Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

TMS P85 motor humidity: high voltage câbles seal leak, error GTW_w176

This site may earn commission on affiliate links.
Hello everyone,

Here is my problem:

26/12: my Tesla model S gave me error messages: GTW_w018 and w157: reduced electrical power; GTW_w174 and w177: provide battery replacement based on voltage (12v); regenerative braking off; GTW_w017 reduced power.

But she continued to drive.

Then on 28/12: it drives, I stop to buy bread and when I want to restart, then it no longer starts: DI_w127: parked carefully, BMS_f027 and GTW_w176: impossible to start.

It is no longer guaranteed, as it was registered in 2014.

Call from the tow truck, direction the Tesla garage.
The garage told me that it had disconnected the + and - of the motor (power side) and discovered moisture there. This is due to the seals on the DC cables which are HS.

By the way, the courtesy vehicle at Tesla is free during the warranty period, then it's 130€ per day. I should have taken advantage of it more often when it was guaranteed.

To change the seals, the cables must be replaced because they are crimped at the factory. The seals are not replaceable because they are blocked between the crimped terminals. (1,000€)

In addition, the humidity having entered the engine (rain), the engine block must be completely replaced. Standard exchange. There is a current leak, visible on the megger (mega ohm meter). Motor insulation fault which causes all the faults mentioned above to appear. (6 500€)

So an estimate of €7,500.

I ask them: is it possible to try to let the engine dry
Answer: yes, we can try but if it works, I will have you sign a liability waiver in the event of electrocution because there has been water observed in the motor. I say ok.

The estimate is updated to 1500€ for the cables, I validate.

A week later, he calls me: still the same, current leakage fault. In addition, the estimate and redone: €8,000

I ask where the extra €500 comes from: answer: we have to look. Do you still have the quote for €7,500? we no longer have the initial (dematerialized) quote.
No luck, me either because I did not print it.


I ask to see the car in the workshop. I note that there is nothing openable on the motor to dry it, apart from 2 holes DN 20 (the penetration of the + and - cable). There is an inspection hatch (~10cm by 5cm) in the center of the engine block but it has not been opened… (it does not help drying in my opinion)

I ask if a second-hand engine is possible: answer: it is already a reconditioned engine in the estimate.
Surprise on my part.
But hey, the quote is in English (Tesla Lille), it may have been written but I'm not good at English.

So, I ask to have my engine reconditioned. Answer: it is an operation carried out in Amsterdam but only for Tesla, internally. Not for customers. At the garage, we only do standard exchanges of the entire engine block.

Info: A reconditioned engine is guaranteed for 4 years.

I indicate that: My engine block was replaced 2 years ago, free of charge because the car was still under the 8-year warranty. Answer: He tells me that it was also a reconditioned engine but that it does not benefit from the 4-year warranty because it is only the remaining time of the car under warranty that is taken into account.

This is where I am.

I'm thinking of bringing the vehicle back to my house to actually try to dry it out.
I hope to be able to send a flow of warm air into one of the DN20 holes + put a heating blanket over the entire engine block for 1 week.
I also need to find a megohmmeter to be able to test before-during and -after.

When do you think ?
 

Attachments

  • DBE935F9-D77E-4031-B017-8C588BAFBF0C.jpeg
    DBE935F9-D77E-4031-B017-8C588BAFBF0C.jpeg
    423.1 KB · Views: 1,125
  • F3AC4050-8159-4149-900E-36DF55113FA5.jpeg
    F3AC4050-8159-4149-900E-36DF55113FA5.jpeg
    396.8 KB · Views: 260
  • D562FA4D-FD93-46D3-867D-D216057D5716.jpeg
    D562FA4D-FD93-46D3-867D-D216057D5716.jpeg
    396.7 KB · Views: 249
  • 411F05F8-9A73-4B86-9AD4-4F8FF04E45A3.jpeg
    411F05F8-9A73-4B86-9AD4-4F8FF04E45A3.jpeg
    381.7 KB · Views: 252
I'm also surprised that water is able to fit inside. Just by the joints of the HV cables?
No, there was no flooding or submersion. There is just the current winter rain here.

