LDU mechanical failure

Topics concerning the Tesla front and rear drive unit drop-in board
jon volk
Posts: 574
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

LDU mechanical failure

Post by jon volk »

It was only a matter of time till something showed up. This unit has seen hundreds of WOT launches, many on slicks. On street tires if traction breaks, it will induce wheel hop, which is quite damaging on its own.

I was powering up a hill as the car shut off and made a whine on deceleration. I did not think much of the sound at the time as the motor was completely dead. As in no precharge, no wifi, nothing. So after some troubleshooting I found there was no power after the regulator on the logic board. This was traced to a broken solder joint on the inductor. Some solder and hot glue for added support and we're back up and running...or not. Powering up the motor resulted in a loud gear noise so I tore it apart. It seems the broken solder joint was likely caused by the shock of the diff exploding.

The bad news is I have to find another set of gears.
The good is that the inverter and motor pinion look fine and I have a Quaife LSD on the shelf that I've been neglecting to install.

There was a good amount of crap in the diff fluid so this likely could have been caught before failure with frequent fluid checks/changes. I would also think that a Quaife diff is a good investment as well.
120370719_349962636152367_1584602465347523353_n.jpg
Formerly 92 E30 BMW Cabrio with Tesla power
JaniK
Posts: 395
Joined: Sun Aug 25, 2019 12:39 pm
Location: Finland
Has thanked: 55 times
Been thanked: 14 times

Re: LDU mechanical failure

Post by JaniK »

Goos to know...

So the addition of an oil filter in model 3 could take some of the metal flakes in before the wear the gears itself?
Any opinions are my own, unless stated otherwise. I take no responsibility if you follow my way of doing things and it doesn't work. Please double check with someone who knows what they are doing.
jon volk
Posts: 574
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: LDU mechanical failure

Post by jon volk »

Couple more
Attachments
651B08E0-5258-47CC-9739-C38C54ACE276.jpeg
6E5F87A4-A6B5-41BD-BADC-3EB54C684AA8.jpeg
D2B63A44-9C19-4DF7-AE6E-52A26C57A850.jpeg
Formerly 92 E30 BMW Cabrio with Tesla power
P.S.Mangelsdorf
Posts: 1057
Joined: Tue Sep 17, 2019 8:33 pm
Location: Raleigh, NC, USA
Has thanked: 223 times
Been thanked: 261 times

Re: LDU mechanical failure

Post by P.S.Mangelsdorf »

Looks like maybe the shaft holding the spider gears snapped, and then the gears ate themselves?

Or do you think the gears were wearing and their failure broke the shaft?
If at first you don't succeed, buy a bigger hammer.

1940 Chevrolet w/ Tesla LDU - "Shocking Chevy" - Completed Hot Rod Drag Week 2023 and 2024

https://www.youtube.com/@MangelsdorfSpeed
jon volk
Posts: 574
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: LDU mechanical failure

Post by jon volk »

Based on the wear on the outer surface of the spider gears facing the carrier, I believe they were eating themselves up prior to the pin breaking.
Formerly 92 E30 BMW Cabrio with Tesla power
P.S.Mangelsdorf
Posts: 1057
Joined: Tue Sep 17, 2019 8:33 pm
Location: Raleigh, NC, USA
Has thanked: 223 times
Been thanked: 261 times

Re: LDU mechanical failure

Post by P.S.Mangelsdorf »

jon volk wrote: Thu Oct 01, 2020 5:59 pm Based on the wear on the outer surface of the spider gears facing the carrier, I believe they were eating themselves up prior to the pin breaking.
I wonder if they seized on the shaft. Seems odd to me that the shaft would break after the gears.
If at first you don't succeed, buy a bigger hammer.

1940 Chevrolet w/ Tesla LDU - "Shocking Chevy" - Completed Hot Rod Drag Week 2023 and 2024

https://www.youtube.com/@MangelsdorfSpeed
jon volk
Posts: 574
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: LDU mechanical failure

Post by jon volk »

That’s my guess is the clearances got to a point of bind. Sort of like running a bent connecting rod until the piston cocks in the bore and the wrist pin gets ripped out. Not that I’ve done that or anything.
It is certainly possible the pin went first. The end result is the same.
Formerly 92 E30 BMW Cabrio with Tesla power
P.S.Mangelsdorf
Posts: 1057
Joined: Tue Sep 17, 2019 8:33 pm
Location: Raleigh, NC, USA
Has thanked: 223 times
Been thanked: 261 times

Re: LDU mechanical failure

Post by P.S.Mangelsdorf »

jon volk wrote: Thu Oct 01, 2020 6:32 pm The end result is the same.
Very true. Best of luck finding new gears.
If at first you don't succeed, buy a bigger hammer.

