Error messages abound, parking brake stuck, and a 5000 lawn ornament. HELP!

[geordi]

My last quote was a search result. Adding "weather" after seeing Nate75D rain issue.

Right, and I had seen that - I didn't have an issue with the premise that the parking brake had errored out b/c of water intrusion from a leaking o-ring.

The problem with that is with the motors disconnected, the problem SHOULD have been confined to an error about the parking system. That's what a logical person would think at least. The thought never occurred that with all the modules connected properly, and only the parking brake motors disconnected, that the car's communication system would go completely haywire.

Why would ALL the communication in the car depend on two cheap Chinese motors being connected in parallel with the rest of the wiring? That makes no sense. Hence all the problems diagnosing this, b/c the errors were intermittent and seemingly unrelated to the ONE known issue that had started it all.

Even now, I'm still not 100% that replacing those motors will solve the rest of it, b/c everything has been so intermittent.
Heck - on my MX, I had an isolation error a few days ago, that ALSO went away shortly after, and has not reappeared. That's ALSO not a comforting thing, b/c it can (and should) brick the car if it is an actual safety hazard.

 

[NewbyMaybe]

Now I know what I'll be doing before it's raining every day here in Florida.

 

[mongo]

No, the last software update still didn't complete, but there was still an error (parking brake) present in the system at that point. I am going to give it ANOTHER try today now that all the errors are cleared.

POSSIBLE SOLUTION:
I spoke with Electrified Garage (VERY helpful, thank you EG!!!!) and it seems that the root cause MAY JUST BE that stupid parking brake motor design. They said that there have been a lot of problems with water ingress with those and the 4 wires in the plug on the motor are: 2 for the motor itself, and 2 for a hall sensor that has a 5 volt reference source. Here's the key to the whole thing: That 5 volt reference is FROM THE SAME SOURCE AS THE CAN BUS.

So if water gets in it and shorts / drains that 5 volts to ground, that would ALSO be draining the CAN bus!!! I think maybe that's the answer to the whole thing, b/c it's the only commonality in the whole thing.

I've now ordered the motors from Ebay / China because EG didn't have any on hand either (said they couldn't keep them in stock, but he didn't know about the Ebay source) and said that they also are potentially not being available anymore outside of Tesla... So I don't know if he's going to pick up any for their needs, but I ordered a couple extra for me since I have three vehicles that all use these. I feel good about finally having a probable answer though. I won't know if this IS the solution for at least 2 more weeks, because I have to go to California while these are coming from China.

I will come back and update the thread afterwards though. I greatly appreciate everyone's posts and assistance in chasing this down.

That would be some some bad engineering.
External power rails should be current limited/ protected and CAN ICs should not load the bus if un or under powered.

 

[NewbyMaybe]

Edmond Tong has a good vid on YT for removing locked pb servo.

 

[geordi]

Now I know what I'll be doing before it's raining every day here in Florida.

Yeah, I have three Teslas in the same generation that use these.... So I'm right there with you, b/c I'm also based in South Florida. Rain is a thing that happens to other places, Florida gets airborne floods. Water intrusion is nearly assured.

Electrified Garage said that the parking brake controller is one of the "master" controllers on the CAN bus, so that should put it in the same general importance level as the drive inverter, gateway, and pedal monitors.

In a logical world, that EPB module SHOULD self-isolate the circuits to the motors when they go open-circuit, or there should be a diode isolation between the module and the rest of the system so something like a dead short to ground doesn't take out the upstream electronics. But then that might require some designers who don't know which crayons taste the best.

In other news, I have not woken the car today or driven it yet, but I did park near it and got my wifi up and broadcasting (I have to bring a hotspot to this location b/c rural) and the car seems to have woken itself up and picked up the wifi - I got a notification on my app that the software update was ready to install, and this time it actually completed. So it seems that if a module is erroring out, then you don't get the software updates.... But if all is well (as I left it last night) then everything works fine.

It has also been raining most of the afternoon so far, and clearly the car is NOT freaking out about that, so it seems that either the water intrusion only happened the last time my sister drove it, OR I jostled it loose in faffing around with the EPB motors. Going to give it a test drive in a few minutes and see what happens. Still not sure if she should try driving it again before the replacements are installed.

 

[mongo]

Yeah, I have three Teslas in the same generation that use these.... So I'm right there with you, b/c I'm also based in South Florida. Rain is a thing that happens to other places, Florida gets airborne floods. Water intrusion is nearly assured.

Electrified Garage said that the parking brake controller is one of the "master" controllers on the CAN bus, so that should put it in the same general importance level as the drive inverter, gateway, and pedal monitors.

