Measuring back emf in a 3-phase BLDC motor

[pyrochk247]

I am currently trying to determine the back emf generated in the floating coil of a 3 phase BLDC motor (simple RC car motor).  I can currently monitor the current and the voltages supplied to each of the motor phases from the controller but I would like to eliminate the use of Hall sensors from my setup and monitor the back emf instead of the Hall outputs (to use as feedback instead).  I am using a CB-68LP terminal block and the NI 6024E PCI card.  To run the motor at full throttle requiers a 5 volt PWM at 60Hz and 12% duty cycle.  I am no EE, just a lowly ME, so bear with my ignorance and give a girl a hand. 🙂  Any suggestions?  Thanks!

[jLo]

Hi pyrochk247,

Out of curiosity, are you using a motion controller to control the motor, or are you using the 6024E's counters to do the PWM?  We don't deal with Back-EMF as a method of feedback so you'll have to forgive me if I trip up my understanding of the topic:

In the way of reading the motor windings for a simple voltage and feedback voltage, that is highly possible.  Using a 6024E card you would be able to read in the analog signals and use the counters as outputs at the same time.  What would be difficult about this is pulling out the Back-EMF signals from the standard motor winding voltage.  I found this article:

http://www.acroname.com/robotics/info/articles/back-emf/back-emf.html

So the trick would be developing a method good enough to read the signals we desire well and rejecting the rest.  If you are using PWM from the 6024E then you could do decent feedback to the motor through a PID loop.

Currently, I'm not trying to suggest anything.  Rather, I'm asking to see if this idea makes sense to you.

Let me know what you think.

Justin Louie

[pyrochk247]

I am currently using a speed controller (simply one that goes on an RC car) and have put in a few shunt resistors in series with the 3 voltage lines to the motor (i.e. thats how I look at the current using the DAQ assistant).  But I think you are correct in assuming that I should be able to filter down the voltages that I am looking at to find the bemf.  I do use the signal generator in LabView to produce the proper PWM for the controller though.  I don't know if this answered your question but I will post tomorrow on some of the readings that I have found.  I have to look them over again to make sure I know what I am talking about before I explain how they suggest to measure the bemf. 🙂 Thanks!