Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register
This site may earn commission on affiliate links.
An update for those seeking to rebuild their LDU

@3k+ mile post rebuild. Found completely rusted and pitted shaft below the triple lip seal (first warning was rusty streak on speed sensor).
Ouch... sorry buddy 😑
Sounds like best thing is to stay away from LDUs...
Second best is to let Tesla replace it n at least have 4yr/50k mi warranty...

I'm guessing this is why Tesla using single lip, less chance of trapping moisture n causing rust
Wonder if water-less coolant can be used?...

Good info for everyone, thanks for following up!
 
FYI, there's a member on Facebook BMS_u029 group with 2018 p100d that had 1st rear LDU replacement at 130k mi
That's a first time i've heard of one lasting that long...
So, i'm inclined to believe that LDUs last longer when driven a lot...

1691082305637.png
 
Ouch... sorry buddy 😑
Sounds like best thing is to stay away from LDUs...
Second best is to let Tesla replace it n at least have 4yr/50k mi warranty...

I'm guessing this is why Tesla using single lip, less chance of trapping moisture n causing rust
Wonder if water-less coolant can be used?...

Good info for everyone, thanks for following up!

yeah, concerns for multi-lip surfaced on this one as is the likely desire for protective plating+coating. Protective layer would need to be on thick enough to then polish down to 20-30Ra desirable for PTFE seal. Quite meticulous process... Probably quite a bit of room for error and sloppiness here.

A nearby DIYer (@asavage) also thought of waterless coolant. Got some along with Tesla coolant pumps but too many other projects and haven't started bench experiment.

FYI, there's a member on Facebook BMS_u029 group with 2018 p100d that had 1st rear LDU replacement at 130k mi
That's a first time i've heard of one lasting that long...
So, i'm inclined to believe that LDUs last longer when driven a lot...

View attachment 961967

Yes a couple of notes here.

There seems to be some that swear LDU seal leak is not a problem even at high miles. I think all are original LDUs including early MS. But not too many early cars have original LDUs just a few year after purchase as Tesla kept on replacing them for 2 main problems (accelerator on/off clunk, milling noise from micro arc welding bearing race in early non ceramic bearing LDUs) so most early cars are on 3+ LDUs (My 2013 is #3 at 40k miles by 2017) But some people seems to have stayed on early LDUs (far away from SC?) What I have since then learned is the Tesla LDU remans process have quite a bit of variance and QC problems. So it is likely the original factory robot assembled LDU with whatever early seal (triple lipped from what LDU rebuilders say) and probably new original protective plating/coating on shaft might have lasted. Reman process might have had QC problem on shaft rust protection which probably doesn't show up with single lipped seals. In hindsight, probably better to say with original LDU's problems then getting the remans. Reman battery packs seems to also be suffering QC problems from reading posts here.

And LDU leak also seem to have a time component as well beside miles so high mileage guys seems to get most benefit. The leak could be static (while parked) or maybe even just in reverse direction. Or only when shaft move slightly axially (when cornering, centrifugal force on that 70lb rotor will move the shaft against a pair of bevel washer springs so maybe 0.5mm max axial travel) Don't really know. Lots of people provide theories. But reality is without a high speed camera inside the seal chamber watching it 24/7 parked and during operation, no one really knows. What is more surprising is researchers that study PTFE lipped seals (Univ of Stuttgart has 20+ years graduate research with seal bench test jigs etc) shows 1 drop/hour on any PTFE lipped seal (they engineer seals with new pumping aid ideas such as micro patterns and varied material density) with oil. This is at lower surface speed than LDU application. But their test jig does go forward and reverse constantly so perhaps more stressful. U Stuttgart also has an X-Ray study of cylindrical Lithium Ion cells non uniform internal swelling during charge/discharge cycles. Quite interesting and lots of modern automotive application graduate lab studies. Graduates go onto Bosch and other EU auto sector manufacturers.