1940 Chevrolet w/ Tesla LDU - "Shocking Chevy" - Completed Hot Rod Drag Week 2023 and 2024

https://www.youtube.com/@MangelsdorfSpeed
User avatar
TheSilverBuick
Posts: 187
Joined: Mon Jun 08, 2020 5:01 pm
Location: Nevada, United States
Has thanked: 3 times
Been thanked: 2 times

Re: LDU mechanical failure

Post by TheSilverBuick »

jon volk wrote: Thu Oct 01, 2020 6:32 pm That’s my guess is the clearances got to a point of bind. Sort of like running a bent connecting rod until the piston cocks in the bore and the wrist pin gets ripped out. Not that I’ve done that or anything.
It is certainly possible the pin went first. The end result is the same.

haha.

That's impressive. Hopefully the damage is as contained as your initial assessment. Picture of the vehicle?
Need to learn. Starting from the basics.
jon volk
Posts: 574
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: LDU mechanical failure

Post by jon volk »

TheSilverBuick wrote: Thu Oct 01, 2020 6:49 pm Picture of the vehicle?
https://www.instagram.com/tesla_bimmer/
Formerly 92 E30 BMW Cabrio with Tesla power
User avatar
Roadstercycle
Posts: 118
Joined: Mon Sep 23, 2019 10:28 pm
Location: California
Has thanked: 3 times
Been thanked: 2 times
Contact:

Re: LDU mechanical failure

Post by Roadstercycle »

Jon, jump on YouTube and hunt down the guy who tore apart a Model s motor just to get YouTube clicks. Maybe he still has the parts you need.
User avatar
Roadstercycle
Posts: 118
Joined: Mon Sep 23, 2019 10:28 pm
Location: California
Has thanked: 3 times
Been thanked: 2 times
Contact:

Re: LDU mechanical failure

Post by Roadstercycle »

Roadstercycle wrote: Fri Oct 02, 2020 3:41 am Jon, jump on YouTube and hunt down the guy who tore apart a Model s motor just to get YouTube clicks. Maybe he still has the parts you need.
Here it is.

bobwind
Posts: 13
Joined: Sat Oct 17, 2020 2:58 pm

Re: LDU mechanical failure

Post by bobwind »

I am going to be swapping the gears on my ldu's for the 4.5:1. If you can't find a set we can work something out. I'm probably 2 months from tearing them down. I think I know someone that has a dead ldu too. I'm close to you John in Boston, if your are in Ct. Pm me.

Bob
jon volk
Posts: 574
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: LDU mechanical failure

Post by jon volk »

bobwind wrote: Sun Oct 18, 2020 7:32 am I am going to be swapping the gears on my ldu's for the 4.5:1. If you can't find a set we can work something out. I'm probably 2 months from tearing them down. I think I know someone that has a dead ldu too. I'm close to you John in Boston, if your are in Ct. Pm me.

Bob
I found a spare set that should be coming next week as well as a full drive unit. When you have them out let me know. I may pick that set up as a spare now that I know what can break. I’ve had at least 3 sources confirm they’ve seen broken differentials/reduction gears so that seems to one of the weak points.
Formerly 92 E30 BMW Cabrio with Tesla power
bobwind
Posts: 13
Joined: Sat Oct 17, 2020 2:58 pm

Re: LDU mechanical failure

Post by bobwind »

It definitely is. The issue is these boxes a 1) geared for a 4800 to 5000 pound car. Drop the weight in half and you have too much torque for any tire to handle. You are going to be loading the gear box all wrong as the tires spin and catch. And b) designed to have traction control so you do not over load the boxes.

What we need is a 6 or 7 to1 gear set and a really good break based after market traction control system.

On my car I am running dual base units in the rear, locked diff, 4.5 to 1 gears, each motor driving 1 wheel. In the front a sport unit, with same gear, and a open diff. I am going to use the racetronics GPS and wheelspeed traction control system, but I think one could do it all in the inverter if they were clever, which I am not. Car is going to be about 3600 pounds, and this gearing is still not perfect, I probably will not have to much torque off the line, with out drag slicks, and even then it will be tough. Ideally I want a 7:1. It would be interesting and probably profitable to make a variety of drop gears for the ldu and sdu, and educate people about gear ratio to car weight in single gear cars. It is something most swap people overlook until the project gets on the road.
jon volk
Posts: 574
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: LDU mechanical failure

Post by jon volk »

What chassis are you using?

