FOC poor efficiency at high speed

Discussions regarding the Field Oriented Control (FOC) implementation on the VESC.
protontim
Posts: 7
Joined: 26 Sep 2017, 18:50

Re: FOC poor efficiency at high speed

Postby protontim » 02 Oct 2017, 14:12

@ ThierryGTLTS - What does "not work" look like? Please let me know what you find out about why the 3mOhm firmware does not work. I am following close behind. Thanks!

rew
Posts: 939
Joined: 25 Mar 2016, 12:29
Location: Delft, Netherlands.

Re: FOC poor efficiency at high speed

Postby rew » 03 Oct 2017, 07:31

If you can, try "compiling firmware" and "uploading firmware" before you start changing the hardware.

pf26
Posts: 296
Joined: 28 Mar 2016, 14:37
Location: FR Valence

Re: FOC poor efficiency at high speed

Postby pf26 » 03 Oct 2017, 13:26

You may also keep the original firmware with the larger shunts, and simply consider all your current measurements and settings are off by a certain coefficient. Be aware when soldering the shunts that there are 4 connection points (2 little ones for voltage sensing accross the shunt).

ThierryGTLTS
Posts: 86
Joined: 09 Aug 2017, 11:10

Re: FOC poor efficiency at high speed

Postby ThierryGTLTS » 03 Oct 2017, 15:26

protontim wrote:@ ThierryGTLTS - What does "not work" look like? Please let me know what you find out about why the 3mOhm firmware does not work. I am following close behind. Thanks!


When i wanna upload the 3.29 modified firmware, VESC Tool upload the HW6 firmware version, not the HW 4.10 version, I don't know why :?: :!:

Hopefully for me I have uploaded back the 2.18 version, and it works now.

Perhaps I will try to use a modified 2.18 version.

ThierryGTLTS
Posts: 86
Joined: 09 Aug 2017, 11:10

Re: FOC poor efficiency at high speed

Postby ThierryGTLTS » 03 Oct 2017, 15:44

Loading modified 2.18 firmware version with VESC Tool and then using BLDC Tool is working.

But it's not better, I have more noise and can't run the motor at a very low speed.

So I will wait for a more stable version of VESC Tool and try again.

ThierryGTLTS
Posts: 86
Joined: 09 Aug 2017, 11:10

Re: FOC poor efficiency at high speed

Postby ThierryGTLTS » 04 Oct 2017, 08:59

OK, the firmware works well now.

I've rebuild a 3.29 firmware for VESC Tool 0.82 and the measured current is right with 3 mohms shunts.

But I've still have a strange behavior in FOC when the motor runs very slowly, less than 1000erpm, and sometimes between 2300 and 2800 erpm.

Is it possible that it comes from the commutation between sensored and sensorless (at 2500erpm) :?: :!:

When the speed is mot very accurate, the current waves are distorded, but the hall sensor outputs seem to be stable.

Have a Nice Day.

Thierry

protontim
Posts: 7
Joined: 26 Sep 2017, 18:50

Re: FOC poor efficiency at high speed

Postby protontim » 04 Oct 2017, 18:56

I swapped out the shunts for 3mOhm and changed the code appropriately. Downloaded the new firmware to the VESC4 and the current waveforms look much better. I increased the observer gain to about 3x the default level and changed my observer gain at minimum duty to 10%. Before those gain changes I could run to about 30% duty cycle. After gain changes it ran well to about 50% duty cycle and then it starts to oscillate. Current waveforms look clean at ~50% duty cycle, but the current is modulated - looks like it is on the cusp of going unstable? Am I running out of processing power? Or is it a mechanical resonance problem? Or??? Any suggestions on how to proceed would be greatly appreciated.
Currents at 50percent.PNG
Currents at 50percent.PNG (533.22 KiB) Viewed 1557 times

pf26
Posts: 296
Joined: 28 Mar 2016, 14:37
Location: FR Valence

Re: FOC poor efficiency at high speed

Postby pf26 » 05 Oct 2017, 09:17

Maybe you want to try to increase the switching frequency (default 20kHz) and see if it helps. I think processing power does allow a bit more, like 30kHz (I never tried though). I guess you are already in the 30kERPM range or so. Is the motor driving a propeller ?
For some reason your phase 1 and 2 currents are a bit different which makes phase 3 still more different. This probably does not help FOC position estimation at high speed.

protontim
Posts: 7
Joined: 26 Sep 2017, 18:50

Re: FOC poor efficiency at high speed

Postby protontim » 06 Oct 2017, 01:51

I messed with the switching speed. No real effect. Still craps out around 50% duty cycle. Is there any way to start in FOC and switch to BLDC at a certain RPM??

ThierryGTLTS
Posts: 86
Joined: 09 Aug 2017, 11:10

Re: FOC poor efficiency at high speed

Postby ThierryGTLTS » 06 Oct 2017, 08:24

I don't think it's possible, but I'm not sure.

BLDC and FOC can switch from sensored to sensorless and that's what I do, because sensorless mode works better :)

Sensors are only there to have a clean start of the motor in my case.

Please can you do a very simple test :?: :!:

Just have a look at the current without any load and without filter at low speed (about 2000 erpm) to confirm that Phase 2 has less noise than Phase 1 and Phase 3.

I also notice that Phase 1 has a positive offset and phase 2 a negative offset.

There are not a lot of people that are working with a TMotor U8 here, so if you can help me, it'll be kind ;) ;)

Have a Nice Day.

Thierry


Return to “FOC”

Who is online

Users browsing this forum: No registered users and 1 guest