Anyhow, all conjecture of course. My shaft seal surface prep on rebuild was 600->1500 grit sandpaper (acetone cleaned after) to target 30Ra ish finish. Barely took off any material (only hardened silicate on sandpaper on first 1-2 rotations) and didn't dare to go too far and left a micro groove from prior seal there. When finished, there is no visible shaft material on my tiny 7-8mm wide sandpaper. Reman LDU was originally manufactured in 2013 and careful inspection of the gears shows maybe high miles + driven hard (I Primary shaft takes the most abuse, highest speed, smallest size, I put 32k miles on the reman in 5 years driving like grandma haha)
 
Last edited:
long time lurker, first time poster so apologies if this is not the appropriate place for this question.

is it possible to swap a failed LDU with a known good one? I have access to lift and tools have spent many years building engines/transmissions. From what i've gathered here and the factory service manual it should be possible with the tesla toolbox software to swap the LDU to a known good one. Just wondering if anyone has done this or if there is anything to be aware of.
 
long time lurker, first time poster so apologies if this is not the appropriate place for this question.

is it possible to swap a failed LDU with a known good one? I have access to lift and tools have spent many years building engines/transmissions. From what i've gathered here and the factory service manual it should be possible with the tesla toolbox software to swap the LDU to a known good one. Just wondering if anyone has done this or if there is anything to be aware of.

I haven't done this but had similar ideas in the past. I think swapping inverter (or maybe even just the control board (1 board on bottom of inverter, lots of connections, its the brain) in the inverter) is enough. Everything else is dumb (stator windings, 3 temp sensors) if swap whole inverter. If just control board, then the 3 IGBT boards of the old LDU would remain. Perhaps check with someone who has done this to confirm.

Removing / installing inverter does have its challenges. Easy to crimp the coolant o-rings. Here are notes on inverter removal + install

 
  • Like
Reactions: Olle
LDU DIY Rebuild 2.0

Completed rebuild 2.0 after finding rust under seal after 4k miles on DIY rebuild 1.0. Here is a summary

Rust Under Triple Lip Seal


My best guess is originally manufactured LDU has anti-corrosion coating on the rotor shaft where coolant seal contacts. This coating was likely compromised during Tesla reman. There is a shallow groove channel on the shaft outside of where reman LDU seal makes contact. Presumably this groove was formed in the original LDU and Tesla reman doesn’t address reconditioning the rotor shaft back to original new spec.

I also lightly sanded the shaft surface with 600->1500 grit seeking to achieve a 0,20Ra finish for the seals. There were no visible materials taken off on the sandpaper nor did I dare to sand down enough to get rid of the above mentioned groove.

Anyway, unless the rotor shaft is pristine and without unknown reman handling, it is probably advisable to avoid triple lip seal.

Flame Spray Repair Rusted Shaft

Flame Spray basically shoot metal droplets onto the shaft to build up material. Then mill down to spec. This process leaves many micro porous holes which then need to be sealed (sealer + 200C cure) The flame spray shop didn’t mask off the shaft chamfer. This chamfer is critical to be smooth as it is a seal installation tool. In particular the PTFE seal lips which exert high force while cleaning this chamfer.

Therefore, flame spray repair alone appears insufficient. There is a Croatia EV rebuild franchise company demonstrating chrome plating.

One major concern is any flame spray or chrome plating shop aren’t experienced in LDU rotor’s high precision and detailed requirements. It is difficult to ask them to study in detail and do detailed work that isn’t part of their major business.


Read following posts on the last link

Solution for damaged rotor shafts

With a damaged rotor shaft (and failed flame spray repair) my rotor shaft isn’t much different than a rusty shaft from a salvaged LDU. Flame spray repair shaft looks nice to the naked eye but finger nail can feel the porous holes and microscope reveal many porous holes. The best solution so far is to use SKF speedi-sleeve along with SKF FKM seal. One person has 15k miles on this solution and counting.

