Tesla Large Drive Unit Support Thread

Topics concerning the Tesla front and rear drive unit drop-in board
Locked
User avatar
catphish
Posts: 954
Joined: Fri Oct 08, 2021 11:02 pm
Location: Dorset, UK
Has thanked: 93 times
Been thanked: 179 times

Re: Tesla Large Drive Unit Support Thread

Post by catphish »

Today I powered up an LDU in a vehicle I've been working on and it's dead! I'm not at all sure what killed it, but I will try to detail the exact series of events that lead to its death.

1) The LDU was tested and working at low voltage (60v testing). The parameters were not configured by me, but were confirmed working.
2) The LDU was installed into a vehicle, including complete replacement of its wiring loom.
3) High voltage was applied, the inverter worked, but the motor would only make small vibrations, followed by an overcurrent fault. This failure was repeated a couple of times.
4) Assuming the problem was the encoder, the wiring to the encoder was replaced.
5) Subsequent attempts to power up the vehicle failed with a precharge issue caused by the BMS underreading the DC link voltage. This is probably not relevant, as I don't believe it would have opened the contactors as a result of this, but it was an annoying twist.
5) Several days later, with a new BMS contactor controller in hand, the next attempt, the LDU starts up, but fails with an overcurrent fault the moment it attempts to apply any current to the motor (ie when >0% throttle is applied).

All this leads me to assume that I have destroyed an IGBT in the LDU power stage, but I'm not at all sure how or why. I don't believe there was ever a contactor event during the failed high voltage tests, even though this is the most obvious explanation. I will test continuity of the power stage and see what I find, but definitely fearing the worst.

Not looking for any particular advice about the above, but thought I'd document it while it's fresh in my mind and I'm still upset about it!

The second problem is that I don't believe this inverter was configured correctly originally. I can definitely see that udcgain was incorrect, which leads me to suspect the device may not have been running a configuration designed for an LDU at all. Unfortunately, most other parameters I can't check because the backup I took didn't include hidden values, and most were replaced when I installed a config example provided in another thread by Damien.

While I realise that most misconfigurations shouldn't lead to inverter destruction, I am keep to make sure in future I start with the correct parameters. Would someone be able to point me in the direction of the current "starting point" parameters for the LDU (including hidden ones)? I'd like to get things back to a safe starting point. I'm particularly concerned about those parameters that shouldn't be adjusted, like voltage and current sensor gains, inverter frequency, and dead times, which most people don't need to touch, but I suspect I don't have correct values for.
P.S.Mangelsdorf
Posts: 753
Joined: Tue Sep 17, 2019 8:33 pm
Location: Raleigh, NC, USA
Has thanked: 89 times
Been thanked: 92 times

Re: Tesla Large Drive Unit Support Thread

Post by P.S.Mangelsdorf »

catphish wrote: Mon Feb 07, 2022 11:47 pm All this leads me to assume that I have destroyed an IGBT in the LDU power stage, but I'm not at all sure how or why. I don't believe there was ever a contactor event during the failed high voltage tests, even though this is the most obvious explanation. I will test continuity of the power stage and see what I find, but definitely fearing the worst.
If I remember correctly, several others have had inverters that were damaged before they bought them, and would function at low test voltages, but fail at full pack voltage.


catphish wrote: Mon Feb 07, 2022 11:47 pm Would someone be able to point me in the direction of the current "starting point" parameters for the LDU (including hidden ones)?
Jon's parameters have worked very well for me.
jon volk wrote: Mon Aug 17, 2020 11:07 pm Updated parameter set. Same car/setup, just improved performance and driveability.

Further details on the what/why behind the changes can be found here.

https://openinverter.org/forum/viewtopi ... 6&start=40
Found here: https://openinverter.org/forum/viewtopic.php?f=10&t=195
If at first you don't succeed, buy a bigger hammer.

1940 Chevrolet w/ Tesla LDU - "Shocking Chevy" - Completed 2023 Hot Rod Drag Week
User avatar
catphish
Posts: 954
Joined: Fri Oct 08, 2021 11:02 pm
Location: Dorset, UK
Has thanked: 93 times
Been thanked: 179 times

Re: Tesla Large Drive Unit Support Thread

Post by catphish »

P.S.Mangelsdorf wrote: Mon Feb 07, 2022 11:55 pm Jon's parameters have worked very well for me.
Thank you!

Sadly these configs largely don't include the low level parameters. They're hidden from view, and not exported by default, but are still configurable, and I would like to verify them.
User avatar
Boxster EV
Posts: 434
Joined: Fri Jul 26, 2019 9:32 pm
Location: UK
Has thanked: 35 times
Been thanked: 26 times

Re: Tesla Large Drive Unit Support Thread

Post by Boxster EV »

catphish wrote: Mon Feb 07, 2022 11:47 pm Today I powered up an LDU in a vehicle I've been working on and it's dead! I'm not at all sure what killed it, but I will try to detail the exact series of events that lead to its death.

1) The LDU was tested and working at low voltage (60v testing). The parameters were not configured by me, but were confirmed working.
2) The LDU was installed into a vehicle, including complete replacement of its wiring loom.
3) High voltage was applied, the inverter worked, but the motor would only make small vibrations, followed by an overcurrent fault. This failure was repeated a couple of times.
4) Assuming the problem was the encoder, the wiring to the encoder was replaced.
5) Subsequent attempts to power up the vehicle failed with a precharge issue caused by the BMS underreading the DC link voltage. This is probably not relevant, as I don't believe it would have opened the contactors as a result of this, but it was an annoying twist.
5) Several days later, with a new BMS contactor controller in hand, the next attempt, the LDU starts up, but fails with an overcurrent fault the moment it attempts to apply any current to the motor (ie when >0% throttle is applied).

