openinverter forum

I tested my SDU while the car was stationary and the motor spun up successfully. After that I drove about 50 meters and everything was fine. Later I tried to drive again, but nothing happened—no throttle response.
I had set the parameters 45 udcmin and 72 udcsw to zero.
An E46 throttle pedal is being used, and the HV voltage is 50 V.
The error messages are:
- 195 Stop overcurrent
- 159 Warn Throttle1
- 160 Warn Throttle1
- 161 Warn Throttle1
The car is a BMW E46.
Can I reset these errors (how), or is there a mechanical defect?

Quax wrote: ↑Wed Oct 01, 2025 8:00 pm Can I reset these errors (how), or is there a mechanical defect?

Errors should reset on a power cycle. If they reappear, then there may be damage. You can do a diode test of the phases to see if they are damaged.

Quax wrote: ↑Wed Oct 01, 2025 8:00 pm After that I drove about 50 meters and everything was fine. Later I tried to drive again, but nothing happened—no throttle response.
I had set the parameters 45 udcmin and 72 udcsw to zero.
An E46 throttle pedal is being used, and the HV voltage is 50 V.

50V is awfully low to try to move a car with a Tesla drive unit. That may have led to the overcurrent error.

Quax wrote: ↑Wed Oct 01, 2025 8:00 pm The error messages are:
- 195 Stop overcurrent
- 159 Warn Throttle1
- 160 Warn Throttle1
- 161 Warn Throttle1

Did you calibrate your throttle? Throttle1 error is throttle signal out of range.
https://openinverter.org/wiki/Errors

Thank you very much for the help. I performed the diode test. The positive‑negative terminals show from 2V to zero, and with reversed polarity I get 0.7 V. All three phases show from 2V to zero, and with reversed polarity I get 0.35 V.
I calibrated the throttle – minimum value 414 and maximum 2356.
The motor still doesn’t spin at 50 V.
Error: 192 Stop overcurrent.
What else could I do?

Try bridging the interlock (HVIL). Theres a jumper link provided on board

Hello Johannes, thanks for the tip. Unfortunately I didn’t fully understand it and couldn’t find any information about it. Do I need to take the SDU apart and remove a jumper on the board?

Yes you need to take it apart and ADD a jumper link where it says HVIL DISABLE.

Thanks, I found it. It’ll be some work, but I hope the engine runs again.

Also what's your udcnom set to?

udcnom = 300V

How can I create this .json-file?

Quax wrote: ↑Tue Oct 07, 2025 12:44 pm How can I create this .json-file?

For the parameters? On the web interface there is a button to download parameter file.

I managed the download. Here are my values:

Ok so a few thoughts:

To clarify, the current behavior is that as soon as you start the unit it shows Overcurrent? or is it once you select a direction? or is it once you start moving?

Question for Johannes/other with experience - I have not tested any of my inverters at a lower voltage - should udcnom be adjusted to account for this?

If the inverter is tripping when moving, maybe try changing idcmin and and idcmax? The other current limits all seem within normal bounds, (but again, that's true for higher voltages, not sure how the algorithm is handling the 50V test voltage).

udcnom linearly accounts for changes in voltage in adapts boost and fweak accordingly. It is well possible that this doesn't work down to very low voltages and for testing it is better set to 0 and boost/fweak set to some dialed back settings such as 3000/50

At least on LDU the minimum voltage seems to be around 220, if you go lower on voltage, power is immediately cut to like 25% and you can feel that there is some kind of limp mode/failsafe mode that it enters. Immediately when voltage goes up, power returns. Running SDU with 50V.. I would think that low voltage can cause all kinds of weird stuff..

What I have done so far:
udcmin 0
udcsw 0
Voltage 50 V
Motor rotates
Then motor installed in the car, drive shafts and wheels mounted, car lifted on a jack.
udcmin 0
udcsw 0
Voltage 83 V
Motor rotates with the wheels
Then motor in the car, drive shafts and wheels mounted, car lifted on a jack.
udcmin 224
udcsw 250
Voltage 360 V
Motor rotates with the wheels
Then
udcmin 0
udcsw 0
Voltage 50 V
Motor in the car, drive shafts and wheels mounted, car driven carefully and slowly for 50 meters.
Next day, same test again: relays switching high voltage but Motor does not start.
Errors: Throttle and over‑current.
Then throttle calibrated. 12 V turned off. Next attempt:
Motor does not start. Over‑current error appears after every high‑voltage (HV) turn‑on.
Two relays (pre‑charge and main relay) have been installed from the beginning.
Resistor: 25 Ω, 50 W.
So what I can still test is higher voltage or the tip from Johu with the HVIL-jumper.
I wanted to be careful, so I did the first street test run not with 360 V but with 50 V – that possibly wasn't okay.