SKF speedi-sleeve will not cover the entire rotor shaft region so triple lip PTFE seals use is not possible. Sleeve also has a somewhat sharp edge which makes installing a single lip PTFE seal risky. One idea is to smooth that edge by painting with nail polish just to enable PTFE seal installation without damage. For elastomer seals such as FKM, a light coat of silicone grease is fine. Probably even dry install is fine.

I chose speedi-sleeve with FKM seal.


Some Additional Things Learned on rebuild 2.0

Reman LDU Inverter Stripped Bolt Holes


Tesla Reman stripped 3 of 6 outer inverter mounting bolts. The design has very small margin and T40 bolt tool will be tilted due to inverter side walls. Very easy to strip.


Also read post #19 and #20.

Jacking Up the Rear to put on Stand

I now do a 3 step process (no lift)
  1. Drive up a shallow ramp on top of 2x8 wood piece to gain clearance to remove the aero shield.
  2. Remove aero shield under the LDU
  3. Jack up by beefy parts of the rear subframe. The frees the rear lift point for placing stand
IMG_0630.jpg

Primary Shaft Spline Grease

Confirmed with professional LDU rebuilders just use any high temp good quality grease. Probably something with similar viscosity. Tesla as usual has a PN for the grease but not for sale.

Eliminate Rotor Cooling Completely?

There seems to be some potential for eliminating rotor cooling with coolant (or anything at all) Tesla parts catalog has a “coolant DELETE manifold” listed for $0.01. Tried to order but was unavailable.


Also recently heard from a Polish owner claiming a local EV repair shop has done multiple LDUs with modified manifold to skip coolant cooling of rotor.

Will post more as I find out more.
 
Last edited:
LDU DIY Rebuild 2.0

Completed rebuild 2.0 after finding rust under seal after 4k miles on DIY rebuild 1.0. Here is a summary

Rust Under Triple Lip Seal


My best guess is originally manufactured LDU has anti-corrosion coating on the rotor shaft where coolant seal contacts. This coating was likely compromised during Tesla reman. There is a shallow groove channel on the shaft outside of where reman LDU seal makes contact. Presumably this groove was formed in the original LDU and Tesla reman doesn’t address reconditioning the rotor shaft back to original new spec.

I also lightly sanded the shaft surface with 600->1500 grit seeking to achieve a 0,20Ra finish for the seals. There were no visible materials taken off on the sandpaper nor did I dare to sand down enough to get rid of the above mentioned groove.

Anyway, unless the rotor shaft is pristine and without unknown reman handling, it is probably advisable to avoid triple lip seal.

Flame Spray Repair Rusted Shaft

Flame Spray basically shoot metal droplets onto the shaft to build up material. Then mill down to spec. This process leaves many micro porous holes which then need to be sealed (sealer + 200C cure) The flame spray shop didn’t mask off the shaft chamfer. This chamfer is critical to be smooth as it is a seal installation tool. In particular the PTFE seal lips which exert high force while cleaning this chamfer.

Therefore, flame spray repair alone appears insufficient. There is a Croatia EV rebuild franchise company demonstrating chrome plating.

One major concern is any flame spray or chrome plating shop aren’t experienced in LDU rotor’s high precision and detailed requirements. It is difficult to ask them to study in detail and do detailed work that isn’t part of their major business.


Read following posts on the last link

Solution for damaged rotor shafts

With a damaged rotor shaft (and failed flame spray repair) my rotor shaft isn’t much different than a rusty shaft from a salvaged LDU. Flame spray repair shaft looks nice to the naked eye but finger nail can feel the porous holes and microscope reveal many porous holes. The best solution so far is to use SKF speedi-sleeve along with SKF FKM seal. One person has 15k miles on this solution and counting.

SKF speedi-sleeve will not cover the entire rotor shaft region so triple lip PTFE seals use is not possible. Sleeve also has a somewhat sharp edge which makes installing a single lip PTFE seal risky. One idea is to smooth that edge by painting with nail polish just to enable PTFE seal installation without damage. For elastomer seals such as FKM, a light coat of silicone grease is fine. Probably even dry install is fine.

