SU1651220A1 - Method for measurement of analogue quantity - Google Patents

Abstract

The invention relates to digital measurement technology and is intended to improve the accuracy of digital measurements of analog quantities. For this purpose, multiple sampling of the analog signal is carried out, and all the drifts of the analog-to-digital converter (A1SC) are recorded. The sampling-storage cell 1 selects a signal, and, at the command of the control computing unit 5, closes 2.1 ... 2.t, integrators 3.1 ... 3.that integrates the incoming signal s each clock cycle, and A1SCh 4 converts it into digital form, and in the absence of a signal at its input in the first cycles, block 5 receives the equivalent (drift) of the zero signal, which takes into account the final signal processing. 1 il. about 8

Description

The invention relates to digital measuring equipment and is intended to improve the accuracy of digital measurements of analog quantities.

The aim of the invention is to increase accuracy and expand the scope.

The method of measuring the analogue value is that a measured quantity is sampled, it is integrated, the next measured quantity is sampled, the signal is subjected to the second integration for the first integration, and the first integration is performed for the first integration, the next sample is taken, the second integration signal is subjected to the third integration, the signal of the first integration to the second integration, and the sample to the first integration, repeating this procedure in N cycles, for-:. they integrate the mth integration signal, each time they convert this signal into a digital code and store it, ^ calculate the final difference of the th order of the sequence of stored digital: codes received in (m + 1) -th consecutive Measurement cycles, and get the corrected measurement result of the corresponding samples of the analog value, i.e.

And thus the error with. q = Y "". approximation of the corrected measurement result to the digital equivalent _{No. + 1} = x k-mind + A of the real value of the reference of the measured quantity x is determined by the finite difference _{k of the} tnth order of error about the analog-to-digital Converter (ADC) in the corresponding measurement step, with N ^ m + 1, and ha and k - whole numbers.

The drawing shows a structural diagram of a device that implements the method.

The device contains a cell 1 sample-storage, analog keys

2.1.2.2 ...., 2.t, integrators 3.1,

3.2, ..., З.т, АЦП 4, the control computing unit 5, the control output of which is a bus 6 clock signals connected to the control inputs of the keys 2.1 ... 2.sh. Cell 1 retrieval-storage, keys

2.1 ... 2. Ш, integrators 3.1 ... З.т,

The ADC 4 and the control computing unit 5 are connected in series.

The device operates as follows.

In the zero cycle, when all the blocks are in the initial state, the output code is a clock signal

ADC 4, which corresponds to its error (zero drift), enters the control computing unit 5 and is stored.

In the first step of operation, cell 1 of the sample-storage of the measured quantity x makes a sample x ^, which is integrated by the closure of key 2.1 by integrator 3.1, at the output of which there is <^

2 _t (t) = 1 Jx ^ i ^ X ,,

A ° where r is the time constant of the integrator;

- the duration of integration, determined by the interval of time of the closure of the key 2.1 timing signal.

At the same time, there is no signal at the output of the Z.t. integrator; at the ADC 4 output, a code is generated corresponding to the new implementation of its zero error.

In the second cycle, cell 1 of the storage sample makes an x ^ sample. First, the key 2.2 is closed and the signal of the integrator 3.1 is integrated by the integrator 3.2, setting the signal z ₂ (t ₁ ) = z _< (t ₄ ) = x _< at its output. After opening the key 2.2, the key 2.1 is closed and the sample x _{4 is} integrated by the integrator 3.1, the output of which is set to the signal z <(t ^) = z ₁ (t |) + x _s , = X (+ X ^.

Subsequently, the conveyor process of signal transmission between integrators is repeated, and the ADC signal 4 is fixed in the computing unit 5.

At measure m, cell 1 of the sample-store sample x _ffl . The first key is closed 2.and at the output of the integrator

Zt, the signal z _m (t _m ) = = zn) _ ((t _ni _ ₍ ) = x _< , which in ADC 4 is converted into a digital code y $ = z _M (t _n ) + ⁺ “η> ^{= χ} ι ⁺ ^ τ » ^{r; ie} is the signal conversion error z _m (t _m ) = x _| . The operation of the keys 2.1 ... 2. ha is carried out similarly to that described.

Code y = x * _(m + 8 is processed in block 5 together with the code sequence y £ = equivalent implementation zero error ADC 4 in previous cycles, the results give final ^tat "A

USCH ^{= x} 1 ^{+ L}

The process can be continued, receiving new corrected measurement results.

Claims (1)

The claims are masks, the second integration signal is subjected to the third integration, the first integration signal is the second, and the third sample is subjected to the first integration, repeating this procedure in N steps, the signal of the integration is generated, the digital equivalents of this signal are stored in all measurement steps, the final m-order difference of the stored sequence of digital A method of measuring an analog quantity, which consists in generating a sequence of 15 samples of the measured quantity, integrating it, converting the signal of the last integration into a digital code, characterized in that, in order to increase the accuracy of 20 and expand the scope, they integrate the first sample, produce the second sample of the measured value, the signal of the first integration is subjected to the second integration, and the second 25 sample is the first integration, the third sample of the measured non equivalent signals of the th — th integration in the corresponding measurement clocks, as a result of which in the (nH-1) th and subsequent clock cycles we obtain the sequence Г У. 1 = Гх. - \ + r mb? I k J I J The corrected measurement results of the corresponding samples ^ of the measured analog value, represented as the sum of the true value of the reference x _k and the final difference of the mth error order of the uncorrected digital value, where Nim + 1, and tick are integers.