Following this approach I found that using testchan 11 on a good BMS I get around 3.7V between connector pin V10 and then probing along V11 I get around 6.6V at Q77 onwards.

I guess this shows a failed fet at Q77 (rather than Q81) which is the next fet along and also gives 6.6V.

So check to check my approach. If I use a known good BMS, set testchan to 10 and measure voltages between pin 9 (negative) and the fets on channel pin 10 (positive) then compare the same setup with BMS7 (with the faulty readings on U10 and U11) would this work? What should or shouldn't I expect to ...

Any chance of some rough dimensions of these? Cheers

So I had some more time on this. Looking at BMS7 I get erroneous values for U10 and U11 (-1.15625mV and 1.15625mV respectively).

Firstly I took a BMS giving good results (BMS3) set testchan to 1 and took readings across every fet terminal from V0-V16.
I then swapped the board for the above BMS7 and ...

Yes im more than happy to try and get to the bottom of the issue myself, unfortunately my knowledge in this area is pretty limited but I'm enjoying learning.

I need to check what is and what isn't working on BMS4 again following this morning events. I'll try again this evening.

I'll check temps ...

So something that might be considered progress...

I connected a working BMS (BMS1) up to 16 cells, set testchan to 1 and measured voltages between V0 and the bottom row of the fet line starting with Q5. Voltage patterns made sense.
I then connected a non-working BMS (BMS4), this board usually gives ...

No, I'll look into it.
So set test channel to any channel I want to check and then manually check voltage on the relevant fets with a multimeter? Or is there something more sophisticated at play?
I did look but could not find any information on the forum or wiki documenting how specifically to use ...

I did some probing with one of the BMS units hooked up to 16 cells and can verify voltage was getting to the FETs and noticed slightly different voltage readings across legs of the fets. Could this be a tell tale sign? Is it worth checking voltage at each Fet leg for signs of issues?

So I psyched myself up to try and identify the issue again this evening, given the rain....
I jumped directly from the connector pin for U2 and U3 to the resistor for the relevant channel but it made no difference to the voltage readings on the webUI....
Any other suggestions on how to test this ...

Delete any .json files from the files section

Put a low ball offer in on a 51kwh pack that got accepted on one if these cell to pack Rubiks batteries.
Made by Saic I believe.
Does anyone have any information on them? Pinouts or anything similar?
Thanks

Glad to see the steel toe cap flip flops in action. What are the pros of this arrangement?

How much?

I believe they come preprogrammed but if not Stmcube programmer and stlink v2 will get you there. There's usually a 3 or 4 pin terminal pad near the stm32 on the board that you can solder in a 3 or 4 pin header to connect the stlink to. Usually labelled VDCG or similar (voltage, data?, clock, ground)

What was the issue? It's always helpful to publicly document this to help others with similar issues.

Whereabouts are you for collection?

Amazing, I had planned on a gs450h into a Mercedes 310D that shared a lot of components and qualities (loud and slow) with these vans but the rust was too much work for me so I sold the van and replaced it with a 2005 VW T5 that im electrifying. I look forward to seeing the results with the GS300H ...

Europa46 wrote: ↑Tue Oct 07, 2025 4:56 pm Thanks, I know you are busy and appreciate the reply. I will download Freecad and have a look.

Mark

freepcb not freecad.

ok, after a lot of faffing I've managed to get the code to generate the CRC correctly.
I dont have access to my van or Can logs of my van moving at the moment but this translator appears to work to translate VW speed signals transmitted on 0x288 to MQB ESP 21 0x0FD and correctly calculate the ...

Can you swap the phase wires iver and see if mg1 inverter can drive mg2 and vise versa? Or try the resolver plugs?

ok, so it looks like I've got the Can filter setup and running with the help of BigPies/RStevens CanBridge board and.... :o ChatGPT/ClaudeAI :o ...
https://github.com/jamiejones85/ESP32VWMITM
just bench tested it but playing a canlog from my VW T5 back at the Can filter hardware and then listening ...

Never commented on this. Never seen channels fail that way. "Normally" some FET fails shorted and skews all other readings.
Are you absolutely sure u2 and u3 voltages make it to the board?



Used the next best online json formatter, then wrapped it in 'code' tags 8-)


Thanks, I'll double check ...

Right im revisiting this. So what would be the recommended method for transmitting the checksum in a live translation from my VW T5?

Probably need a location

Ok thanks, I can only really assume mine are OK....