Taking foot off the break pedal disables the vehicle.

[tjlee]

It's not mandatory to replace the rotor bearings -- Johan didn't on his RAV4 EV -- but in my world it's Best Practice. I've got rotor bearings on the way from Acorn in the UK -- supposedly to arrive by tonight, but tracking showing still in Belgium -- best price for genuine SKF (6007-2RZTN9/HC5C3WT) that I've found at about $252 shipped for (2). I initially had trouble getting past the VATIN field on the Checkout page, but a couple of messages to them via their Contact Us page sorted that out. There are a lot of counterfeit "SKF" ceramic bearings being sold.

The axle seals are always in need of replacement, because they're always leaking Same problem on Model S, too. Tesla doesn't sell parts (a generality), so best I think they're best sourced via Toyota. It's been posted that they're Toyota 17292-0Z010 but I haven't verified that. I do need to get some coming. I'm surprised there isn't a generic seal available.

You're not supposed to re-use the axle nuts, 90080-17238, $20 for two from my local Toyota dealer (where I can be sure they're genuine). I'm picking those up Friday, they're in Will Call.

I'm doing a "coolant delete" manifold modification via a machined 6061 plug "hat", so I'm not buying a rotor seal; I have a sort of prototype seal from Germany, which I paid ~$105 for a few months ago that I'd make you a very good deal on, should you want it, but I have to say that you're going to have a real challenge getting a reliable liquid seal to that rotor, with its wear and damage.

Are you planning to try a Speedi-Sleeve? Or a custom machined spacer (has to very very thin, hard to do)? Or the flame-spray + machining that Howard tried unsuccessfully? Or re-plating + machining? All these options have been tried, and none have proven 100% successful. Since I doubt I'll ever overheat this rotor in RAV4 EV usage -- I just don't drive that hard, or tow heavy loads up mountains -- I'm going with the "don't need liquid cooling" route for mine, via the coolant manifold modification, though it's tempting to just cut off the last inch of the rotor to simplify the lathe work for the manifold plug, down to a simple 3/8" thick circle, ~55.05mm OD.

I'm not 100 percent sure the car will run after rotor seal replacement. Once the car runs ok with repair, I would like to do "coolant delete" option as well. I have question regarding what you mean by "I'm going with the "don't need liquid cooling" route for mine, via the coolant manifold modification, though it's tempting to just cut off the last inch of the rotor to simplify the lathe work for the manifold plug, down to a simple 3/8" thick circle, ~55.05mm OD"
So far I have order axle seals, nuts and rings from local toyota dealer. Is the chengming seal best option for rotor seal so far. I've read so many threads it's so very confusing. I feel my rotor shaft is pretty bad so I might use the lathe to smooth it out and polish and see if chengming seal would work. If not use speedi sleeve option. Is it my understand that I don't need to anything on rotor shaft if I go with speedi sleeve.

 

[asavage]

If using Speedi-Sleeve, no shaft prep (other than very light cleaning) is needed. Howard's failed within 1k miles or so; BoxsterEV says his is still sealed after many thousands of miles. The sleeve is not quite long enough for a 3-lip seal, and no alternative longer sleeve seems to be available (I looked hard).

Cutting the shaft down via lathe is problematic for three reasons: it's thin and you can't remove much material. An undersized shaft will not give the same seal lip pressure as original 30mm (and none of the teflon seals use garter springs behind the lip, so they're very sensitive to shaft diameter). And, it's surmised that the original seal finish was a plated surface of some kind (that's Howard's guess, and he may be correct).

You can buy for $$$ ($700?) a replacement coolant manifold from QC Charge. It's a bolt-on coolant bypass of the rotor, while maintaining coolant flow to the inverter. QC Charge doesn't answer their phone nor emails, in most cases (certainly, that's been my experience with them). Good luck. I know of nobody at this time who has actually had one installed and had any experience with it. It looks gorgeous.

You can DIY mod your manifold to do the same thing, using some aluminum tubing and some drilling and epoxy (I think I posted a link the the thread above; you can do your own research) and removing (cutting off) the steel tube on the manifold that inserts into the hollow rotor. It's messy but definitely do-able DIY, and has been done at least twice. Here's a pic of the coolant manifold (with a seal installed) with the steel tube that must be removed:




Howard is seriously having a SS "hat" plug made in China, to press-fit into the coolant manifold to perform a less-messy DIY mod of the coolant manifold. You still have to cut off the long steel tube, but don't have to drill the manifold for an aluminum bypass tube, nor use epoxy to seal it. Howard likes SS; I prefer 6061 for this particular design. His design is formed SS sheet and quite inexpensive.

