Figure 2 - uploaded by Victor Aksenenko
Content may be subject to copyright.
Source publication
The paper describes the principle of operation of a resolver-to-digital converter based on signal processing in DSP, analyses the main conversion error components, including the one due to quantization of signals that occurs when the latter are entered in the DSP. A number of conversion algorithms considered in the paper make it possible to enhance...
Contexts in source publication
Context 1
... us perform the numerical estimation of the error in the angular positions ξ ξ 1024 , where Figure 2 shows the error for N=16 and k=12. Hereinafter the error value is given in arc minutes. ξ ξ 1024 , where Figure 2 shows the error for N=16 and k=12. Hereinafter the error value is given in arc minutes. Of particular interest for choosing ADC by a specified error is the maximum AnDC error value with different N and k. The calculation results of the maximum error value for the samplings numbering from 8 to 64 for a period N and ADC resolution k – from 8 to 16 are given in Fig.3. The error values given above caused by signal quantization and sampling have been calculated for idealized conditions without regard for noise and interference. In real conditions the ADC converts a mixture of signal and noise. Under the effect of noise the error value of the system varies in time even if the shaft angle does not change. Further, the error value fluctuations are smoothed by the system when the average angle estimate is being formed. Modulation and averaging provide decrease of the error caused by signal quantization and sampling. The extent to which the error decreases depends on the level and noise spectrum as well as transfer function of the virtual servo system of the converter. Gains can be calibrated when the sensor shaft turns within the conversion range, for example, within the whole revolution. For this the amplitudes с 1 and с 2 of the first harmonic of the sine and cosine signals are calculated at angular positions equally spaced over the prescribed range and stored in the ...
Context 2
... us perform the numerical estimation of the error in the angular positions ξ ξ 1024 , where Figure 2 shows the error for N=16 and k=12. Hereinafter the error value is given in arc minutes. ξ ξ 1024 , where Figure 2 shows the error for N=16 and k=12. Hereinafter the error value is given in arc minutes. Of particular interest for choosing ADC by a specified error is the maximum AnDC error value with different N and k. The calculation results of the maximum error value for the samplings numbering from 8 to 64 for a period N and ADC resolution k – from 8 to 16 are given in Fig.3. The error values given above caused by signal quantization and sampling have been calculated for idealized conditions without regard for noise and interference. In real conditions the ADC converts a mixture of signal and noise. Under the effect of noise the error value of the system varies in time even if the shaft angle does not change. Further, the error value fluctuations are smoothed by the system when the average angle estimate is being formed. Modulation and averaging provide decrease of the error caused by signal quantization and sampling. The extent to which the error decreases depends on the level and noise spectrum as well as transfer function of the virtual servo system of the converter. Gains can be calibrated when the sensor shaft turns within the conversion range, for example, within the whole revolution. For this the amplitudes с 1 and с 2 of the first harmonic of the sine and cosine signals are calculated at angular positions equally spaced over the prescribed range and stored in the ...
Citations
... The fundamentals of the approach that provide full and automated measuring of the error of angular transducers in a system were presented in [8][9][10][11]. This approach is based on the dependence of the DAS error harmonic frequency on the speed factor of the resolver. ...
Digital angle sensors built on multipole resolvers and angle-to-digital converters have found a wide application in mission-critical systems operating in severe environment and determine their accuracy and efficiency to a significant extent. A method for self-calibration of digital angle sensors based on integration of two conversion channels with the errors of different spectrums is suggested. Two approaches specially intended to increase the accuracy of self-calibration for the case of high initial errors of the channels are discussed. The results of the simulation and experiment are presented
... Various engineering solutions that realize automatic measurement and compensation of larger or smaller portion of error components are available from many different references, such as [3][4][5][6][7][8]. All of them are based on structural redundancy and the use of additional angle-measuring channels with the error sign opposite to that of the main channel. ...
... All of them are based on structural redundancy and the use of additional angle-measuring channels with the error sign opposite to that of the main channel. In [3][4][5][6][7][8], error compensation is achieved by summation of the output data from the main and additional channels. Application of these methods allows significant improvement of conversion accuracy in the static mode. ...
... The present paper discusses the results of the theoretical analysis of the channel residual error dependence on the dynamic characteristics of the DAS in the case that this error is corrected by the known methods given in references [3][4][5][6][7][8]. The approaches used to increase the efficiency of correction in dynamic modes are described. ...
The aim of the present paper is to study how the variation of the angle converted by the digital angular sensor influences the correction of its systematic error. The principal causes of decrease in error compensation for the case that the shaft rotates are analyzed. Two approaches that allow improving the correction accuracy in dynamics are described. The results of the analytic investigation and numerical simulation with the use of these approaches for different modes of the shaft motion are presented.
... Recently a tendency has arisen to decentralize the data processing in systems; data pre-processing is performed mainly by microprocessors. Sometimes the microprocessors are also used to process the resolver signals in order to estimate the angle [3][4][5][6]. ...
Angle-to-digital converters consisting of multipole resolvers and devices converting their signals to digital representation are widely used in mission-critical systems working in severe environment; they define substantially the accuracy and efficiency of such systems. The report presents the resolver self-calibration technique based on integrating two conversion channels having different error spectra. A goniometric setup and technique for the converter error investigation are also described. Experimental results are presented.