In a logical world, that EPB module SHOULD self-isolate the circuits to the motors when they go open-circuit, or there should be a diode isolation between the module and the rest of the system so something like a dead short to ground doesn't take out the upstream electronics. But then that might require some designers who don't know which crayons taste the best.

In other news, I have not woken the car today or driven it yet, but I did park near it and got my wifi up and broadcasting (I have to bring a hotspot to this location b/c rural) and the car seems to have woken itself up and picked up the wifi - I got a notification on my app that the software update was ready to install, and this time it actually completed. So it seems that if a module is erroring out, then you don't get the software updates.... But if all is well (as I left it last night) then everything works fine.

It has also been raining most of the afternoon so far, and clearly the car is NOT freaking out about that, so it seems that either the water intrusion only happened the last time my sister drove it, OR I jostled it loose in faffing around with the EPB motors. Going to give it a test drive in a few minutes and see what happens. Still not sure if she should try driving it again before the replacements are installed.

FWIW:
Diode would only help against reverse polarity (backfeeding). Best thing is a voltage tracker or other current limited switch to handle low resistance faults. Open circuit on a power feed shouldn't break anything.

CAN doesn't have master nodes. It does have termination nodes, but that is usually just a passive resistance. Gatway is an important node, but still not in control of the bus. EPB data is used by other modules (as shown by error logs), so in that sense it's important.

CAN is driven by transmitting nodes, otherwise CAH_H=CAN_L=2.5V or so.

 

[geordi]

Huh, I know the main function of a diode is a one-way power flow gate, like in a rectifier bridge that stops AC and makes it DC. But I thought there was a way they were used to open if there was an imbalance from one side to the other like in a short. Dunno what I was thinking of then, but this is why I'm more of a hardware hacker. I haven't done circuit building work in MANY years since messing around with Heathkit stuff with my grandfather in the 80s.

It does seem like on a $100k vehicle based entirely on voltage and circuits, there should be more concern about module protections from unexpected circuit surprises. As far as the CAN communication, EG were the ones saying it was a master, I don't know that side of things.... But clearly the computers weren't happy about SOMETHING b/c the firmware refused to take without everyone playing nicely.

I learned a bit about that side of things (installing updates) with my convert-2016-X-pedestal-to-folding-seats project, and my X has the expected "seat unlatched" errors... BUT the computers otherwise must be happy with themselves b/c there hasn't been any issue with updates there either.

 

[mongo]

Huh, I know the main function of a diode is a one-way power flow gate, like in a rectifier bridge that stops AC and makes it DC. But I thought there was a way they were used to open if there was an imbalance from one side to the other like in a short. Dunno what I was thinking of then, but this is why I'm more of a hardware hacker. I haven't done circuit building work in MANY years since messing around with Heathkit stuff with my grandfather in the 80s.

It does seem like on a $100k vehicle based entirely on voltage and circuits, there should be more concern about module protections from unexpected circuit surprises. As far as the CAN communication, EG were the ones saying it was a master, I don't know that side of things.... But clearly the computers weren't happy about SOMETHING b/c the firmware refused to take without everyone playing nicely.

I learned a bit about that side of things (installing updates) with my convert-2016-X-pedestal-to-folding-seats project, and my X has the expected "seat unlatched" errors... BUT the computers otherwise must be happy with themselves b/c there hasn't been any issue with updates there either.

Diodes are handy to combine two (or more) power feeds such that, if either feed fails, the module still works.

In standard automotive testing, outputs are typically exposed to open, short, and variable resistance conditions. Possibly also 12V.

I know you were relaying EG info, just adding in some other data.

I think a car needs all modules reporting properly (and possibly no logged serious faults) to do an update.

FWIW, yhe new S/X have the EPB ECU functions built into another body module and don't appear to have feedback sensors.

 

[ewoodrick]

No change. It isn't the 12v battery, stupid thing is STILL disabled with the parking brake locked. Will not shift gears because one of the errors is "gear shifter CAN node MIA" which is pretty much the error from EVERY module in this thing apparently.

So what now? after much fighting with it, I finally managed to get a service ticket in - and there's no "leaving it in the driveway" because OF COURSE this problem requires it be brought to the service center, more than 60 miles by road away. If I can get the wheels off and disable the parking brake, I can't even trust that the stupid thing will let me move it under power, because it can't communicate with itself.

Horn doesn't work, gear shifter doesn't work. Round knobs on the wheel work fine. Things that I can do in the MCU seem to work. Anything outside of that is not working. Is there a fuse for a CAN gateway maybe?

How dd you determine that it wasn't the 12V battery? What voltage is it showing?

 

[geordi]

The battery was showing 13.2 volts when removed and under no load. More importantly, changing the battery had zero effect on the other issues.

