Hi VESC hard-core nerds, supporting crew and developers,
I've one of the very original VESC6.0's and recently bought recently a newer version VESC6.0MK. They are used to control T-Motors MN705S, KV140, 28mag poles (or 14 pole pairs). Both originally sensorless and no temperature compensation.
The troubles I had:
* Current control: none (all good ;-) )
* RPM control: VESC 6.0 "the new" had a very unreliable, jerky, rough start up from low rpm. I mean moving back and forth, no acceleration for a long time, then suddenly increasing speed. (maybe).
Setting comparison: Identical settings fon both VESCs
Switched motors: no improvement.
Ordered and integrated AS5047D rpm sensor and integrated temperature sensor NTC 10k into one motor with the "the new". Configured.
RPM control at lower speed nice, but multiple times a "high speed run away" when started (tested with experimental and via CAN message). -> The new VESC only has 5V sensor out? Possible noise?
Finally figured out, that if I disable the Phase Filters on "the new" it has also a nice smooth run up as my very old VESC6.0. This is surprising because if I understood, the phase filters should help at low rpm, but they don't, they actually cause troubles.
Q1: In sensored mode with AS5047d -> very often "high speed run away at start up" -> I guess until Open Loop ERPM is met. Why?
Q2: If phase filter cause troubles at low rpm, why should they improve low rpm?
Best regards,
Tobias
Your max "sensorless ERPM" (in Encoder tab) with AS5047 in SPI mode should not be more than 2000-8000ERPM. You can not run it sensored until max erpm, also because the signals are sensitive.
Link to forum post: VESC help offered for private persons and companies
Website: www.electricfox.de
I have not changed the settings, and I checked, it is, and was set to 4000 ERPM (~285rpm). I tried to repeat the run-away by using RPM control from the experimental taps. Startup from 0...to...2000;250step;2s. hold 10000end. (FOC=Sensored Mode, Phase Filters = False)
Worked 15 times and then:
The first figure shows the same experiment in sensorless mode, phase filter=false.
The second figure shows the experiment in sensorless mode, phase filter = true; as you can clearly identify, a tough stuttering at the beginning and very rough low rpm.