Rebuilding Salvaged Model S 75D

[desoukya]

I'm rebuilding my second Tesla Model S; if you're interested in my first rebuild experience, see here. This is a 2017 (December build) 75D with 4k miles.

So the first thing I did after moving the car into the garage was replace the pyrofuse and connect the cut loop. I gave the battery a jump although eventually the dcdc fully charged the 12v battery and the car was operating on its own power. Off to a good start.

Unfortunately though, what comes next is not as glamorous...

I decided to put the car in drive to back out of my driveway and then back-in again. That was fine. Although next, I decided to see if the car chargers, so I connected the HPWC and the car began charging! However, after 5-10 minutes the car stopped charging and I noticed the charge port was red.

I then made the unfortunate discovery the car was extremely low on coolant. So I decided not to touch anything until I installed a new radiator and added G-48 coolant. This is mistake #1.

In the meantime however, I noticed an update from Tesla, so I decided to perform the update. This is mistake #2. Evidently, you should never perform an update on your car if it's still undergoing repairs and all ECU interfacing components are not connected. I suppose this seems obvious in retrospect, but nevertheless something to always keep-in-mind. So, naturally, the software update failed. Although, the update seemed to have flashed some components, but not all. So, I started seeing this error:

Well, this sucks. The car will not drive anymore. In addition, the contactors are not closing and the dcdc is not charging the 12v anymore. Also, asking Tesla to send me an OTA will not work since the car is salvaged. So, I had to rely on my software engineer background and help from some members to gain access to the car and derive a bit more information, but first, I tried the update again from factory/developer mode:

Here are the syslogs.

Some notable references from the logs:
```
gwmon 10261: BMS contactor close request
gwmon 15444: timeout - BMS contactors did not close, ctr state OPEN (0)
gwmon DCDC NOT present, NOT updateable (NO bootId, NO versionMsg, NOT enabled), BMS contactors NOT closed
gwmon chgph1: no hwid
```

As you can tell, the BMS, DCDC, and Charger are not responding, which contributed to the upgrade failure. Here is a list of diagnostic errors I'm seeing:

It seemed suspicious the on-board charger was MIA, so I figured there could be a blown fuse. Will expand on this in the next post.

 

[ACA Man]

Thanks for sharing.

 

[desoukya]

So I started making my way towards the on-board charger. My suspicion was a blown fuse:

I checked the 500v fuse and there was continuity; so it was fine. In addition, all the other small fuses were fine as well. I'm also assuming the cable between the charge port and charger is fine, but perhaps it could be the charge port assembly itself?

This is how far I've gotten. If anyone is able to make heads or tails out of this, please let me know. I'm trying to avoid pulling the dcdc as a last resort because it's behind the front motor and thus not easily accessible; although, I feel it'll probably come down to that.

 

[JRP3]

What caused this car to be a salvaged vehicle?

 

[iCharge]

congratulations on the project and kudos for having the balls to do this !!
although many people may be capable of, I only know a few people . RuchBuilds and Ingineer are 2 names that come to mind

 

[tedsk]

Check 12V fuses. Mia errors are happening because GW can't communicate with particular ECUs. They do not communicate via HV wiring obviously. There was no need in opening charger and checking HV fuses. Check that there is 12V on charger's logic connector. Check corresponding fuse.
Important thing is when it refused to close contactors - before or after update. If it happened after then it is software-induced failure, no point in checking fuses.

You learned your lesson right - never update incomplete car. It won't update normally now, even if you push update as you do.

What state are you in?

 

[desoukya]

Thanks for your reply.

Important thing is when it refused to close contactors - before or after update. If it happened after then it is software-induced failure, no point in checking fuses.

My speculation is after charging stopped abruptly, due to the low coolant, the contactors opened and refused to close. Shortly after, I performed the update but it was too late as the BMS request to close contractors was not responding, which ultimately contributed to the software update failure.

I’ll check for the 12v signal on the charger’s logic connector; do you happen to know where that is exactly?

 

[Kuhz]

What caused this car to be a salvaged vehicle?

Looks like front end damage judging by the hood

 

[tedsk]

It is easy to find - HV cables are orange and thick, logic connector is just a regular harness with a connector that you can unplug by hands.
If you run chargers without coolant it just overheated and shut itself down since it has temperature sensors. It shouldn't blow it and shouldn't cause MIA afterwards. Still there is a chance that you killed your charger.

 

[tedsk]

Can't judge by your tag. Are you in MD? If you are I can help you to recover your ECUs after deployment failure.

 

[desoukya]

Looks like front end damage judging by the hood

Correct, it was in a front collision with airbags deployed. Here is the car after disassembling the front:

 

[desoukya]

Can't judge by your tag. Are you in MD? If you are I can help you to recover your ECUs after deployment failure.

Will send you a PM.

 

[desoukya]

It is easy to find - HV cables are orange and thick, logic connector is just a regular harness with a connector that you can unplug by hands.
If you run chargers without coolant it just overheated and shut itself down since it has temperature sensors. It shouldn't blow it and shouldn't cause MIA afterwards. Still there is a chance that you killed your charger.

I had started to (safely) remove the HV cables because I heard of Tesla technicians attributing charger problems to fualty chargers & HV cables, although I decided to walkback that approach. In any case, in the following image, you’ll notice 4 HV cables feeding into the charger (2 from charge port and 2 from the battery pack). There are also two coolant hoses and a blue connector (not sure what that is). Finally, betwen the coolant hose and HV cable, there is a balck cable. Not seeing the harness you are referring to.

 

[tedsk]

Check the opposite side of charger. I don't see logic connector where HV cables come in.

 

[desoukya]

Check the opposite side of charger. I don't see logic connector where HV cables come in.

Found it:

I also noticed red LED lights on the board. So I think the charger is getting power and hopefully not fried:

 

[nemSoma]

Based on your alerts, the core problem is the charger not communicating with the gateway properly. By the way, you have to put the lid back on the charger, it detects when the lid is off and won't let the BMS close contactors (although that shouldn't cause a charger MIA alert, but a HV interlock alert)

It won't update normally now, even if you push update as you do.

I disagree. Why would that be?
Once the charger is present, the BMS should close contactors and all should go just fine.

Btw, sometimes the gateway or the BMS get's sort of stuck, not wanting to close contactors. If everything's intact on the hardware side, then a gateway reboot should be all that's needed.
(But you have a charger MIA alert, that's different. I'm saying this for when you solved that part.)

 

[nemSoma]

Okay, so after inspecting your full syslog, one thing I can see is that during HWIDACQ it was able to get the HWID for chgvi, so your charger is alive. Now I don't know how chgph1-2 are supposed to respond on US models, but HWIDACQ itself didn't seem to have a problem with the charger. The only thing it failed at was the DCDC and the CMP, caused by the open contactors. And again, those won't close with that lid off

 

[desoukya]

Back in business! Special thanks to @tedsk for his time and help; we were able to re-flash the ecu components that had a version mismatch and redeploy.

 

[desoukya]

The only thing it failed at was the DCDC and the CMP, caused by the open contactors. And again, they won't close with that lid off

Thanks for your input @nemSoma! Contactors are closing now and car is moving into drive gear!

Thankfully this was a software cluster mess and not hardware related. Leason to be learned is never do a software update when the car is not fully repaired yet.

 

[tedsk]

I disagree.

It is your right to agree or not. I worked on this car and I know my job very well.

I got to work with it at the point when it refused to do even hardware acquisition routine which is part of normal redeployment.