Need Help: Inverter goes to full throttle during testing

[Ember]

Hello everyone,
I am currently testing my Gen 3 Prius inverter and transaxle with a V1d board from EVBMW. I’ve got them hooked up to a HVJB and I am currently testing with a 80v battery. I am trying to get my throttle working to spin MG2, but keep running into the issue of the motor “running away” and not stoping when I release the throttle. Here is a video of the issue I am having:



I was able to spin MG2 successfully with sine firmware and then with FOC in manual testing mode.
I currently have v5.24 FOC firmware that Pete edited to have a fixed flash size because the chips on my board were having trouble flashing.
I was first testing with a gen 3 Prius throttle and now I am using a gen 2 Prius throttle with the same issues.
I have tried using single regen mode with pot min at 1400 and pot max at 4095 but that made the issue worse.
I am currently using dual channel potmode with pot 1 at 700-3800 and pot 2 at 1500-3800 and have not been getting throttle errors from the inverter.
You can also hear in the video that the motor is making a knocking noise until it runs away, I’m not sure what this is.

If anyone has any advice or has experience with a similar issue any help is greatly appreciated!

Thank you!
Ember

Here are all of my params:

params (4).json

(1.46 KiB) Downloaded 96 times

[Zieg]

Can you tell which way it's going? Is it regen or is it spinning forward?

[Ember]

It is spinning forward

[Ember]

I am now pretty sure that this is an issue with my syncof value, because when I change it the behavior changes. At syncof 12900 the motor spins smoothly in the forward direction, and yields good throttle Controll, but seems slow, and if I put it in the reverse direction it runs away immediately. I tried adding 32,768 to the value, but this did not solve the issue. I am having a lot of trouble finding the correct syncof value. It appears there is one trough around 10k and another around 45k. My testing battery is only ~75 volts. I am wondering if anyone can comment on how this resolved feedback looks? I have tried changing the resolver wiring configuration without success. The first photo is with Data points 1000, burst length 10, and the second is data points 10000 and burst length 100. Thank you very much!

[FFMan]

See if you can find my Prius thread from when I started with Prius motors etc. on mobile so can’t link to it easily.

The upshot is there are syncoff points which look right but only one is really right.

There is also a Prius thread about achievable output and on the back of that I swapped to a phev setup before I installed anything.

[Ember]

Thank you! Currently my parameters are:

params.json

(1.46 KiB) Downloaded 104 times

I now have snycofs at 10,000, and when field weakening was disabled the unloaded motor spun well in both directions, with the same resolver as provided the plots above.
I already had it mounted in my project, and installed the cv axles. I have a V1d board but currently am only using MG2. My project is super light (1980 VW Rabbit Pickup) and I have a 360v BMW i3 battery. Currently I am able to go 47 mph (76 kmh) and cruise comfortably at 45 mph (72 kmh) according to the GPS on my phone. Luckily that is suitable for me, but not ideal lol.