Position Measurement in Industrial Drives by Means of Low-Cost Resolver-to-Digital Converter

Speed error mitigation for a DSP-based resolver-to-digital converter using autotuning filters

“…Coil imbalance, amplifier gain imbalance and dc offsets are the most significant sources of resolver signal error which may introduce estimation errors in the RDC output [1], [5], [6], [12]- [14]. We describe these errors onto the resolver signals in (1), adding a constant voltage, , to both , representing the dc offset, and the term, sin( ) cos ( ), on representing coil and amplifier gain imbalances (called the gain error), where is a constant depending on the degree of imbalance [13], [14].…”

“…In Fig. 5, 1 and 2 have been used to interpolate a new filter 3 with 3 equal to 350 Hz based on (10)- (12). Note the accuracy of the interpolated filter in Fig.…”

“…These errors appear in the speed output of the RDC as oscillations of a frequency that is an integer multiple of the resolver mechanical frequency of rotation (e.g. a harmonic) and of a magnitude that is proportional to the resolver speed [12]- [14]. Calibration methods to deal with these errors based on the least squares as in [3] are more suitable for open loop systems.…”

“…1). This is especially so at high RDC speeds, since the magnitude of the error harmonics is proportional to the resolver speed [12]- [14]. This, therefore, may require additional pairs of SRF at higher than the 2 nd order harmonic frequency, which will add a significant computational overhead due to the complex transformation computations required by each rotating frame.…”

Design and Implementation of a PMSM Rotor Position Detecting System with High Speed and High Precision

“…52.2, R1-R2 is the exciting winding or the primary winding, S1-S3 and S2-S4 are the secondary windings. When the primary winding inputs the highfrequency sine wave, the pulse magnetic field is rotating with the rotor, thus the winding S1-S3 and S2-S4 induce corresponding sine and cosine signal varying with the rotor position [4]. The relationship between the input and output is shown in formula 52.1.…”