I chose speedi-sleeve with FKM seal.


Some Additional Things Learned on rebuild 2.0

Reman LDU Inverter Stripped Bolt Holes


Tesla Reman stripped 3 of 6 outer inverter mounting bolts. The design has very small margin and T40 bolt tool will be tilted due to inverter side walls. Very easy to strip.


Also read post #19 and #20.

Jacking Up the Rear to put on Stand

I now do a 3 step process (no lift)
  1. Drive up a shallow ramp on top of 2x8 wood piece to gain clearance to remove the aero shield.
  2. Remove aero shield under the LDU
  3. Jack up by beefy parts of the rear subframe. The frees the rear lift point for placing stand
View attachment 974271

Primary Shaft Spline Grease

Confirmed with professional LDU rebuilders just use any high temp good quality grease. Probably something with similar viscosity. Tesla as usual has a PN for the grease but not for sale.

Eliminate Rotor Cooling Completely?

There seems to be some potential for eliminating rotor cooling with coolant (or anything at all) Tesla parts catalog has a “coolant DELETE manifold” listed for $0.01. Tried to order but was unavailable.


Also recently heard from a Polish owner claiming a local EV repair shop has done multiple LDUs with modified manifold to skip coolant cooling of rotor.

Will post more as I find out more.
I would definitely be interested in that Polish owner’s claim. I’m stationed in Germany with my 2015 P85D and have been considering my options for a LDU rebuild. Right now it’s Germany, Norway or Croatia.

As far as removing the coolant from the LDU, on the face it just seems like a bad idea. I would be curious to see if air cooling is sufficient or if they modify in such a way to introduce more air to the LDU.
 
I would definitely be interested in that Polish owner’s claim. I’m stationed in Germany with my 2015 P85D and have been considering my options for a LDU rebuild. Right now it’s Germany, Norway or Croatia.

EV Clinic in Croatia has the best LDU rebuild info I've seen so far. They chrome plate the rotor shaft. Can even build a new and better stator. They wrote a long open letter to Tesla on the poor quality reman LDU units.


As far as removing the coolant from the LDU, on the face it just seems like a bad idea. I would be curious to see if air cooling is sufficient or if they modify in such a way to introduce more air to the LDU.

Yes, on the face seems counter intuitive. However, there are probably more variables. For example, I'm sure can roll at 10mph safely for longevity so power demand probably has a lot to do with it as well. Here is a quick write up. Trying to find a way to get quantitative so have back of envelop calculation to see how realistic.

 
There seems to be some potential for eliminating rotor cooling with coolant (or anything at all) Tesla parts catalog has a “coolant DELETE manifold” listed for $0.01. Tried to order but was unavailable.
Here you go. Turns out it is actually real. No details beyond this pic. Pictured is a reman Base LDU part number 1002633-00-U.
 

Attachments

  • manifold delete.png
    manifold delete.png
    3.4 MB · Views: 80
Here you go. Turns out it is actually real. No details beyond this pic. Pictured is a reman Base LDU part number 1002633-00-U.

Interesting. The printed PN on the coolant manifold in the pic (1492787-00-A) is what I found a few months back on Tesla parts catalog ( link ). Current catalog shows the part isn't listed anymore and latest LDU PN is now revision -U per your post.

Without seeing inside of the manifold, can't confirm if coolant no longer travels through the rotor center. The external shape suggest the coolant inlet passage is directly routed to the top exit to the external tube without routing through the center of the rotor. The round shape in the middle of the manifold is probably just to accommodate existing rotor's extruded coolant shaft. Coolant shaft would be useless now but can avoid needing to make a new 70 pound squirrel cage aluminum copper rotor. Made diagrams to show existing coolant circuit here ( link ) I noted in the circuit diagram the external tube from the coolant manifold eventually need to go through a 4.5mm hole so likely not much coolant volume is passing through the rotor.

Here is an annotated pic of the existing coolant manifold internal coolant passage with the seal removed.