I am going to farm out the lathing of a 6061 round bar chunk I bought (three, actually) . . .




. . . to make an aluminum "hat" plug instead of formed SS sheet like Howard is doing. User philip295 over at that link I posted, he's done this "in principle" on his home lathe but hasn't installed it. An early drawing I made a while back:



(concept drawing only: do NOT use those dimensions)

philip295's actual part:







I aim to replicate this part, but finalizing the dimensions. In particular, there is a finite amount of room at the "top" of this hat; it must clear the end of the rotor, but on the other side also allow coolant to flow from below to above. It's tight. Grooving the manifold where the steel tube was removed is a probable step.

And, the interference fit of the 55mm OD to the old seal's counterbore has to be finalized. I'm using a cut-and-try approach, but it's a reasonable guess that maybe .002" (~.05/.06mm) oversize on the "hat" will be adequate for a light press fit in the manifold, but we'll make a couple of passes and see how it "feels".

Having said all that: if I were to cut off the end of the rotor where the seal rides, that "hat" could be a flat disc. Hacksaw work. Less machining, but if someone wanted to go back to liquid cooling, it would be more expensive to find a good rotor than an unmodified coolant manifold.

HTH

 

[TinkerTan]

I have precisely this same issue on mine. I have rebuilt the bearings and seal once (late 2022) and it lasted 5k miles, although the shaft and everything else was in quite nice condition - I must've damaged the seal when I struggled installing it.
The car sat for 6 months before I learned of the coolant loop delete solution then I did that on my own. I spent a day cutting the coolant flange in half and isolating the two halves (Aluminum MIG), and plugging the rotor coolant return tube up top. (By the way, if your car is otherwise in good shape and you plan to keep it, I recommend just buying a qc charge flange. I'm personally trying to repair this as cheap as possible and get rid of it since I'm unemployed).
I removed the inverter housing and dried out one ounce or less of coolant, although it appears as if the batt - terminal had sat in coolant.
The problem persists (ONLY at a stop i.e. right after putting it in R or D or even at stop lights - it otherwise drives perfectly fine) and I may need to drop the ldu to dry everything out more thoroughly... I will measure the phase resistance beforehand, as I think a damaged winding should have obviously higher resistance or continuity to ground.
I don't have a cable/adapter or tesla/rav4 ev software at this point so I can only assume it likely is due to:
- some coolant internally still shorting Batt - to housing
- wet phase wires/buses shorting phase-phase or phase-chassis
- damaged inverter
- damaged stator
- damaged encoder/wires (perhaps less likely but I can try to tap A/B to verify both are OK)

 

[TinkerTan]

To add to the list of helpful videos, Alex @ QC Charge posted a video on removing the LDU from a RAV4 that is pretty comprehensive. I've bookmarked it because I need to pull my LDU out for a coolant leak next month:



IMO, it's much easier to R&R the LDU in the RAV4 EV than a Model S, because the Model S requires dropping the rear subframe first.

Mind, this is only getting the LDU out of the car for shipment. Teardown/inspection/cleaning is quite a bit more work, but it's bench work. The unit is around 300 lbs.; the rotor is ~90 lbs. alone. I worked with one on a 2013 Model S last year, we had a dry stator but wet inverter harness, causing codes.

I will be doing the "coolant delete manifold" DIY modification to eliminate coolant flowing through the rotor. There at least three ways to DIY this mod, plus QC Charge's off-the-shelf $700 bolt-on version. It requires some reading, but a good thread on this begins here:

https://teslamotorsclub.com/tmc/posts/7898613/

After modification, coolant continues to flow through the stator and inverter sections, but not the rotor. No teflon seal is needed, and rotor shaft condition is irrelevant. If you're replacing the seal, the rotor shaft end's condition can be "fine" or "must be machined/sleeved/plated to be serviceable", and that can easily run an additional $50-$600.

How did you determine the wet harnesses were the issue? Or was it not until you resolved it that you knew?
How did you dry them?
Mine are wet and I cleaned out all the LV connectors on the bottom of the inverter with electronics cleaner spray (hoping the hydrocarbon spray would fetch some glycol but that G48 is hard to dry or wipe or dissolve)

 

[asavage]

Your timeline is off: I haven't dropped my LDU yet -- that's next week -- and my car is sitting in my driveway for the past three weeks or so, since I had ONE contactor-opening event and the attendant "Check EV System" message, and the next day I pulled the speed sensor to find blue coolant on it. It still drives, but I'm not using it until I've dried it out, so my issue isn't resolved yet.

