openinverter forum

This might be a good place to crosspost a link for an LDU issue that might be becoming more prevalent nowadays, with the aging LDU population: the LDU Control PCB has a resistor network consisting of R105/R106, 40.2Ω SMD resistors, that link chassis ground and encoder ("speed sensor") grounds. One resistor becomes intermittent, making the paralleled resistance ~40Ω instead of ~20Ω.

On the RAV4 EV, this manifests in Tesla codes:

• DI_f020 speed sensor problem detected
• DI_w020 Speed sensor problem detected
• DI_f108 Drive not possible because of other system problem

And the inverter will stop driving the motor. The BMS does not open the HV contactors.

@ogmm in Guatemala chased this symptom publicly @ myrav4ev.com and tracked it down to those resistors Nov2024 (link to his "aha!" post). Meanwhile, Aug2024 I had purchased another RAV4 EV that had that symptom. Last week, I was able to reproduce the test conditions (test resistance between ENCG and INV GND, found ~20Ω, pressed on PCB and obtained ~40Ω), and I replaced those two resistors to stabilize the reading. Yesterday, I was able to test drive ~2 miles and had no reoccurrence (other issue prevented me testing further), so I'm calling it "probably fixed". My few posts after the above link document what I found on mine, and how I repaired it.

As this LDU Control PCB is exactly the same part (and Tesla part No.: 013416-00-x) as the one installed in the Model S, I expect more reports of this issue to occur. Once the PCB is removed from the LDU -- and it pretty much has to be removed, as SMD rework on a partially inverted PCB is nigh impossible -- the conformal coating can be heated and scratched off around the resistors, and due to the relatively large format of these resistors -- 2512 -- they can be replaced easily. Probably even without a hot air station, though I didn't have to try that.

New and used boards are plentiful right now via eBay, but knowing what I do and seeing two failures of one of these resistors, I think I'd consider replacing them any time I have a used PCB in my hands, going forward. But that's just because it happened to me plus one other that I know of.

I know that pics are only thumbnails over there unless you register, but photo handling is so much easier there than here that I can't repost them all here, but here's a few.

While I tested at connector D14, it's probably a lot easier to test between chassis ground (LDU case) and ENCG at the encoder. You can't press on the PCB, but if you can get the LDU into a fault state and quickly measure, if it's not 20Ω, you know what the issue is.

Boxster EV wrote: ↑Fri Jul 12, 2024 4:39 pm Having completed the delete in my circa 1250kg Porsche, so far I’ve seen a maximum motor temp spot value of 182 degrees C during spirited driving. That seems pretty hot, however it appears to be running well enough.


Edit: obviously this is the stock motor sensor, not the centre of the rotor.

I think the hottest temp sensor is embedded with the stator windings near the gearbox casing. Thats the temp sensor wire that goes through the tunnel when splitting the gearbox. The other 2 are on the coolant passages into and out of the inverter (I think)

Nothing measures the rotor itself and given rotor is in the center of the motor with not even air cooling... should be hot. This is induction motor's achilles heel... how to cool the rotor dead center in the unit. Industrial induction motors spin a fan mounted on the rotor shaft externally to cool the stator (we have better solution with coolant) and open air channels to the rotor itself.

When I've researched inductor motor failure. The study seems to point at the squirrel cage welding points in the rotor. The cage is just a bunch of bars welded onto 2 rings at the each end of the rotor. Current flows through this cage and its these welding junctions that weaken and fail... then current flows move through neighboring welding junctions and cause them to hotspot and fail.

But 2+ year on since Tesla and aftermarket followers have removed the rotor cooling channel and we don't seem to be hearing about failures. However, I do wonder about 8-10+ years on. Most people were just trying to get past the coolant seal motor failure <= initial 8-10 years so maybe no one really cares.

muehlpower wrote: ↑Mon Jun 10, 2024 6:32 pm What do you think of two plugs, one at the end of the pipe, the other in the screwed-on line? The one in the pipe is held in place by spreading, but can of course also be pinned, in which case a hole must be drilled. It took less than an hour to make.

20240610_201452.jpg20240610_201703.jpg20240610_201948.jpg20240610_201948.jpg

Cool solution. Can be made in China for way cheaper... Probably in the $1-$2/unit range in 500+ MOQ (Minimum Order Quantity) via Made-in-China.com. Does require a bit of supplier interviewing to evaluate which manufacturer can meet the necessary precision.

Hello Howard, long time no hear from you. I like my solution, but the response was minimal. The rollback option is not in high demand, as Trent is going to be permanently deleted. The required precision is low, as the O-rings compensate for most of it.

howardc64 wrote: ↑Sat Jan 24, 2026 7:46 am I think the hottest temp sensor is embedded with the stator windings near the gearbox casing.

Nothing measures the rotor itself and given rotor is in the center of the motor with not even air cooling... should be hot.

Upthread I showed a temp reading of the end of the rotor via an IR temperature sensor pointed at the end of the rotor, which I painted flat black prior to reassembly to aid in IR measurement.
https://openinverter.org/forum/viewtopi ... 126#p71126

At the time, the in-car reading while driving never exceeded 203°F, though that is on a lighter vehicle than a Model S. I only ran that temperature monitor for a couple of months, then removed it, but I have put around 18k miles on the car since.

howardc64 wrote: ↑Sat Jan 24, 2026 7:46 am But 2+ year on since Tesla and aftermarket followers have removed the rotor cooling channel and we don't seem to be hearing about failures.

I admit to be quite surprised as well. I'd have thought that people who push the LDU hard would have grenaded a few rotors by now.