IMG_2994.jpeg

Existing manifold can probably be modified : Remove tube on the manifold and close off the hole (A). Close off the hole that feeds the top manifold exit to external tube (B, hole not visible in the picture). Tap the internal coolant passage and route a circuit on the outside from manifold inlet to top exit to external tube. But there isn't much room externally as the manifold has a small gap to the rear subframe. Quite a shame Tesla won't publish a TSB or make the new manifold available. It will take awhile for 3rd party/DIY community to see a Rev U LDU manifold.

IMG_1099.jpeg

I wonder if also require a firmware update as well. LDU has no thermal measurement of rotor temperature. However, there are research papers outlining how to accurate estimate rotor temperature of typical electric motor rotors without direct measurement within the motor controller software.
 
Last edited:
Maybe block the passage from the lower inlet hose to the rotor and simply add a T to that hose to run a small line to the top hard tube. A brass heater T and some properly sized heater hose shouldn’t be too difficult to source.

I had my Tesla contact search that new manifold part number in the system and it came back with quantity zero at all locations worldwide. Maybe the printed number is no longer current.
 
Maybe block the passage from the lower inlet hose to the rotor and simply add a T to that hose to run a small line to the top hard tube. A brass heater T and some properly sized heater hose shouldn’t be too difficult to source.

I had my Tesla contact search that new manifold part number in the system and it came back with quantity zero at all locations worldwide. Maybe the printed number is no longer current.

yeah, was thinking similar after my post...

I do remove the plastic manifold inlet when pulling and installing the LDU into the subframe. There is so little room and plastic parts on both sides (coolant inlet on left side, HV connection plastic housing on right side) I worry swinging nearly 400 pounds to remove/install could break them.

Another way is
  • Tap the inlet coolant hose with a T, cut top aluminum tube between manifold exit and gearbox casing entry and feed coolant directly externally. Can be removed/installed with LDU in the subframe.
  • 2 channels would need to be sealed : Manifold's top coolant exit (easy with metal plate + sealant). Manifold internal passage feeding the rotor ( JB weld can make a good long term water tight seal?)
One note on the changed thermals. Electric motors require cooling for longevity. Would be good to have a induction motor engineer comment on what the impact might be. With a 4.5mm constriction on the rotor coolant circuit and ~3? psi of pressure. I'd imagine its possible to calculate thermal cooling loss and approximate rotor temp based on known power feed to the motor.

The coolant delete manifold also makes it entirely possible for SC to do field upgrades. Just remove LDU and change manifold. This would reduce the repair cost significantly without requiring customer trading in an expensive cost at $0 value. Speed sensor inspection is easy without removing LDU and inverter chamber inspections easy post LDU removal to determine coolant ingress at inverter. But with Tesla's general avoidance for SC in field repairs, not publishing repair procedures, not supplying parts... Basically force the aftermarket+DIYers to mods seeking more cost effective repairs.
 
Shared the coolant bypass manifold pic in openinverter. A good mod suggestion for modding existing coolant manifold is

- remove the rotor tube
- fit a custom cap into the coolant seal bore to seal off the chamber. Proper shape to clear the rotor shaft and leave enough gap for coolant to flow out of the removed rotor tube hole.


BTW, Tesla not enabling coolant delete manifold update at SC is like asking the customer to pay for an engine to update a badly designed water pump. Tesla has long long way to go to reach efficient repair cost+process.
 
We've got a new aftermarket manifold option we've been working on at QC Charge. It can be used either for a complete rotor cooling delete, or to split the rotor cooling to a separate circuit, which can be oil cooled. Hopefully we will have them available starting in January after we do some further testing.

We expect to sell the "manifold" for $599. Oil cooling would add additional cost for a pump, reservoir, and heat exchanger. We will sell these either as standalone parts, as well as with availability for installation at our shop.

Rotor Manifold Oil Cooled.png


Rotor cooling diagram.jpg
 
Last edited: