Tesla Gen 3 charger board assembly as of mid-2024

[RobCote]

I have to really spell this stuff out so I know I understand. Are the spot values shown in the web interface derived from the can bus? I'm curious because it does show some values, but none of them that would be coming from the onboard charger. So perhaps you're right that the charger is dead. I'm a little hesitant to jump to that conclusion because it came from a car that was in an accident, so presumably it was charged just before that. Of course I could have ruined it by disassembling and reassembling it, but it actually seemed to go just fine, though it was a challenge. I believe everything went back together the same.

Who knows. I'll pick up a can adapter and see what I can find, I guess. Do you have any recommendations?

[tom91]

No, I do not live in the US so no idea what adapters are available that have easy to use software.

Spotvalues come from the controller software. The values that are listed with the numbers are derived from the CAN bus information sent by the charger modules.

[RobCote]

I disconnected can wires from one module, but left the power, ground, activate, and enable wires connected. My main motivation was to see if perhaps the can communications were being disturbed by a wonky connection somewhere, so I removed the 2-into-1 connection and just went straight to the module with an intact temperature sensing capacitor. No change, unfortunately.

However, I measure 2.4-ish volts on each of the can terminals of the unconnected module board. Would that indicate that it's trying to transmit? Or is that meaningless?

[RobCote]

So 2.5V is nominal for inactive CAN bus. A message should be about 2V differential between high and low lines. So it seems like the module bus is coming alive but not sending messages? Am I interpreting the limited data I have correctly? I do realize I need to get a can sniffing tool still.

[tom91]

No point,as you say the resting bus average to 2.5V. Bursts of data does not get picked up by a multimeter. A multimeter averages.

[RobCote]

Gotcha. That's kind of what I figured would be the case, but I was curious and wanted to take a look with what I have on hand.

On the other hand though, does 2.5V tell us that the modules aren't totally dead?

[tom91]

No

[RobCote]

Just to be thorough here, can you confirm the annotations below represent the only connections I need to make on the "external" side connector? Just 12V (2 places), ground, and the CP/PP wires from the charge port? I realized on review that the annotation is a little misleading. The connections are made on the lower connector as shown in the picture, not the top. I included the schematic I've been using for reference so you can correlate the pin numbers to the board.

[tom91]

can you confirm the annotations below represent the only connections I need to make on the "external" side connector?

Yes

[RobCote]

Okay so it seems I am only getting one message from 0x42C over and over and over. It says:

42 BB 00 00 64 00 00 00

does that tell us anything?

[tom91]

0x42C is phase one command message from the controller.

Id this is the only one you are getting even with the leds on on the modules thats not a good sign.

Were you logging with turning the modules on to see if they send something and then fault out?

[RobCote]

Can you clarify for me which module is 1, 2, and 3? If you're looking at the charger with the OEM control board closest to you, is number 1 on the left or the right? I assume 2 is in the middle (in mine it's empty).

The only difference between the two modules at this point (unless I've messed something up without knowing) is the temperature sensing capacitor. I only repaired one so far, because it was a bunch of work and I wanted to see if it'd actually help things before I did it twice. The one I repaired is on the module board on the right (i.e. - closest to the HVDC connections). Is that perhaps why that module is transmitting and the other isn't?

I started logging after I turned the modules on. If I start logging, then power on the modules that's still the only message received.

[tom91]

Numbering does not matter. If it matters to you no way of finding out if they do not send CAN.

Okay, so your modules are not sending any messages.

When the controller is in run mode it should send:

0x45C - Voltage setpoint
0x42C - Module 1 AC current Limit
0x43C - Module 2 AC current Limit
0x44C - Module3 AC current Limit
0x368 - Another message required

No idea how you are looking at the CAN frames, in theory all these should be present with your current controller configuration.

[RobCote]

I'm using this CAN to USB adapter:

https://amzn.to/3Aguclv

Which is running candle, and I have cangaroo on a Windows PC. I'm not sure this is correct, but I have the adapter connected in series with the CAN wires. In other words, the CANL wire from the OI control board goes to CANL terminal on the USB adapter. CANH wire from USB adapter goes to CANH on the module boards.

I completely disconnected CAN wires to the module board that I haven't repaired the broken capacitor on, to remove it from the equation. That's why it's not communicating. I thought working with only one module might help simplify the troubleshooting, but I forgot to write that here.

One thing that's somewhat odd is that across the CAN terminals with the R120 switch on, I measure 239.0 ohms. With the resistor switched off, I measure 81.6 ohms. I'm not sure if that's relevant because I am getting one message successfully multiple times. If the bus resistance was out of limits, I'd expect not to receive any messages.

[tom91]

No idea on that can to usb. So cannot help.

[RobCote]

Well I really appreciate all the help you've offered thus far. I'll continue to poke around. Might try a didn't interface.

[tom91]

https://github.com/jsphuebner/stm32-teslacharger

The code that should be on your V5 board, it should always send the can messages I listed above. No real reason it should not be sending them, specially if you see just one.

[RobCote]

Are you thinking it's possible the open inverter board is faulty?

[tom91]

Are you thinking it's possible the open inverter board is faulty?

No. You are getting one message or so you are saying, I see no reason why you should not see more beyond issues on your CAN adapter end.

[RobCote]

Is there a dbc for the Tesla charger or the control board? I'd like to know what this message is telling me, at least.

[uhi22]

If the control board is the only node and does not get an ACK bit, it may be that it repeats the first message endless with maximum speed. Or should the CAN to USB send an ACK? What is the cycle time of the message?

[royhen99]

with the R120 switch on, I measure 239.0 ohms. With the resistor switched off, I measure 81.6 ohms.

This suggest the CANL and CANH lines are not correctly connected, and there are potentially three 120R resistors in the system. R120 switch should be off and the termination resistors at the controller and charger. Use a short length of twisted pair to connect the CAN /USB adaptor to the CAN bus.

[RobCote]

If the control board is the only node and does not get an ACK bit, it may be that it repeats the first message endless with maximum speed. Or should the CAN to USB send an ACK? What is the cycle time of the message?

Most of the lines have a delta T of 0.0003s

[RobCote]

This suggest the CANL and CANH lines are not correctly connected, and there are potentially three 120R resistors in the system. R120 switch should be off and the termination resistors at the controller and charger. Use a short length of twisted pair to connect the CAN /USB adaptor to the CAN bus.

Should it be connected in series as I have it, or do I need to put it in parallel? It's not well documented.

[royhen99]

Parallel so the resistance measured across the bus is 60 ohms. It's CAN bus, which is very well documented. Just look at the diagrams in wiki page. https://en.wikipedia.org/wiki/CAN_bus