AC Drive China Forum
Forum » General Discussion » PMSM sensorless vector control
Topics: PMSM sensorless vector control on General Discussion
Start by
Ravi Kumar
09-16-2013 04:04 AM

PMSM sensorless vector control

I am working on sensorless vector control . I am using high frequency signal injection to estimate rotor position using heterodyne approach . I have issue in stabilizing the system when estimated rotor position used to control the motor . Please share in your thoughts how to tune the heterodyne system for proper position estimation.
09-16-2013 06:37 AM
Top #2
Marek Stulrajter
09-16-2013 06:37 AM
Hi Ravi,
What does mean "issue in stabilizing the system"? ... you can switch to the sensorless mode, but you have a ripple on the rotor or you even have a problem to switch and run the motor without the sensor?
What is the injected frequency and what is your sampling frequency? What electrical frequency you expect to track with your sensorless approach? What type and the order of filters you use in signal processing (heterodyne technique)?
I am sorry for too much questions at the beginning, but this is very complex topic and I am sure nobody can give you an advise in 2-3 sentences.
09-16-2013 09:16 AM
Top #3
Sandro Calligaro
09-16-2013 09:16 AM
Hi Ravi,
have you already tuned your system while estimation is "off-line" (i.e. not used for FOC)?

I mean injecting and estimating position/speed while moving the rotor by hand or with sensored control (pay attention to the different Park transforms involved), and looking at the estimated vs. actual position/speed.
By doing so, by trial and error you can find proper gains for your estimator (and reach a limit above which estimation gets unstable). Then closing the loop (i.e. using estimated position/speed for FOC) should be a lot easier.
Reply to Thread