View Full Version : Sensorless BLDC motor with back emf? Is there a solution?
I do need to drive a brushless dc motor without hall sensors or encoder. Did anyone implement a solution?
08-31-2007, 08:02 PM
There is a slight confusionregarding terminology. A BLDC does have its own electronics and is fed with DC. Other motors, like induction motors, permanent magnet (PM) motors etcetera need something that feeds them with a suitable waveform. Is it one of the latter you are asking about? Or is it really a BLDC you are asking about?
Yes, it is. A BLDC motor is a 3phase synchronous AC motor with trapezoidal BEMF. Commutation is made electronically in phase with the rotor position, normaly detected by hall sensors. But there are solutions for sensorless drivers, but not with the propeller. Freescale shows it in AN 1913
08-31-2007, 09:24 PM
It can be done by softly applying a DC field to the coils. You shall make the currents proportional to sin(0), sin(120) and sin(240). When the rotor has aligned itself, you can turn the resulting field by increasing the angles (0 goes to 1, 2, 3 etc, 120 goes 121, 122, 123 and 240 does the same thing). Ramp up the frequrncy you are incrementing the angles with and the motor will follow.
There may be a problem, a PM motor (which it actually is) usually doesn't have much damping. So fast acceleration may be a problem if you do not have a rotor position feed-back. You may need an encoder for that.
Another problem is that the trapezoidal back EMF. It is not an exact fit for the applied sine, so you will have some extra current flowing to compensate for the voltage difference that occurs in the "corners". Adjusting the waveform so it is more like the back EMF helps. And I think that the Propeller is powerful enough to take care of that adjustment automatically. Just feed back the current when running the motor unloaded and let the Prop adjust the voltage profile so that the "parasitic" currents are minimized.
08-31-2007, 11:29 PM
Whenever someone asks about back EMF BLDC, I refer them to this paper: http://scholar.lib.vt.edu/theses/available/etd-09152003-171904/unrestricted/T.pdf·Since it's a thesis it concentrates more on the theory side of things, but there is enough about his implementation to adapt it to another app.
Paul Baker (mailto:email@example.com)
Propeller Applications Engineer
Parallax, Inc. (http://www.parallax.com)
Post Edited (Paul Baker (Parallax)) : 8/31/2007 4:34:35 PM GMT
Paul, many thanks. This paper will give me a lot of information on the state of the art. But I have to study it and hopefully, I will implement a solution with the propeller!