Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

[spiff]

Just to document what I did (in case anyone has the same question): There is an excess of grease in the female spline of the primary gear that fits on the end of the rotor shaft that gets pooled up on the surfaces where it’s not used. I just rolled the shaft around these surfaces and used that. The volume of grease there was much greater than what was sitting on the rotor splines so I don’t think it’s a huge issue to have that little bit less.

Rotor half is assembled! Now moving on to cleaning up the inverter mating surfaces and getting the wiring passage ready to do the final merge. Wish me luck!

[howardc64]

DELETED

[asavage]

ATF still full so likely no ATF leak into motor

A tablespoon of oil will make an enormous mess on the floor or distributed by moving air over external components, but won't show up on a level check. That said, the color of your streak/stain alone makes one think of rust, not ATF.

[howardc64]

DELETED

[jrbe]

I'm late to this, some ideas, best practices and info from going through this thread.

Sealing surfaces have a specified finish to work properly. Rougher than specified means seal wear from abrasion. Smoother than specified means more surface area touching seal lip to mating surface than required to seal = extra friction which = extra seal wear.
PTFE seals typically have 1 lip. I'd be weary of them if they had 2 unless they were sealing with fluid on each side.

When you are pressing bearings you must ensure you are pressing on the race that is being installed. Outer race must be used when pushing into the housing, inner race must be used when pressing intro the shaft. If you don't, you load / overload the bearing and there's a high likelihood of damaging the bearings. This gets tricky when assembling case halves with bearings.
Tesla may have decided to leave some bearing bores loose for easy - just press it together without worrying about damaging the bearings for production. It makes sense to have retaining compound in those loose bores if they went this way. Deciding to leave the retaining compound out isn't a great idea. The aluminum case expands much more than the steel in the bearing as they heat up. It can get loose / spin if you skip it. In fact, heating the case is the easy way to get stuck bearings out typically.

As for the coolant leak rusting different areas, would it make sense to cut a weep slot in the position sensor cover? Coolant leaking on the ground at least gives you the hint something is wrong. It should also not get nearly as bad as some of the pics show with a weep hole if it leaks. I'm not sure a weep hole could work in some of these areas tough, I haven't been in one of these yet, the pictures don't all fall into where they are in the assembly for me yet.

[howardc64]

DELETED

[jrbe]

I think SKF's FKM seal excluder lip to keep dust out is just a tiny bit larger than the shaft diameter which make sense. The triple lip PTFE seal's dust lip hugs to the seal. And yes, 1 or multiple lip is debatable and latest research in bi-directional pumping aid probably a bit of a challenge as well (pumping structure probably easily gets clogged as coolant evaporates around the lip due to heat and leaves silicate behind) I think bottom line is, its impossible to seal without leak in this application. Latest seal designs use face seal technology along with leak routing/exiting channels in the seal cartridge for the minute leaks that will occur.

There were some strange choices made in the design. Watching the breakdown videos had me scratching my head in a bunch of spots. But hindsight is 20/20. Getting any coolant that leaked into sensitive areas out without pooling is imperative, I missed the weep hole you mentioned pages back, that's great.

Bearing Retaining Compound

Thanks for the thermal expansion index of case/bore and bearing. In my experience on this LDU, here is what I encountered on the 3 counter bore bearings

- Primary shaft's 6207 bearing dropped into the retaining compound coated bore slowly seated with gravity alone. All surfaces clean (wiped off frozen bearing surface ice coating from condensation) other than retaining compound on the bore where it mates to the bearing's outer face.

This one probably gets installed on the shaft assembly and slips into a heated casing (250°F max likely for the aluminum, unsure what the magnets can handle.) This is from my still somewhat poor understanding of the assembly (even after watching videos.)

- Intermediate shaft's FAG bearing was much tighter if all surfaces are dry and will jam if just slightly angled into the bore. Lubed insertion would be better but no way to do retaining compound this way. So I chose not to apply retaining compound here (original bearing had a 1cm red dab on the side wall of outer race (I didn't follow this as was suggested it makes no difference and original bearing was pulled out by fingers while lubed with ATF so this dab achieving nothing) Perhaps thermal trick would have worked (heat bore and/or freeze bearing). Unfortunately didn't try.

I'm not 100% on where the oil seal inside the motor seals on its sealing surface. Ideally the assembly starts there to watch that seal can be installed properly and without worry of damaging the oil seal. That looks like the center point / starting point of the assembly from what I'm seeing.

- Big bearing on the differential shaft has no evidence of retaining compound on original assembly.

Likely slid into a heated case while its already on a gear / countershaft assembly. The electronics would handle 200°F which should also make the bearings slide in easier.

Anyhow, definitely experienced knowledge on installing tighter fit bearings with retaining compound AND inverter removal/install along with custom jigs to position the 3 feet tall 300+lb drive unit for assembly would help.

Jigs, stepped press plates, and bearing press tools to match shaft heights would be really helpful.