Removing the G48: I will wash everything with plain tap water. Garden hose stuff. Water will bind with and dilute the glycol, reducing it to nearly none left.

Then, to remove the water. Soaking parts in an alcohol solution binds water to the alcohol, which is then easier to move out of areas that are hard to reach otherwise, such as windings, crevices, blind holes and other places where compressed air won't reach well. Vacuum boiling works too, but few have access to the size vacuum chamber to readily accomplish that. 99% IPA is available at my local hardware store for ~$25/gal. I'm going to drizzle and spray the stuff over my inverter boards, internal harness, and stator, as a wash. Outdoors, of course: it's quite flammable, esp. at warmer temperatures.

The 71% IPA stuff is much cheaper, but also a bit less effective. Dilute the remaining water, then blow dry. Then apply low heat, if possible, before assembly.

I used to have a "garage oven" for this kind of work, but I've moved and left the garage oven behind, and now I'd have a hard time finding room for one in the new digs.

 

[alflash]

....
How did you dry them?
...

One way to dry RAV4EV motor assembly

 

[tjlee]

If using Speedi-Sleeve, no shaft prep (other than very light cleaning) is needed. Howard's failed within 1k miles or so; BoxsterEV says his is still sealed after many thousands of miles. The sleeve is not quite long enough for a 3-lip seal, and no alternative longer sleeve seems to be available (I looked hard).

Cutting the shaft down via lathe is problematic for three reasons: it's thin and you can't remove much material. An undersized shaft will not give the same seal lip pressure as original 30mm (and none of the teflon seals use garter springs behind the lip, so they're very sensitive to shaft diameter). And, it's surmised that the original seal finish was a plated surface of some kind (that's Howard's guess, and he may be correct).

You can buy for $$$ ($700?) a replacement coolant manifold from QC Charge. It's a bolt-on coolant bypass of the rotor, while maintaining coolant flow to the inverter. QC Charge doesn't answer their phone nor emails, in most cases (certainly, that's been my experience with them). Good luck. I know of nobody at this time who has actually had one installed and had any experience with it. It looks gorgeous.

You can DIY mod your manifold to do the same thing, using some aluminum tubing and some drilling and epoxy (I think I posted a link the the thread above; you can do your own research) and removing (cutting off) the steel tube on the manifold that inserts into the hollow rotor. It's messy but definitely do-able DIY, and has been done at least twice. Here's a pic of the coolant manifold (with a seal installed) with the steel tube that must be removed:

View attachment 206


Howard is seriously having a SS "hat" plug made in China, to press-fit into the coolant manifold to perform a less-messy DIY mod of the coolant manifold. You still have to cut off the long steel tube, but don't have to drill the manifold for an aluminum bypass tube, nor use epoxy to seal it. Howard likes SS; I prefer 6061 for this particular design. His design is formed SS sheet and quite inexpensive.

I am going to farm out the lathing of a 6061 round bar chunk I bought (three, actually) . . .


View attachment 207

. . . to make an aluminum "hat" plug instead of formed SS sheet like Howard is doing. User philip295 over at that link I posted, he's done this "in principle" on his home lathe but hasn't installed it. An early drawing I made a while back:

View attachment 208

(concept drawing only: do NOT use those dimensions)

philip295's actual part:

View attachment 209

View attachment 210

View attachment 211

I aim to replicate this part, but finalizing the dimensions. In particular, there is a finite amount of room at the "top" of this hat; it must clear the end of the rotor, but on the other side also allow coolant to flow from below to above. It's tight. Grooving the manifold where the steel tube was removed is a probable step.

And, the interference fit of the 55mm OD to the old seal's counterbore has to be finalized. I'm using a cut-and-try approach, but it's a reasonable guess that maybe .002" (~.05/.06mm) oversize on the "hat" will be adequate for a light press fit in the manifold, but we'll make a couple of passes and see how it "feels".

Having said all that: if I were to cut off the end of the rotor where the seal rides, that "hat" could be a flat disc. Hacksaw work. Less machining, but if someone wanted to go back to liquid cooling, it would be more expensive to find a good rotor than an unmodified coolant manifold.

