Sometimes motor builders don't provide this important piece of information that is critical to Field oriented control of sinusoidal (AC) permanent magnet brushless motors. In this question I assume that the rotor angle is the angle between the direct axis of the rotor magnet and the axis of phase "A". If one could know this offset delta (and whether it leading or lagging the N-pole axis with respect to a given sense of rotation), it is then easy to decide the angle of the stator current vector necessary for FOC.
The easiest way I think is that if you rotate the rotor then it will induce voltage on stator winding this is back EMF. After getting back EMF, you shoul see the enconder index pulse on oscilloscope at same time. Finally, you can see and know how much phase degree has the motor between back EMF and index pulse. actually back EMF is the best way to understand direct and quadratue axes of motor.
Thanks too much Yusuf, This is what I do actually in a lab where you can isolate the motor terminal and the encoder terminals and just feed the incremental encoder with its rated voltage and run the motor as a generator with another prime mover. However, in other situations, this procedure may be too difficult(e.g. an elevator driven by a gearless PMSM with 58 pole pairs has a malfunctioning encoder that needs replacement. on replacement, the technician is not sure whether the replaced one has the same zeroing as the old one or not), here you can isolate the cables and excite the encoder but rotating the motor as a generator at constant speed is not straight forward as it is always coupled to the elevator car (yes it can be driven by the counter weight gravity force, but it may accelerate to dangerous speeds!)
I was asking if some one has another doable check in such circumstance
This is an interesting real-world problem. The issue here is that whomsoever installed the system did not think about this. They should rather have used a resolver.
However, if possible you should first try to determine where the current encoder index pulse occurs.
First make marks which are all aligned on both the motor shaft and encoder shaft and stator; these form the reference point.
If you cannot run or turn the motor then, after applying the brake and locking the lift into position.
Then undo the encoder coupling and turn the encoder shaft (do not move encoder stator housing) until you locate the first index pulse and make another mark on the motor shaft and encoder housing. This second mark will be the actual location of the index pulse of the existing encoder which is to be replaced.
Once you have done this you can do same as you always do for the replacement encoder on the bench and make marks on both the rotor shaft and stator so that they are aligned with one another at the point at which the index pulse occurs on the replacement encoder. It is tricky for high resolution resolution encoder.
Thus when you remove the faulty encoder you will have the point at which its index pulse occurred and you may then align the replacement encoder to ensure all three lines line up.
If they do then it will ensure that the old and new index pulses are lined up as was before repalcement.
Alternatively if lining up is problematic then you can determine the offset as the difference between the old and new index pulses and compensate for this.
This should be doable if the motor encoder is accessible and whether either its its rotor or stator may be rotated by hand.
Hope this helps :)
Thanks Gregory too much, your input is really helpful. Just for sharing ideas, we have gone through a different procedure. We have made the counter weight and the car load critically balanced such that the motor sheave could be rotated with hand. Then we applied a steady zero-angle vector from the inverter (at current saturation limit) thus an assisted-rotor -homing (or alignment to the zero angle) could be done. We dismantled the old encoder, and prepared the new one with a mark on its rotor and housing that corresponds to its zero. Then we tightened the new encoder such that the zero marks on both its members are aligned . Now, the rotor at home position gives a zero angle by the encoder and it works fine
Great, that what good engineering is all about.
Glad you managed to sort out.
I hope you have made marks so that it may be easily done next time :)
Connect your motor to VFD.
Now as we know , electrical angle(theta) = (p/2)*mechanical angle.
So for 8 pole motor,VFD will show 4096 count per revolution (for π) as 4096=8π.
Hence π/2=256.
- Badges
- Science topic
- Similar topics
- Engineering
- Control Theory
- Control Systems
- Control Systems Engineering