There is a slight corrosion on each side of my O-rings. Not big.
Not as big as the Gtech example (in post #9)

View attachment 896345

I still think it could be coolant (most common moisture intrusion problem). Pull speed sensor and see if have coolant seal leak.

The technician opened the inspection hatch in the middle: it's dry.
The technician says that the water may not have reached the middle.

middle hatch is for unbolting 3 bus bars from inverter to stator. Its slightly higher than inverter but still very low.

if someone here is used to repairing LDU with an insulation fault, if it is a breakdown (and not from coolant or humidity): what needs to be replaced inside, most often? electronics? motor windings?

Tesla LDU - Failure Testing (google.com)

Maybe need to dry out windings. Probably replace inverter harness (only salvage available now, I'm planning to develop new harness source eventually). Probably clean up inverter control board and casing.

If speed sensor shows coolant seal leak, replace seal and prepare shaft seal surface + install seal carefully (I will add dedicated page on LDU rebuild website soon on this since its so important) Maybe also regrease/change rotor bearing. If no leak, then maybe consider not changing since only 2 years old LDU? Its a gamble.

Tesla LDU (google.com)
 
Last edited:
Upvote 0
Thanks. I salute your experience and your desire to pass it on. I continue to immerse myself in it and I am actively looking for a mechanic, near me, who can help me carry out your operations.

I have some additional information:
- The car was registered in March 2014. (First registration)
- The LDU was replaced in February 2015 for a clicking noise. (New / reconditioned ? Miles ?)
- The LDU was replaced in July 2020: broke down at full throttle (130km/h) on the motorway (was new LDU / car have: 100,000 miles)
- Now: (130,000 miles)

The Tesla garage tells me that the price of a new LDU is 2x more expensive than the price of a reconditioned LDU. This while they will have the same guarantees: 4 years / 50,000 miles.
I still asked for both quotes. I am waiting to receive them.

The car is coming to my house tonight.

I will send an email to Tesla to report my dissatisfaction:
Tesla reconditions LDU to recover from failure. It is therefore that there is an economic interest in reconditioning LDU. Why can't the LDU be directly reconditioned (repaired) for the customer?
 
Upvote 0
What is the actual 12v voltage? The MCU may not turn on if well below 12v (i.e. shorted cells in the 12v battery). I'd expect once the voltage is high enough the MCU should boot up. You might also try a reset (press both scroll wheels in for 15 seconds or so). The boot process typically takes between 60-120 seconds.

As for Tesla doing DU repair on specific DUs at the service center, they are just not equipped for it. I doubt any EV vendor's service centers will do DU work in-house. They are sent to the factory, where they are pulled apart and evaluated, parts replaced as needed, sealed backup, and tested on the bench likely with custom equipment (i.e. very expensive). Testing will be done for current leakage, power output, speed, vibration, and pressure as they do for all new motors. Now they could in theory take your DU out, ship it to the factory, repair it, and then ship it back. Depending on the location, this might take several months. I doubt many customers want to wait that long. Shipping a 300 lb DU means it will go by the cheapest means (i.e. slow).
 
Upvote 0
Would there be any feedback on:
approximately how much time can elapse between the moment the seal leaks (coolant appears on the speed sensor) and the electrical insulation fault (alerts on the dashboard) ?
Difficult to say. I noticed coolant on my speed sensor and did a rebuild half a year later. There wasn’t any condensation in the stator or inverter (and my revision has the hole between reluctor housing and stator).

There are probably other cases where the coolant needs less time to travel to the stator and inverter.
 
Upvote 0
Hi,

I am trying a (provisional) solution that is accessible to me. I am a beginner handyman. I attach the photos.

The goal is to dry out the rotor and stator, without removing the LDU.

I leave the radiant heater and the vacuum (smoke pen shows airflow) cleaner on for 4 or 5 days. If the megohmmeter indicates that the fault is gone, I hope to be able to restart the vehicle.

I will also drill out the drain hole (emergency drain #1) so the coolant doesn't go back into the stator.


My end goal is to buy time to find a re-builder for the LDU and to be able to bring the vehicle to him without a tow truck.
 