I have had overcurrent error that comes immediately when switch direction to D or R when I broke my LDU with regen. Contactors got lifted during regen and one phase got fried. Had to get new motor to "fix" that. Maybe something similar accidentally happened to your SDU?

Yes, it could be possible. The diode test is fine, though. It was just a short ride on a flat street. The power was turned off after it stopped — but who knows?

In my case, diode test failed on one phase. So it might be different for you then..

Quax wrote: ↑Thu Oct 09, 2025 8:20 pm Then motor in the car, drive shafts and wheels mounted, car lifted on a jack.
udcmin 224
udcsw 250
Voltage 360 V
Motor rotates with the wheels
Then
udcmin 0
udcsw 0
Voltage 50 V
Motor in the car, drive shafts and wheels mounted, car driven carefully and slowly for 50 meters.
....
I wanted to be careful, so I did the first street test run not with 360 V but with 50 V – that possibly wasn't okay.

Yeah once it worked with 360V, going back to 50V doesn't seem like a good idea to me. Especially 50V with a load. Power is power, and if it doesn't have the voltage, its gotta draw more amps to make up for it.

Quax wrote: ↑Thu Oct 09, 2025 8:20 pm Motor does not start. Over‑current error appears after every high‑voltage (HV) turn‑on.

I don't know about how the SDU behaves, but on the LDU, when this has occurred for me it has indicated a blown inverter phase.

Then I’ll try it again at 360 V. If it’s already defective, it doesn’t matter.

Every used SDU has already been run at 400 V, so the 50 V test mentioned in the forum is explicitly intended only as a gentle trial, while the unit is stationary at low speed and should not be considered a test drive under any circumstances.

Hi, I am testing a v9 SDU board before installing it into the inverter itself. I tested basic funcs - so turning on with 12V, connecting via CAN to the ESP32 web interface and simple throttle pedal, but no current sensors soldered in or any HV. I used params as supplied. I had to set udcmin to 0 and cruisemode to off and then I was able to get opmode to RUN by doing ign start. Spot values were showing udc to be c. 196V even though no HV connected.

However, it went straight to OVERCURRENT.

I assumed that was because I had no current sensors so I soldered new Melexis ones and repeated but get same error. However I noticed this time that il1 was oscillating between 0 and 14A, and il2 between 0-2A (from memory). I reflowed the current sensors and tried again, and this time I had same behaviour but il1 oscillating beween 0 and minus 14, il2 the same. Screenshot and param file attached.

The behaviour of il1 and il2 being so different make me think there is something wrong but I wanted to check in case the above is normal?

HV Interlock would also result in OVERC message. Bridge it with the jumper

il1 is a but shaky indeed, have you modified il1gain at some point?

I haven't touched il1gain, I'll bridge HV interlock and report back. Much obliged.

So, I bridged the HVIL disable but am having the same issue with immediate OVERCURRENT and odd values from one current sensor. Board is not connected to any HV or the power stage of the inverter.

I did some measuring on the board, using the v6 schematic as a guide although the board is v9 and I don't know what the differences are.

both current sensors getting 4.93V and reading 2.46V at output.
IL1_21 at R10/C5 = 1.682
IL2_26 at R12/C8 = 1.682
IL1_IN at R1/R2 = 2.46V
IL2_IN at R3/R4 = 2.46V

on IC4, pins 5,7, 8 ,10 = 1.682,
pins 1,2,13,14 = 3.723
pins 4 and 6 = 0.206 ?OCURLIM_NEG_59
pins 11 and 9 = 3.1 ?OCURLIM_POS_61

on IC5, low on pins 1,2,4,5
pin 3 = 3.73
pin 6 = 4.9V
pin 8 = 4.9V
pins 11,12 4.9V
13/14 = 4.9V

The fault A, B, C are all low so the PWM_Inhibit is being set high, and maybe that turns into an overcurrent error in the software, but at this point I am guessing. The current sensors seem to be OK though so I don't understand why the IL1 and IL2 are so different. I am at a bit of a loss.

I can't seem to get the the commands to work in the web interface either - I wanted to check IL1 and 2 gains as those values are not exposed. Any help there too would be gratefully received.