Parameters disappear after throttle! [SOLVED]

[ekohn04]

Hey all,
I'm trying to bench spin my LDU. I was having an OVERCURRENT error, then I swapped my params to some basic LDU params that were posted on the forum. I'm able to get it into run and fwd. When I apply throttle I hear a small blip of a whine from the inverter and then nothing. My pedal doesn't do anything after this. However, the contactors stay closed and the voltage (measured w/ multimeter at busbars) is still present. When I go to my web interface to look for an error, nothing is displayed. I see the headers for the sections, however there are no parameters or spot values visible. The board is still putting out a WiFi signal, but there's no information being displayed! Once I turn my LV switch off and then back on I can see my params again. Here is a copy of my params file. Does anyone know what's happening?

[johu]

The parameters seem slightly odd, try these: https://openinverter.org/parameters/view.html?id=16
You have tripmode set to "1" or "DcSwOn" which is generally recommended as it doesn't open the contactors on error. In the linked parameters that is not the case so you might just want to apply the motor parameters.

Loosing comms must be down to a restart. Basically ESP module and STM32 negotiate a higher communication speed on startup. But if for some reason one of them reboots it is back to the slower communication speed and the two can't communicate. I wonder is your LV supply stable? The inverter pulls more current when PWM is running and that could lead to a voltage dip.

[ekohn04]

My LV connection was the problem! I was using some pretty thin wire to get my 12v to my harness. I think this is definitely a lesson on the importance of having strong and capable connections. Thank you johu, for being so helpful. I finally spun my motor for the first time on the bench.

[alexbeatle]

I'm curious. Is this what your web-ui looks like when you "loose parameters"?
My SDU board has good comms and runs fine when I test it on the bench
However, when I put it in the drive unit in the car, bolt and torque it, it misbehaves. My laptop connects and I can browse the web-ui, but after few minutes it disconnects. Though the inverter7 network still shows in the list of networks. When I reconnect to the same inverter7 network, I see everything in the web-ui except for the parameters. Contactors stay closed and motor spins with accel. I can change directions too.
This gets fixed after rebooting the Ignition, but only for a few mins.

[johu]

This indicates that for some reason the wifi module did a watchdog reset. Afterwards the stm32 still communicates at 1M baud and the wifi mod at 115k baud. It might be caused by the data rate being so low that the watchdog resets the wifi module.

It doesn't affect drivability as the wifi module is just complementary.

[alexbeatle]

This indicates that for some reason the wifi module did a watchdog reset. Afterwards the stm32 still communicates at 1M baud and the wifi mod at 115k baud. It might be caused by the data rate being so low that the watchdog resets the wifi module.

It doesn't affect drivability as the wifi module is just complementary.

What triggers the watchdog reset and how can it be avoided? Very difficult to try different parameters when this keeps happening

[johu]

It seems downloading large chunks such as the full refresh. Use the plot to observe values if signal strength is poor

[uhi22]

... the wifi module did a watchdog reset. Afterwards the stm32 still communicates at 1M baud and the wifi mod at 115k baud.

Opened an issue to improve this: https://github.com/jsphuebner/esp32-web ... e/issues/1

[ekohn04]

'm curious. Is this what your web-ui looks like when you "loose parameters"?

Yeah, that's exactly what my interface looked like. The wiring for the power to my OI board was really thin, all I did was increase the gauge of the wires supplying 12v and it solved my issue. It sounds like the problem you're having is different because I wasn't able to spin my motor after I lost view of the parameters.

[alexbeatle]

Yeah, that's exactly what my interface looked like. The wiring for the power to my OI board was really thin, all I did was increase the gauge of the wires supplying 12v and it solved my issue. It sounds like the problem you're having is different because I wasn't able to spin my motor after I lost view of the parameters.

What was your original wire size and length? What is the new?
Mine are 20awg and ~10ft long. Voltage drop ~ 0.3V from the source.

[ekohn04]

original wire size and length?

Since I was just bench testing when this happened I was really using whatever I had lying around. A little embarrassed to admit I was using premade alligator clips with 22AWG wire. I used some ridiculously thick 8AWG when I got it to work, just because I wanted to go overkill for peace of mind.

Mine are 20awg and ~10ft long. Voltage drop ~ 0.3V from the source.

I'm by no means an electrical engineer but 20AWG seems a little thin for 10ft, especially for supplying 12v power to the board. I feel like it would be ok for the signal wires but when the board starts pulling more amps from the 12v supply it might be a different story. It might have the right voltage but need to be pulling more amps. I would suggest getting some thicker wire to see if it solves your problem.