voltage) to the current (at the voltage) of the branch to be adjusted according to the formula (8), we obtain

t 2liio2i 3; i; t

 2 100 a for the control voltage taken from the output of the converter 4, this formula can be represented in its final form

2 100-e; -l;

(12)

and.

{N

(13)

2-100 where and; „- the initial voltage from

enable i-ro bit

but

a 59887

i-ro, defined up to cm o prep (

and, - „- for NLP 9 (Fig. A) is the same as for the equivalent circuit (Fig. 1).

As a result of fitting all the bits, taking into account the corresponding (13), the DAC is fitted according to such accuracy parameters as linearity and differential nonlinearity.

After aligning the steps of the static conversion characteristic of the DAC, it is adjusted with respect to the full scale of the conversion to the desired value, thereby changing the transfer coefficient of the A converter current - voltage.

A device that implements the method „„ functional fit, works as follows.

When fitting the i-ro bit, the DAC 9 includes the i-th bit of the adjustable 9 and the reference 11 DAC. Measured by a digital 25 voltmeter 13, the output voltage of a driven AC 9, connect the input of a digital voltmeter 13 to the output of the test DAC 9 via a switch 12. Then through a switch 12 connect the inputs of a digital voltmeter 13 to

35

40

45

50

55

The bridge is formed by adjustable 9 and reference 11 D / A converters, its indication is copied into the memory of microcomputer 14 Then all bits of the adjusted DAC 9, lower i-ros, including the additional, not changeable code at the input of the reference PAP 11, are connected. A new reading of a digital voltmeter 13, the inputs of which are connected via switch 12 to the diagonal of the bridge of the DAC 9 - DAC 11, is recorded in the memory of the microcomputer 14, which then calculates the difference D% and compares the obtained value with the allowable value for the given bit, if and exceeds the allowable value of d this discharge, the microcomputer 14 instructs the resolution adjusting unit 15, controlling an adjusting tool. In this case, the microcomputer 14, using the expression (13), determines the reference value of the voltage of the i-ro discharge.

A piece of i-ro resistive material is removed with a fitting tool.

a discharge resistor (i-ro connection resistor) before the output of the adjusted DAC 9 reaches a calculated control voltage value, and the comparison produces a micro-computer 14.

 i: 3

Upon reaching the output of TsL11 9 of the originally calculated core (the microcomputer's 14 magnitude, the command is given to the Kone laser control unit 15, the fit. After that, the voltage is measured in the diagonal DAY 9 - NAL 11, the value of b% is estimated by microcomputer. With a new value of L% above the norm, a new measurement of the output voltage of the DAC 9 is made and the calculation of the new control voltage, to the value of which the i-ro bit is re-adjusted.

After fitting all the bits of the D / A converter 9, the transfer coefficient is adjusted, affecting the large-scale

resistor adjustable tool

(by laser), with all included bits of the D / A converter 9, including an additional one, until the module reaches the voltage at the output of the PAP 9 value of the reference voltage of source 8,

Claims (1)

Invention Formula . The method of functional adjustment of digital-to-analog converters, according to which sequentially, starting with the lowest bits, forms the i-ro unbalance signal by forming the difference between the first output voltage of the converter, generated when the i-th discharge is turned on, and the second output voltage of the converter, formed when all low-order bits are turned on relative to i-ro, the imbalance signal is compared with the unbalance control value and, if they are unequal, the corresponding resistance is corrected ezistora i-ro discharge dekodid 5 0 five 0 five 0 matrix P) (; the former, then ps again form the 1-bit unbalance signal and compare it with the unbalance control value until they are equal, after which the resistance of the scale resistor is equalized until the output voltage of the converter is equal, all of its included bits, its nominal value, characterized in that, in order to increase the fitting speed, the formation of the second output voltage of the converter is carried out while simultaneously In addition, an additional 1; discharge bit is used, the imbalance signal is compared with a zero value of the unbalance control value, and before the corresponding resistance of the corresponding i-bit discharge resistance of the decoding matrix to the resistance of the corresponding i-ro discharge resistor, the formula l iiioo /};)  2. 100 where i is the number of the bit to be adjusted; Uj is the larger of the first and second output voltages of the converter; L; - Relative unbalance,%;  +1 t- yy-t coefficient depending on customizable bit numbers, which is used in the correction of the resistance of the corresponding i-ro discharge resistor of the decoding matrix for comparing with it the larger of the first and second output voltages, after reaching equality between which the correction is stopped. and  Rcs2 Have one Yas5, 1/71 1Н2 four Ic. F t ff / 2 (; 0i / 2j / tf%; Ai7. W / GM FIG. 2 KZH G Control FIG. H FagM