Attachments

  • AF76E2D1-3202-480E-AF9E-5B788C960C64.png
    AF76E2D1-3202-480E-AF9E-5B788C960C64.png
    2.2 MB · Views: 279
  • AFF36597-1968-4B3D-A0DC-49C0E6D28BB1.jpeg
    AFF36597-1968-4B3D-A0DC-49C0E6D28BB1.jpeg
    1.2 MB · Views: 145
  • 0B9BE208-9ED1-467A-A766-7AFEB806D3B6.jpeg
    0B9BE208-9ED1-467A-A766-7AFEB806D3B6.jpeg
    166.8 KB · Views: 138
  • 4B1F47BF-8B92-4027-B2FD-A911202C98CE.jpeg
    4B1F47BF-8B92-4027-B2FD-A911202C98CE.jpeg
    832.4 KB · Views: 131
  • B1FD7AD7-46EE-4FEA-94BC-38EA55067D13.jpeg
    B1FD7AD7-46EE-4FEA-94BC-38EA55067D13.jpeg
    554.5 KB · Views: 133
  • 3434ADF6-17F7-48D1-82C2-4A132BECD742.jpeg
    3434ADF6-17F7-48D1-82C2-4A132BECD742.jpeg
    462.1 KB · Views: 132
  • B44D7FE0-88ED-4166-93C7-86669E8BF9D8.jpeg
    B44D7FE0-88ED-4166-93C7-86669E8BF9D8.jpeg
    548.4 KB · Views: 142
  • E2DE44B1-65EA-4471-82D0-FAD4A936C376.jpeg
    E2DE44B1-65EA-4471-82D0-FAD4A936C376.jpeg
    490.9 KB · Views: 145
  • 69F9AB5F-B3AF-4FBF-B426-6B5C037636B1.jpeg
    69F9AB5F-B3AF-4FBF-B426-6B5C037636B1.jpeg
    500.6 KB · Views: 136
  • A3DFEDCF-A5C1-443C-AD7D-E0EDD6BAB840.jpeg
    A3DFEDCF-A5C1-443C-AD7D-E0EDD6BAB840.jpeg
    718.8 KB · Views: 148
Upvote 0
Yikes, that’s a lot of coolant in the stator. I would expect that there’s coolant in the inverter also. I would definitely keep the 12 V disconnected unless necessary for testing. When there’s a voltage applied to the PCB and wiring, it will accelerate corrosion.

Did you measure isolation resistance of the stator? The isolation will fail if they are soaked in coolant long enough (possibly beyond repair).
 
Upvote 0
I removed 4 syringes.

10x4= 40ml of coolant on the back of the rotor.



The car is currently raised on the driver's side: I just passed the cotton swab to the lowest of the BUSBAR (3 holes in the center of the LDU): No humidity and no blue color. I put the cotton swab at the lowest in hole B-: No humidity and no blue color.



Also, the first alert is displayed on the trip computer while I was driving. I kept driving. I stopped to go buy something. I came back 5 minutes later and the car wouldn't start. (Secured).



From these fact, it seems to me that the inverter is intact. I hope.



The 12v and the high voltage have been disconnected for 3 weeks, when the car was put in safety.



The mechanic Tesla had shown me his megohmmeter under 400~500V. From memory it was 0.4 (M ohm, K Ohm?). It was weak.



A friend lends me the megohmmeter tonight. I will tell you the exact value tonight between B+ and mass / B- and mass. The engine block is at 45° C with the small airflow inside at the moment.
 
Upvote 0
Would there be any feedback on:
approximately how much time can elapse between the moment the seal leaks (coolant appears on the speed sensor) and the electrical insulation fault (alerts on the dashboard) ?

My leak's evidence is similar to @Wim_V, rotor and stator both have some rust suggesting leak has been for awhile (maybe > 1 year) but don't really know for sure.

On coolant getting to inverter. The vent hole in speed sensor chamber likely plays a key role. Reluctor wheel will agitate and aerosolize the pooled coolant and travel by air to the inverter. It will leave no trace on the path... just pooling and collecting on the low spot in the inverter which is usually under the CANBUS connector. See pics here for mine


However, people have now reported LDUs without the speed sensor vent hole. In this case, coolant tends to get past the outer rotor bearing and enter the stator and soak the stator. This member had this situation. His inverter had no coolant but stator windings were soaked. Stator low isolation detection stopped his car.