 

[math2999]

I'm having similar issue with 2015 MS - see screenshot. Happens daily on start up and goes away after a few minutes. Tesla SC have given up diagnosing ! Does this look similar ?

 

[geordi]

Yeah, that looks pretty similar - especially the last one on the list of the parking brake needing service. Then everything else losing its mind after that, is likely much the same situation I have.

Don't tell the idiots at the service center, they won't pay for the knowledge and don't deserve to know. But maybe we should file NHTSA complaints about this EXTREMELY poor design that links the power source for ALL the data lines, to something which is subject to water contamination. When it gets wet and shorts that sensor to ground, then everything else goes haywire.

Right now, here's what you need to do. It's not hard if you know which end of a wrench to point at the car.
Jack it up in the back, and pull the rear wheels. You will see the two different calipers, you only care about the smaller ones. There are TWO T-25 screws that hold the motor to the gearbox and THREE that hold the gearbox to the caliper body. You want to remove the THREE. The motor will come right off then. Clean that O-ring, although if you are already having issues.... Maybe don't worry about sealing it up right now.

This is what I ordered, haven't installed them yet b/c I'm on another job. Parking Brake Caliper Motor For 12-21 Tesla Model S/X 1621620888C 40C07814 | eBay

Since my X was not malfunctioning, before I left on this trip I pulled the motor and gearbox the same way I just told you how, used a Q-tip to push some basic grease into the gearbox spaces, and sealed both the motor-to-gearbox and gearbox-to-caliper with a thin layer of RTV silicone.

It's hard to prove a negative, but on a VERY wet trip so far, it has not caused any issues with more mileage on these than on that 2013 S that caused this whole thread.

 

[geordi]

Coming back to this thread with an update - After a month on the road I finally got back and was able to swap the EPB motors. Observations:

The original gearboxes were in a metal body, probably aluminum. The replacements (and what I found on my X) looked like the gearbox housing or at least what mates with the o ring and the caliper body, is ABS plastic. So a different design. The interface between the motor and the gearbox is still an o-ring. Both new motors were disassembled and the seals augmented with a layer of RTV silicone. There was a healthy amount of grease in the new motors, so I’m happy about that.

After replacement, the car behaved properly for multiple days and did not error at all even going through a vigorous car wash. But has since tossed the “parking brake did not function car may be free rolling” a couple times with no obvious moisture cause. Each time the error cleared on the next drive. Don’t know if this is just a problem with Chinesium parts and a buggy sensor…. Or something more is still going on. SO FRUSTRATING to try and chase intermittent issues!

 

[PromisInterior]

Hello, I have similiar problem.
Since i bought Model S P85 2013 i have had errors for handbrake or regenerative braking temporarlly reduced, usually after car wash but it always came back after 5 minutes.
But now I have had car sit for 2-3 days with DRLs and screens on because errors GTW u014 ; GTW u099 ; GTW w027, EPAS W002 ; ESP W002.

I have checked both handbrake motors, everything was clean, now i would check module and wiring, can somebody give me ideas where to start checking wires and how to check module?

Important, both wheels were locked(parking brake did it's job)
I could not switch from N to D, R or P.

 

[geordi]

If it has to do with the parking brake motors, unplug them and see if the errors resolve soon after. The sensor in the motors relies on a 5 volt reference that is also the reference source for the entire communication bus, so if that is shorted to ground then the communication bus fails. You have many errors (and the names would be helpful here) which seems suspiciously coincidental that all those things could simultaneously fail. That says communication failure instead.

 

[PromisInterior]

If it has to do with the parking brake motors, unplug them and see if the errors resolve soon after. The sensor in the motors relies on a 5 volt reference that is also the reference source for the entire communication bus, so if that is shorted to ground then the communication bus fails. You have many errors (and the names would be helpful here) which seems suspiciously coincidental that all those things could simultaneously fail. That says communication failure instead.

I already did everything, checked fuses, I tried new motor, reinstalled software,

At first, cars pinned motors but now nothing happens even when i tried new motor.


Is there way to check wires?

 

[NewbyMaybe]

Did you check the pads? If they are dry and crusty then sometimes they won't move.

 

[PromisInterior]

Did you check the pads? If they are dry and crusty then sometimes they won't move.

Car does not react at all. Motors are removed and conected

 

[NewbyMaybe]

The pads need a small layer of grease on the outer edges so they can move inside the caliper. They will sometimes require too much force from the system and that triggers the errors. After sitting the pads will stick to the brake rotors and require more force to retract.

 

[PromisInterior]

The pads need a small layer of grease on the outer edges so they can move inside the caliper. They will sometimes require too much force from the system and that triggers the errors. After sitting the pads will stick to the brake rotors and require more force to retract.

I understand, but both motors are detached. I assume that wiring is broken somewhere, but I don't know how to check.