Hi, I am currently working to tune a VESC to power a vehicle with 4214 Xing Quadcopter motors (660kv). I am turning a flywheel with the motor to provide some resistance and also measure the motor's torque. It was all going well until I started to up the current limit, at 75A a harsh noise started coming in that sounds like loose bearings but I believe it is actually due to the motor and how it is being controlled. Here is a (Crappy unedited) video showing the issue here: https://youtu.be/d-BT6kmxFf4. Looking at some logged data there seems to be a major difference in current between phases during the issue and also some huge spikes in BEMF (see images here). If anyone could make any suggestions as to what might be the cause and/or what to change next that would be much appreciated. My goal is to get the powertrain running smoothly and reliably before fitting it (and three others) into the vehicle!
Thanks in advance!
I have tried numerous things to improve it with limited success (based on trawling through the forums):
What didnt work:
Trying higher or lower zero vector frequency (30khz standard, tried 24 and 35)
Cutting the observer gain (currently 831, tried 450)
Raising motor resistance slightly
Saturation compensation setting of factor 10%
"heavy inertial load" setting in FOC sensorless
Re-doing detection
Decreasing inductance 10%
Decreasing the time constant
What did work:
Change observer to ortega original (still makes the noise but doesnt overcurrent cut as much)
You can first check in Terminal -> "print faults" to see details of the fault. You can upload your motorconfig xml file so its easier to see your settings. You might need a time constant of about 150Us to 250uS on such motors.
Link to forum post: VESC help offered for private persons and companies
Website: www.electricfox.de
Thanks for the reply Electricfox, I'll do what you suggest tomorrow.
https://youtube.com/@steveeng1826?si=6EsA0w03ctFcJrjw
Hi, Here is the motor config file. here is some logged data from when the noise is being made, showing the spikes in BEMF and the uneven currents across the phases (as shown in the images I shared earlier). When it does cut the error log simply states over current (I believe because of the uneven current across the phases and one phase spiking up to the set limit).
I tried reducing the time constant to 200us and updating the kp and ki in accordance with that, and it made no difference. I've also tried going up to 50khz on the zero vector frequency with no improvement. I experimented with the current controller decoupling modes since it sounded like they could help in this situation of rapidly changing rpm, but it didnt make any difference. So I'm currently a bit stuck and would welcome any suggestions! Thanks
https://youtube.com/@steveeng1826?si=6EsA0w03ctFcJrjw
I just took a look. Never set Max Duty Cycle higher than 95% ! Zero Vector frequency should be good at 30khz. Depending on your hardware you can give Control sample mode V0 and V7 and Current sample mode all sensors combined a try. Your CSV file is broken and is missing data.
Link to forum post: VESC help offered for private persons and companies
Website: www.electricfox.de
Hi and thanks for having a look and for the suggestions. I'm using a Makerbase 60100 as hardware. I believe this has three current sensors so I thought I would be able to try current sample mode all sensors as suggested, but when I looked that option is not there (only high current sample mode true/false). Do you think that would change if I update the firmware from 6.02 to 6.05 or have I got a hardware limitation there? I tried control sample mode V0 and V7 with no change to the results. I'm not sure what is up with the .csv, it seemed to open fine for me. I have zipped it and re-uploaded here if you fancy giving it another go (thanks).
https://youtube.com/@steveeng1826?si=6EsA0w03ctFcJrjw
The functions are available on 6.05 yes, so you will need to update vesctool and the firmware. The file does not work for me, its still broken. There is no need to zip it.
Link to forum post: VESC help offered for private persons and companies
Website: www.electricfox.de
Hi Electricfox, I updated my firmware (It was way easier than I expected!) and was able to try those other settings you mentioned, with no change in behaviour unfortunately.
On the plus side I'm not experiencing the issue at 75A (the screeching noise happens at 80A plus), and in my final application there will be less acceleration as the speed increases (due to air resistance) so maybe under that situation 90A would work if it related to the rapid change of erpm.
Thanks for the suggestions so far, let me know if anything else springs to mind.
https://youtube.com/@steveeng1826?si=6EsA0w03ctFcJrjw