Agreed on the gearing. Even from a usable top speed perspective 9:1 is tough with a 25” tire. 4.5 goes a bit too far for my purposes so a 7:1 would be a great middle ground. Until then I’m diving in the rabbit hole of managing things electronically and already starting on v2 of my VCU.
Formerly 92 E30 BMW Cabrio with Tesla power
User avatar
johu
Site Admin
Posts: 6639
Joined: Thu Nov 08, 2018 10:52 pm
Location: Kassel/Germany
Has thanked: 346 times
Been thanked: 1496 times
Contact:

Re: LDU mechanical failure

Post by johu »

I've implemented traction control using the stock ABS sensors. Very simple, but I "only" have 100 kW available and just one driven axle so I can take the "real" vehicle speed from the undriven axle and control the driven axle to match that +X
Support R/D and forum on Patreon: https://patreon.com/openinverter - Subscribe on odysee: https://odysee.com/@openinverter:9
bobwind
Posts: 13
Joined: Sat Oct 17, 2020 2:58 pm

Re: LDU mechanical failure

Post by bobwind »

My chassis is a rcr superlite coupe. Aluminum semi monocoque, basically a semi street legal track car. It is a lot of work to convert over to electric, but it was the best platform I could find without building new from the chassis up.

I really like the GPS and wheel sensor solution racetronics uses in concept, but I have not seen it in action yet. Zero ev is said to have a open inverter fork with TC enabled but I don't know much about it or how it works.
jon volk
Posts: 574
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: LDU mechanical failure

Post by jon volk »

Sounds like an awesome project. I have a dual vr sensor conditioner in my VCU for taking wheel speeds off the abs sensors but haven’t implemented anything code wise yet there for TC. With slicks traction wasn’t really a big concern. Without the instant there was slip it induced wheel hop so traction control wouldn’t necessarily solve that. The threshold for normal wheel speed differences would mask such minimal slip. I need to solve that issue with proper dampers then focus on reeling in wheel speeds. Tesla had a few years and an army of developers to work all that out. I’m just an idiot bashing the keyboard until stuff compiles.
Formerly 92 E30 BMW Cabrio with Tesla power
User avatar
johu
Site Admin
Posts: 6639
Joined: Thu Nov 08, 2018 10:52 pm
Location: Kassel/Germany
Has thanked: 346 times
Been thanked: 1496 times
Contact:

Re: LDU mechanical failure

Post by johu »

You can find the relevant code here, I hope it's self-explaining: https://github.com/jsphuebner/stm32-car ... r.cpp#L358
Obviously wheelXY are the ABS wheel speeds obtained via CAN, throtmin/max is transmitted to the inverter. The actual control method is just a P-controller with a gain of -10
diff subtraction would have to be swapped for rear wheel drive.
Support R/D and forum on Patreon: https://patreon.com/openinverter - Subscribe on odysee: https://odysee.com/@openinverter:9
bobwind
Posts: 13
Joined: Sat Oct 17, 2020 2:58 pm

Re: LDU mechanical failure

Post by bobwind »

Ok makes sense. for awd you would need to compare all four wheels, or through a GPS on the Can and have it calculate expected wheel speeds for a given mph to compare too.

This sounds good for straight lines, but in turns the left and right wheels need different speeds. And in a car with 3 or 4 motors you can adjust one set or all wheels independently.

I have been thinking about something very similar could be used for torque vectoring if you had steering wheel rotation input, and maybe a IMU on the Can. I even played around with the idea of a lever on the steering wheel that can real time front/rear and left/right bias the power delivery.
jon volk
Posts: 574
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: LDU mechanical failure

Post by jon volk »

...and there were two. Replacement LDU going in the car and a fresh set of gears to repair the original as a backup.

*not sure what’s up with the orientation. They’re straight on my phone
Attachments
07693AE4-FD75-4DF4-B6BA-B40A421ECA3C.jpeg
6C15C32C-8D4E-43AD-9C24-FD51448B7C5A.jpeg
Formerly 92 E30 BMW Cabrio with Tesla power
User avatar
mohalkh
Posts: 7
Joined: Sun Oct 11, 2020 10:11 am
Location: Amman - Jordan
Been thanked: 3 times

Re: LDU mechanical failure

Post by mohalkh »

jon volk wrote: Tue Oct 27, 2020 8:58 pm ...and there were two. Replacement LDU going in the car and a fresh set of gears to repair the original as a backup.

*not sure what’s up with the orientation. They’re straight on my phone
where can i buy a Tesla unit drive for mercedes b250e?
Mohammad Al-Khaddash :)
jap
Posts: 32
Joined: Fri Jun 26, 2020 5:14 pm
Location: Finland
Been thanked: 1 time

Re: LDU mechanical failure

Post by jap »

mohalkh wrote: Wed Nov 04, 2020 6:34 am
where can i buy a Tesla unit drive for mercedes b250e?
https://finndel.no/bildeler.aspx?bk=354705&dk=0137
jon volk
Posts: 574
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: LDU mechanical failure

Post by jon volk »

Back in action.



Interesting observation is this motor is much quieter than the last. There was a pronounced whine, only when the inverter would start pwm generation that is notably absent from this one. It is a lower mileage, newer revision.
Formerly 92 E30 BMW Cabrio with Tesla power
Post Reply