Weep hole

We cut a weep hole in the lowest spot in the seal chamber. Position sensor hole is higher than this but any accumulated liquid here would just run down to the lowest spot. Here is a pic of the weep hole and speed sensor mounting location for reference.

https://teslamotorsclub.com/tmc/attachm ... eg.874371/

That looks great.
An adapter could be spun op to be able to use a different seal. Like a collar the seal presses into (with o-ring) then install that assembly into the coolant manifold.

Is it possible the coolant manifold is available as a spare part with a seal installed?

[howardc64]

DELETED

[muehlpower]

Deleted

[howardc64]

DELETED

[WimV]

As already announced, I had a professional sealing company make sealing rings for the LDU. These have now arrived. Haven't tested them yet but look good. They should work with or without Speedy sleeve.

Very interesting!

Anyhow, definitely experienced knowledge on installing tighter fit bearings with retaining compound AND inverter removal/install along with custom jigs to position the 3 feet tall 300+lb drive unit for assembly would help.

I made an adapter plate to mount the drive unit on an engine stand. Makes it at lot easier to work on.

[2oldteslas]

Hello, I've been watching this thread for a couple weeks since my 2012 S with 180,000 miles threw some error codes and shut down. There was coolant leaking into the motor causing an isolation fault. This P85 large drive unit was last replaced in about 2017 and has about 80,000 miles on it. Using the information from this site I pulled the motor and split the case. All your posts were very helpful! The rotor spins freely and has no lateral play but will not budge when I try to remove it. I've seen what happens to bearing mating surfaces when they are pounded on with sledgehammers in other applications so I decided not to go that route. I built a horizontal press to push the rotor out in a controlled manner. It uses a 12 ton bottle jack for the force. I fabricated a pusher nut that fits over the roller bearing contact surface of the rotor shaft and bears on the flange behind it so I won't damage the shaft. The other end of the press has a hole large enough for the bearing on that end of the rotor shaft to fit through so there will be no binding. Here is my problem: With at least 10 tons of force from the bottle jack there is no movement in the rotor. I tried tapping parts of the jack, press frame and push nut with a 3 lb hammer while it was under pressure to no avail. It has been under full pressure for hours with no changes. I fully expected the ceramic bearing to break apart. I have a new set coming from Ukraine so I'm not worried about destroying the bearing. I have more coolant seals coming from China to the Chicago area as well. Does anyone here have suggestions or comments on what I may have missed? My other 2013 S 85 has a bad bearing in the drive unit so that one will be next. I wonder what Electrified Garage and Tesla do to remove the rotor shaft. Gruber said in mid July 2023 that they won't take the job because they can't get parts. I'd rather do it myself for the $6000 US id have to pay someone else. Thanks for your help. Attached are some pictures of the press I made specifically for this job.

[asavage]

Without a top view, and not having a lot of in-person experience with the LDU at this stage (other than visiting Howard a couple of times during his journey), I can't tell if there's clearance between the end of the rotor and your end frame in this picture.

Regardless, you are not pushing on the rotor, but on the pinion gear. I believe you're pushing on the back of this:

Pinion shaft/gear, with KM 7 Axial Nut retaining the brg.

You are pushing against the pinion gear/shaft, which abuts a retaining ring in the case:
[later: forgive me: this is a pic of a RAV4 EV pinion, which is why it has that "extra" externally-toothed doo-dad on the end, for the Toyota-mandated park pawl that Tesla made just for this vehicle and the MB B250e. It's the picture I have handy, though . . . ]

Pinion shaft, out

Pinion shaft, installed (w/Aegis ring)

Lemme go find a vid (no, not Idiots at Work, something less combative . . . I'm well into "wine time" tonight) . . .

. . . found it:



Worth it to see the vertical alternative, but now that I've reviewed that vid again, it doesn't mention the important part . . . I think.

I'm guessing that you want to use a drift and press through the pinon to the end of the rotor.

Howard, do I have this correct?

[later still]

You'll want to press on the end of the rotor, which looks like this when it's got rusty bits around the rotor's bearing, to the inside of the pinion shaft/gear:

[2oldteslas]

After pouring WD40 in it last night and sleeping with this on my mind I came out this morning to find exactly what you described. I'm pushing on the pinion shaft. I will now attempt to remove the collar inside the pinion shaft so I can get to the end of the rotor shaft. I'll probably use a slide hammer with a course thread on the end to grasp the collar. The flared end of the collar is inset 2.320" (59mm) from the end of the pinion shaft. A brass mandrel should be soft enough to not damage the splines as I push the rotor out. I should make a video of this process but I don't have the necessary equipment. As a side note, this 2012 Model S has 182,000 miles on it and the motor has been out at least twice for bearing noise repairs. The tag does not specify REMAN so it might have been a new revision R motor when it was installed by Tesla under warranty. I'll check the exact mileage and dates shortly. Updates to come...

[asavage]

They're pressing on something. IDK what, but it stands to reason that pushing on the pinion shaft/gear won't work.

[muehlpower]

I just got a call from GFD. Someone asked about my sealing rings. This is bad because I did not have permission to publish the drawing. This means I will keep my ideas and results of my tests to myself in the future.
Thanks very much!