HTH

I have question for you. I was originally thinking about doing coolant delete like you've decided but I didn't have any round aluminum piece laying around so I came up with this option and wanted to see what you think. As the picture below shows i've carved out mil or so deeper into the cover and some more channeling towards the outlet/inlet port on the side. My question to you is if I use a play putty to create a channel to the port and use jb weld on top to create the enclosure, would the jb weld seal the area so no coolant will enter the rotor. Similar to hat cap your going to lathe, I am also going to create the groove for the shaft. Appreciate your input.

 

[asavage]

I'm always suspicious of adhesive and non-mechanical interfaces, if a problem can be solved by mechanical means.

Put another way, what you describe, forming a channel using JB weld and play putty as a filler for a mold, definitely doesn't appeal to me, but I'm certain it could be done, it's just that I don't have enough experience with those materials and construction methods to feel confident, so I would not attempt it.

Put yet another way, the consequences of a failure are that you have a potentially large leak that would within a few minutes lead to the inverter, destroying it.

The rotor can live with being soaked for a while, so can the stator, but the inverter would be quickly destroyed if, say, it was inundated by a cup of coolant within a couple of minutes, which would certainly be possible if the JB weld failed its adhesion to the manifold.

 

[tjlee]

Here is an update for stater cover modified for coolant delete. Jb weld was used

 

[asavage]

Are you shortening the rotor? Or, will you bore that epoxy to form a cup?

I just got a call from my machinist, my plug is ready for pickup.



I gave him some 6061 2.25" round bar to work with, and a drawing with dimensions (pointing out which were critical and which were arbitrary) and now, two days later, it's done.

I hacksawed off the tube, then used an end mill to remove the stub. Another end mill to form a channel "north". I bumped the seal's counterbore (oops) but it went rather well, and none of it is high tolerance, as it doesn't need to be.

 

[tjlee]

Are you shortening the rotor? Or, will you bore that epoxy to form a cup?

I just got a call from my machinist, my plug is ready for pickup.

View attachment 226

I gave him some 6061 2.25" round bar to work with, and a drawing with dimensions (pointing out which were critical and which were arbitrary) and now, two days later, it's done.

I hacksawed off the tube, then used an end mill to remove the stub. Another end mill to form a channel "north". I bumped the seal's counterbore (oops) but it went rather well, and none of it is high tolerance, as it doesn't need to be.

View attachment 227

View attachment 228

View attachment 229

I've pretty much done what you have except yours look way cleaner. I won't be cutter the end of the rotor nor boring out jb weld. I haven't double check but I think it should work the way it is.
Do you have some insight on troubleshooting the stator. After washing with water and using alcohol to complete dry probably not dry completely yet but i'm getting less resistance than before. I'm getting continuity on three prongs as well. I wonder whether my stator is bad. Each prong currently gives only 3.4 mohms. Before the wash It showed 9-10 mohms.

 

[asavage]

Do recall what the spec is? I would think >3Mohm ought to pass, but I can't recall the spec.

If you've done alcohol and water wash, I can only think of heat as an aid.

Anything that can be reached with air will benefit from air movement, more than heat: think fans, not heaters. But if the moisture is embedded where air cannot reach, then drying will be proportionally slower, and good dilution (clean water + alcohol again) helps a lot. The glycol basically never dries and wants to hold on to water, so diluting it is the only way I can think of that has a chance.

Heat sounds nice, but heat without air doesn't do a whole lot (until you reach boiling point, and I don't think that's an option, unless you strip the case). If your stator is bare, maybe heating above the boiling point of water is reasonable? I hope someone else with chime in with useful information here.

---

I set up a pressure test fixture and tested my plug at 20 PSI -- dry fit, no sealant, as a test -- and it failed: it leaks. I kind of thought that would happen, and it feels tight enough that if I'd applied sealant, I'm certain it'd work fine. Pics and more info over on TeslaMotors.com

 

[hokiematt]

I couldn't find the Tesla spec, but the Toyota standard for HV motor insulation is >100 Mohms between any HV conductor and ground/chassis/motor case.

It might be overkill, but bagging a part and pulling a decent vacuum (>29" Hg) will lower the boiling point of water to room temperature. This is only useful if the glycol has been sufficiently diluted.

 

[asavage]

Ah, 100Mohms, OK. I was way off with 3Mohm -- I must have remembered that from some class I took long ago.

I'd mentioned vacuum boiling upthread, but few have the capability to pull a sufficient vacuum on a part the size of the stator housing. I'd love to see someone's setup to do that though!

