Thanks, @ziploc
1 Like
No problem. I’ve got another thread here that describes my build, and I believe I have the source code there if you need it.
2 Likes
I’m having issues with my footpads. When I switch to the balance app, the state is stuck at calibrating. I read you have to run the detect calibration for the IMU. However, after running it, applying, and switching back to the balancing app it still says calibrating. Any suggestions?
Tested today a CheapFOCer2 0.9 on my GF’s board and had some issues when accelerating hard from standstill as @doomy describes, but without going into limp mode. Some DRV OC faults were reported. Went back to the configuration and made sure my ERPM limits were correct…and they weren’t (now using variESC to avoid setting them up every time). Also lowered motor amps to 50/-30 and battery to 60/-20 (10S LiPos, plenty to go around…but defaults are kinda insane). Everything works fine now, so I’m even giving HFI a go! Thanks a lot @shaman for your hard work.
4 Likes
@thunkar how did the eRPM limits improve hard acceleration from a standstill? I’m guessing sensorless operation? If so, are talking about the openloop eRPM settings? Does using any of the presets help?
I’ve found that selecting the 'large outrunner" option for motor detection, even for smaller outrunners, yields better results. Detection accuracy affects low-rpm performance. Then also playing with the switching freq can help too.
4 Likes
I figured it out. I had to hit the write app configuration button.
2 Likes
Btw, I’m finally getting somewhere with my first eMTB build. It’s not in its end state yet, but hopefully this will help me understand all the factors that affect performance for this kind of application
7 Likes
Yest, it is a sensorless motor, running on FOC after an apparently normal detection. I am sorry, I confused the settings…I was actually talking about the duty cycle limits and the Duty Cycle Current Limit start. They were basically set to 100%. I know that should only affect the top end…but weirdly setting them to “sane” values completely got rid of the issue. I’ll keep experimenting and will report back…
Also my PH-188 motor is here now and the next FOCer should be getting some love soon. Since onewheels seem to also trigger the issue, please let me know if I can be of assistance.
1 Like
Yeah man just posting about what settings cured which issue for what kind of setup would be massively helpful.
1 Like
Gonna go downstairs to test the latest stuff on 5.02 firmware and will report back.
Also, fuck single drive boards, they’re so finnicky and put way too much stress on everything…at least I think we can get good data, since the electronics are going to suffer.
2 Likes
Here I am, it runs so nicely on HFI. Better than sensors, if it weren’t for the noise… Board and settings:
Single drive SK3 6374, belt, 12/36
2x 5000mAh 5S LiPos in series
Firmware 5.02 on a CheapFOCer2 v0.9
Motor: 50/-30A
Battery: 60/-20A
HFI: 14/4/8V
Everything else default as per variESC settings. 0 cogging, limping, resetting or issues whatsoever. Tested on slight uphill, from standstill and hitting full throttle, then hitting brakes until belt skips. Trying this on the firmware from Github master branch would give me DRV faults unless duty cycle was limited to 85%. Will try to put some more km on the board this week, but seems very promising. Hope it helps!
2 Likes
I’m stuck again. I’m trying to do a simple bench test by applying pressure to my footpads and have the motor/wheel spin. My motor only works when I set it UART and enable key control. When I switch to Balance and apply pressure to the footpads it does nothing. The values for ADC1 and ACD2 goes from 0 to 3.2. What am I missing?
1 Like
@JRMN You need to select ADC as the App To Use since you’re using analog analog footpads. I suggest reaching out to @NuRxG, @Flux, @Fungineers or anyone else that specializes in the balance stuff.
This post can help too
1 Like
In the tune tab you need a P value. As you are benchtesting i suggest taking a low P value of 1 and a low motor current value of 5amp or you are hitting max duty in a flip of a second. This way it is more gentle to control.
1 Like
Thanks Shaman and Dummer. Much appreciated.
1 Like
@shaman running into issues again on steep hills…DRV faults everywhere when maxing out the throttle. Tried increasing commutation frequency to 30KHz and battery max to absurd levels (130A) to no avail.
2 Likes
After a whole day of troubleshooting trying to find a way to mitigate the DRV fault issue under high current/low rpm conditions, these are my findings:
- No settings seem to particularly improve the behavior besides limiting motor amps as much as possible. This includes DRV OC settings, current filter settings, switching frequency, current controller decoupling…everything I felt I understood and wouldn’t destroy the controller.
- Also run detection multiple times at different currents, RPMs and time constants with consistent results, so no luck on that front.
When that didn’t work, I dissasembled the heatsink and redid part of the soldering, in particular the DRV (for which I used the hole in the bottom and went to town on the ground pad with tons of solder and flux), C35, C36, C37, C15 and C31 (which I had to solder myself) and the inductor. Finally, I bridged the hell out of the resistors next to the phase wires, which I missed during assembly.
Repeating the testing seemed to improve things a lot, and I went up to 60 motor amps (which in the end triggered the fault), but could not get it to fail at 55A. The fault was the same but as before, but reported absurd current values, such as -328A instead of 50/70.
Bear in mind that my testing was inside the house, clamping the board to the floor which was covered in a thick rug, which stressed the drivetrain enough to trigger the fault consistently.
I hope it helps someone and maybe @shaman can add as to why it got better, but my money is on the DRV ground pad, which was improperly soldered by JLCPCB. I’ll try to test the board in real conditions ASAP and maybe add some more data.
6 Likes
Hello all,
After a few weeks I got the time to test my CheapFoccer 2. I would like to thank everyone involved and this welcoming community.
After flashing my CheapFoccer2 with the files from Shamane’s github account, I tried to setup the wizard, but there seems to be a problem. A weird one, I don’t know if it’s a hardware or a software one, but I’m thinking that it’s probably a software problem.
Here is a video of the problem :
What do you think ?
Thank you very much in advance.