[jrbe]

I just got a call from GFD. Someone asked about my sealing rings. This is bad because I did not have permission to publish the drawing. This means I will keep my ideas and results of my tests to myself in the future.
Thanks very much!

Your previously posted solution and references made it sound like there's a solution available to replace the seal. There was no mention of how to buy one, anything about confidentiality from your post, and nothing noted on the drawing. Usually there is a confidential stamp across the drawing if it is. The breadcrumbs were there to source them with the info you provided but not to know it was confidential. I don't think anyone intentionally got you in trouble, they were likely trying to solve their own issue with the solution and info you provided. If I was in either of these situations I'd be currently apologizing for getting you in trouble.

If I needed a seal I would have reached out to you first. If that failed or you didn't reply after a week or 2 my next step would be to contact GFD and try to buy from them. If I was an engineer at Tesla and was tasked with solving the sealing issue I would have gone straight to GFD to get one and see how they stack up.

I appreciate you posting / sharing your work, thanks for sharing it all. You are extremely talented. I think most /all on this site appreciate your contributions as well. Please consider sharing exactly as you have in the future, just not the confidential bits.

[muehlpower]

There is a reference to the Copy Right on the drawing, but I haven't seen it either. I have ten rings that would sell at cost price of 100€ + shipping. Actually, I wanted to wait until I have made meaningful tests. So who wants to buy at his own risk, please contact me. Shipping Worldwide!

[2oldteslas]

I got the gearbox and inverter put back together for my 2012 P85. All the bearings spun smoothly and had no play so I didn't replace any of them. My rotor bearings were the same. Johan mentioned in one of his videos that he damaged the dust seal on a motor bearing while removing it to repack the bearing. On mine the inner bearing stays in the case when I remove the rotor and it seats very tight. I have to press the bearing in using the bolts on the end cover of the case. I'll get a blind puller for the inner when I remove it. I have access to the outer bearing so I pried the dust cover off with some tools I got from an old machinist. These worked perfectly with no damage. I highly recommend getting something similar to this for repacking the bearings. I didn't have any of the right kind of grease so I just redistributed the grease that was already inside the bearing. It should last until I take it apart for the leaking coolant seal...again. I ordered 3 different kinds of seals: single lip, double lip and triple lip. I ordered 3 of each kind to set up a longevity test between them or if someone needs a seal in less than 3 weeks for a repair.

[2oldteslas]

While cleaning the stator with an abrasive pad I noticed small metal shavings (hairs) sticking straight up from where the potting material was missing. (The grey strips in the picture.)
I attempted to clean them out and apply some quick set epoxy but the metal filings stuck up through the epoxy as it set. The filings had to come from somewhere and I don't think I created them with the green kitchen abrasive pad I used to clean the rust. An electric motor rebuilder told me to check for continuity between the individual strips and between the strips and the case as this was a sign of a short in the windings. They all have continuity. I found that this is because the strips are spacers connected to the case between the windings. Make sure you talk to people who know about this type of motor. Three phase AC brushless. I will attempt again to remove the metal filings and repot the windings. Has anyone else run across this? The motor is not a reman but has about 80,000 miles on it.
While I was checking continuity I decided to check everything else. I separated the motor buss bars from the inverter side by using plastic shims to isolate the connections from each other (picture). All 3 phases on the motor side have continuity with each other and none have continuity with the case. On the inverter side none of the 3 phases have continuity with each other or the main + and - power cables or case. This all makes sense and I believe it is working as designed. I'm no expert on electric motors though. Please correct me if I am wrong. Also be very careful to not bump the electrical pin plug on the inverter when you have the case removed. It is VERY fragile and will tear the ground strap that is potted into the plug itself. I had some brass sheet metal so I soldered that to the original ground strap to make the connection. The ground strap is a location mount for the plug so it has to be shaped and mounted properly (picture). I'm learning a lot working on this car and look forward to facing more challenges as I dig deeper. This site is a good resource for people to exchange ideas.

[howardc64]

DELETED

[howardc64]

DELETED

[asavage]

Not sure where you got this pic but I think its for pressing in the 2x ceramic bearings.

Listen for about 20 seconds, after Alex describes rusted-in-place rotors, he says "push the rotor out" whilst pointing to the press.

[2oldteslas]

howardc64

I had the gear case split apart. Inside the pinion shaft there is a steel collar that looks to be a grease stop for the rotor splines. It has a small hole in the middle so I screwed a self tapping screw into it connected to a slide hammer I made to pull the cap out. Then the rotor splines are visible through the pinion shaft. I used a brass slug and a steel bar to push the rotor out with the press I made. The inner ceramic bearing stayed in place when I pressed the rotor out. I took the grease seal off with the tools I got from the machinist and redistributed the grease without removing the bearing. The outer one was much easier to access since it came out with the rotor. The entire assembly is back together and waiting for the coolant seals I ordered. I'll probably use the 2 lip seal first, I think 3 lip is overkill and I've read that it doesn't fit properly.

[2oldteslas]

Now that I'm looking inside the stator with a magnifying glass these metal shavings "hairs" sticking up look more like fibers from the insulating casing around the windings. They don't appear to be metal at all. It sucks getting old, haha.