Agreed: the glycol wants to hold on to water, and diluting the glycol is essential for later removing the water.

 

[asavage]

Here is the pictures of all the motor parts. . . . I've checked the bearing by turning them and they seem ok. . . . I will take the cover off and check bearings inside to see if it needs new grease or replacement.

I missed this before: your DU still has an Aegis ring (whisker brush), implying that it isn't a late Reman, and the rotor bearings may still contain steel balls rather than the later ceramic ones.

If you pull a cover off a rotor bearing and see steel balls, I would definitely replace both rotor bearings. The Aegis brush proved to not be a reliable diversion of incidental rotor current, which eroded the steel balls in the rotor bearings.

 

[asavage]

I couldn't find the Tesla spec, but the Toyota standard for HV motor insulation is >100 Mohms between any HV conductor and ground/chassis/motor case.

The Tesla component isolation spec is contained in article 6230600, but that's only available to Tesla Toolbox subscribers and I'm not one

Tesla's Isolation/Insulation Test is here:
https://service.tesla.com/docs/Mode...UID-30A90537-F54A-4DF8-86BD-41A80462D8ED.html

It calls for (amongst others) a Fluke 1507 Insulation Resistance Tester (functionally a megohmmeter), which I own. On my (wet) stator, testing at 500v, I obtain 4.7Mohm.
Using my Fluke 79 -- which is a passive ohmmeter with a range to 40Mohm -- on the same stator I obtain a reading of 6Mohm.

The NHTSA copy of the Tesla SB SB-10052460-6095, which is publicly available but from Sep2013:
https://static.nhtsa.gov/odi/tsbs/2013/SB-10052460-6095.pdf
says that if the BMS reports <1.8Mohm to the the diagnostic software (TDS for Tesla in 2013), that that is a Fail.

This post over at teslamotors.com has interesting info, and implies that <1Mohm is OK on the early Model S (reported by the BMS, displayed on the in-car display), and says that other similar Teslas display 3.4Mohm.

https://teslamotorsclub.com/tmc/threads/high-voltage-isolation-issue-on-2022-model-x-plaid.303457/

 

[alflash]

"Fault threshold" for the parameter Isolation resistance for RAV4EV
https://alflash.com.ua/2019/to_rav4ev/isol_resist.png

 

[asavage]

1.8Mohm is a "total" value the the entire HV system must not cross. Each HV component contributes to that total, so the LDU should be well above 1.8Mohm minimum.

 

[tjlee]

1.8Mohm is a "total" value the the entire HV system must not cross. Each HV component contributes to that total, so the LDU should be well above 1.8Mohm minimum.

I’ve done a continuity test with a multimeter between the hv contactors and I get a continuity. Is this ok?

 

[asavage]

Before I forget . . . see the blue oval below? You have coolant corrosion on that wire harness. This is common, and is probably the reason behind the error codes you saw in TPD: the inverter control electronics can't work right. This is why so many LDUs get driven until they won't move, because just rust on the motor side won't keep it from driving, and the increasing noise from rusty parts rubbing or failing bearings won't get most people to look into it, but when the inverter gets wet . . .

I’ve done a continuity test with a multimeter between the hv contactors and I get a continuity. Is this ok?

I'm not understanding: there are NO HV contactors in the LDU -- when we speak of "HV contactors", we are referring to the ones inside the battery case, bolted to the bottom of the vehicle.

You are testing the three big stator windings to stator ground. alflash has a short vid of this, but I can't locate it right now. It's easy to do, if your LDU is apart as yours is, and if you have the tool to measure insulation resistance. My repair doesn't require splitting the gearbox (I have NO coolant in the stator wiring "tunnel" nor in the inverter side cover) so measuring mine was more difficult, and I had to install plastic spacers between the stator leads and the inverter leads that they bolt to (under the orange HV cover and T45 screws).

As noted above, a multimeter or std ohmmeter will give a reading, but a megohmmeter actually applies 500v to the windings and is a much more reliable way to measure insulation resistance than a ohmmeter. But you use the tools you have, and testing with an ohmmeter is better than not testing at all.

I paid $382 for my Fluke 1507 via eBay a couple of years ago, but it's a $7-800 tool new. I figured I'd have enough use for it to own one, and this project (coolant leak on my LDU) was the 2nd time I'd had a real use for it. Previously, I'd been using one from work, but I've retired.