Tesla Large Drive Unit (LDU) Motor Teardown and maintenance - Page 8 - openinverter forum

There are reports of RevQ an RevR Sport without this vent hole. Maybe you also don't have vent hole? Can you check and also provide LDU sticker image for revision details?

Tesla LDU - Revisions (google.com)

Hi,

I am trying a (provisional) solution that is accessible to me. I am a beginner handyman. I attach the photos.

The goal is to dry out the rotor and stator, without removing the LDU.

I leave the radiant heater and the vacuum (smoke pen shows airflow) cleaner on for 4 or 5 days. If the megohmmeter indicates that the fault is gone, I hope to be able to restart the vehicle.

I will also drill out the drain hole (emergency drain #1) so the coolant doesn't go back into the stator.


My end goal is to buy time to find a re-builder for the LDU and to be able to bring the vehicle to him without a tow truck.

A couple of thoughts

I posted this idea on leak mitigation with LDU in the car. Maybe solution A and B in diagram below are both possible. I would suggest having some very precise drill depth measurements and practice on aluminum blocks before proceeding. I don't know if can easily reach the drilling location as its hiding under the manifold body. May need a long drill bit.


Screenshot 2023-01-23 at 11.46.20 AM.png

I guess you need to worry about metal shavings in the stator chamber since you drilled a hole. But I think its aluminum (Please check with magnet) so probably no reaction with magnetic field. However, metal shaving dust inside high speed rotation chamber probably can cause some damage. Maybe some will penetrate the bearing seal and destroy the bearing race? I don't know. Unless you are sure metal shaving didn't reach inside, you may need to remove LDU to clean up :(

I am pessimistic on drying out stator without removal. Even fully opened under hot lamp 24 hours a day, the member in above link took 1+ week. He said he can smell coolant every day as they seep out of the micro passages inside the stator winding varnish.
 
  • Like
Reactions: OttoPV60 and Gtech
Upvote 0
Hi,


Here is some news.

The values measured daily with the megohmmeter are still very low. The Needle is close to 0.
The values are identical, regardless of the measurement point (B+, B-, 1, 2 or 3 of the BUSBAR).
Here are the values measured with a simple multimeter, selector set to 20 Mohm:

Monday :
Black (COM) to Ground: 6.3
Red on ground: 3.6

Tuesday :
Black on Ground: 4.4
Red on ground: 1.1

Wednesday :
Black on Ground: 12.5
Red on ground: 2.6

Thursday :
Com on ground: 6.3
Red on ground: 1.5

Friday :
Black on Ground: 10.4
Red on ground: 2.60

The results are variable.


There is no hole (on the back of the speed sensor) between the reluctor and the stator. The reference and T19… (see photo) It was installed in July 2020.

It is aluminum, the magnet is not provided.

There doesn't seem to be any rust on the speed sensor. (Photos taken during the first opening attached)

The standard exchange with a refurbished (or even new) LDU at Tesla does not seem interesting to me. Indeed, the LDU will be guaranteed for 4 years. It will therefore be impossible for me to make any changes to it. And, if it starts filling up with coolant after 3.5 years… game over.

I installed a temperature regulator on the rotor to maintain it at 45°c permanently. Is this temperature sufficient?

I think to maintain the ventilation and the heating of the rotor (without opening it) for 2 weeks. (I will have tried)

I want to charge the HV lithium battery. Can I reconnect all the control connectors (disconnected by Tesla garage under the rear seats) and not reconnect the power to the LDU (B+ and B-) in order to recharge the HV battery?
 

Attachments

  • 1E301BC2-D4D9-4228-B8A9-1DFAB398AC70.jpeg
    1E301BC2-D4D9-4228-B8A9-1DFAB398AC70.jpeg
    973.8 KB · Views: 145
  • 1DCC6017-EA73-4C9E-B41A-17E4AEC394C8.jpeg
    1DCC6017-EA73-4C9E-B41A-17E4AEC394C8.jpeg
    898.1 KB · Views: 138
  • 478D1743-C08F-4A07-AB5E-B9051B58E023.jpeg
    478D1743-C08F-4A07-AB5E-B9051B58E023.jpeg
    322.5 KB · Views: 132
  • B1ABF2D1-CD28-428E-934A-26806E318580.jpeg
    B1ABF2D1-CD28-428E-934A-26806E318580.jpeg
    585.4 KB · Views: 133
Upvote 0
The values measured daily with the megohmmeter are still very low. The Needle is close to 0.

Here is a video of low and proper isolation resistance. Turn on CC + auto translation to understand what they say


Here is Tesla's isolation fault diagnostic procedure. Unfortunately I don't know what should be the spec. Above video showing ISO cleaned and dried stator seems to be showing 4M ohms.

June 2013 Report.pdf (nhtsa.gov)

I think simple multimeter probably doesn't tell anything for HV isolation. Need to inject 250-500V (depending on HV device) to check isolation.

There is no hole (on the back of the speed sensor) between the reluctor and the stator. The reference and T19… (see photo) It was installed in July 2020.

It is aluminum, the magnet is not provided.

There doesn't seem to be any rust on the speed sensor. (Photos taken during the first opening attached)

T19 = 2019 initially manufactured drive unit. No REMAN on label so not remanufactured unit.

1002633-01-T is the latest revision.

No vent hole behind speed sensor chamber means all leak went to stator and didn't travel by air as aerosol to inverter. So if you have moisture in the inverter (I still guess it must be coolant) maybe enough coolant leaked into stator and traveled by busbar tunnel to the inverter. Another potential worry is if any coolant leaked past gearbox oil seal into the gearbox.

The standard exchange with a refurbished (or even new) LDU at Tesla does not seem interesting to me. Indeed, the LDU will be guaranteed for 4 years. It will therefore be impossible for me to make any changes to it. And, if it starts filling up with coolant after 3.5 years… game over.

I installed a temperature regulator on the rotor to maintain it at 45°c permanently. Is this temperature sufficient?

I think to maintain the ventilation and the heating of the rotor (without opening it) for 2 weeks. (I will have tried)

I want to charge the HV lithium battery. Can I reconnect all the control connectors (disconnected by Tesla garage under the rear seats) and not reconnect the power to the LDU (B+ and B-) in order to recharge the HV battery?

When I encountered HV isolation error (Coolant heater plug corrosion) HV battery contactor doesn't close, charge port door doesn't allow plug in. So no way to charge battery. But HV battery also doesn't drain much if first responder loop disconnected. So probably no need to worry.
 
  • Informative
Reactions: Gtech
Upvote 0
Here is the result of the day with the megohmmeter: 2.2 Mohm. (Picture attached)

the value is increased

It is the same result between all the points (B+, B-, BUSBAR 1, 2 and 3) and the mass.

In view of this value, do I continue drying or do I try to reconnect it?
 

Attachments

  • 16733850-AFBB-4E5C-B561-D2AD7D2D0189.jpeg
    16733850-AFBB-4E5C-B561-D2AD7D2D0189.jpeg
    449 KB · Views: 136
Upvote 0
Here is Tesla's isolation fault diagnostic procedure. Unfortunately I don't know what should be the spec. Above video showing ISO cleaned and dried stator seems to be showing 4M ohms.

June 2013 Report.pdf (nhtsa.gov)
Tesla isolation fault diagnostic procedure say :
(page 19/23, point 15, third indent)

« If either measurement is below 3.5 MOhm, replace the drive unit »

I will dry the rotor until I get a value above 3.5Mohm.

Little question to those who have been able to test, what was the value of your rotor?
 
Upvote 0
EV HV is DEADLY. Understand the risks ( link )

Since your LDU is in the car. Need to be extra cautious as HV cable is always connected. Please understand the risks and necessary precautions.

Tesla isolation fault diagnostic procedure say :
(page 19/23, point 15, third indent)

« If either measurement is below 3.5 MOhm, replace the drive unit »

I will dry the rotor until I get a value above 3.5Mohm.

Thanks for the reference. Explains measurement from B+ B- on the drive unit. The videos I shared measures from the stator bus bar. Maybe its the same. I don't know.

====
Tesla LDU Rebuild ( link )
Model S DIY ( link )
2013 MS85 75k miles (1/25/23) Original RevB battery. 3rd (RevQ) LDU with rebuild
 
Last edited:
Upvote 0