All this leads me to assume that I have destroyed an IGBT in the LDU power stage, but I'm not at all sure how or why. I don't believe there was ever a contactor event during the failed high voltage tests, even though this is the most obvious explanation. I will test continuity of the power stage and see what I find, but definitely fearing the worst.

Not looking for any particular advice about the above, but thought I'd document it while it's fresh in my mind and I'm still upset about it!

The second problem is that I don't believe this inverter was configured correctly originally. I can definitely see that udcgain was incorrect, which leads me to suspect the device may not have been running a configuration designed for an LDU at all. Unfortunately, most other parameters I can't check because the backup I took didn't include hidden values, and most were replaced when I installed a config example provided in another thread by Damien.

While I realise that most misconfigurations shouldn't lead to inverter destruction, I am keep to make sure in future I start with the correct parameters. Would someone be able to point me in the direction of the current "starting point" parameters for the LDU (including hidden ones)? I'd like to get things back to a safe starting point. I'm particularly concerned about those parameters that shouldn't be adjusted, like voltage and current sensor gains, inverter frequency, and dead times, which most people don't need to touch, but I suspect I don't have correct values for.
What error messages are you seeing (just overcurrent?). Any reason why you’re allowing the BMS to control the contactors over the OI logic board?

When you say the wiring to the encoder was replaced, do you mean that the A and B signals were swapped over?

What firmware pare you using and can you post your parameter file?
Porsche 986 powered by a Tesla large drive unit. Backwards. Build documented here and Instagram @tesla_porsche here.
User avatar
johu
Site Admin
Posts: 5684
Joined: Thu Nov 08, 2018 10:52 pm
Location: Kassel/Germany
Has thanked: 153 times
Been thanked: 960 times
Contact:

Re: Tesla Large Drive Unit Support Thread

Post by johu »

The parameter database openinverter.org/parameters contains also the hidden ones
You can pull a full dump with 192.168.4.1/cmd?cmd=json%20hidden
Support R/D and forum on Patreon: https://patreon.com/openinverter - Subscribe on odysee: https://odysee.com/@openinverter:9
User avatar
catphish
Posts: 954
Joined: Fri Oct 08, 2021 11:02 pm
Location: Dorset, UK
Has thanked: 93 times
Been thanked: 179 times

Re: Tesla Large Drive Unit Support Thread

Post by catphish »

johu wrote: Tue Feb 08, 2022 8:52 am The parameter database openinverter.org/parameters contains also the hidden ones
You can pull a full dump with 192.168.4.1/cmd?cmd=json%20hidden
Thank you. I will start over with https://openinverter.org/parameters/view.html?id=16 when I can. These in fact appear to be the parameters I was running when the failure occurred, so I guess the device was configured correctly (or at least safely). I believe the udcgain in these default parameters isn't correct for the LDU, or at least the LDU I'm working on, but I guess this isn't critical. Thanks again.
User avatar
catphish
Posts: 954
Joined: Fri Oct 08, 2021 11:02 pm
Location: Dorset, UK
Has thanked: 93 times
Been thanked: 179 times

Re: Tesla Large Drive Unit Support Thread

Post by catphish »

Boxster EV wrote: Tue Feb 08, 2022 7:26 am What error messages are you seeing (just overcurrent?). Any reason why you’re allowing the BMS to control the contactors over the OI logic board?
Yes, just overcurrent. This vehicle is using a ZeroEV CCS setup, which handles contactor control. I don't know why it was decided to use this to control the main inverter contactors too, but it works, and to the best of my knowledge never opens the contactors once closed. I will verify this with Seb though.
Boxster EV wrote: Tue Feb 08, 2022 7:26 am When you say the wiring to the encoder was replaced, do you mean that the A and B signals were swapped over?
Yes, the other lines but tested but not modified.
Boxster EV wrote: Tue Feb 08, 2022 7:26 am What firmware pare you using and can you post your parameter file?
I was running 4.97. Parameters attached, which I now believe to be the defaults. The only issue I noted was that udcgain was causing DC bus voltage to appear a little higher than reality. So as most people suggest, this problem is very likely a hardware one, caused by the usual contactor / fuse opening at some point.
Attachments
params(1).json
(1.15 KiB) Downloaded 86 times
User avatar
Boxster EV
Posts: 434
Joined: Fri Jul 26, 2019 9:32 pm
Location: UK
Has thanked: 35 times
Been thanked: 26 times

Re: Tesla Large Drive Unit Support Thread

Post by Boxster EV »

Hmm. I’m not familiar with the Zero EV contactor control logic. Although the original issue does sound encoder related if the motor was just oscillating before it fully let go.

Either way, sounds like you need to remove the cover and test the IGBTs.

I have some spare inverter parts if you get stuck.
Porsche 986 powered by a Tesla large drive unit. Backwards. Build documented here and Instagram @tesla_porsche here.
User avatar
catphish
Posts: 954
Joined: Fri Oct 08, 2021 11:02 pm
Location: Dorset, UK
Has thanked: 93 times
Been thanked: 179 times

Re: Tesla Large Drive Unit Support Thread

Post by catphish »

Boxster EV wrote: Tue Feb 08, 2022 1:13 pm Hmm. I’m not familiar with the Zero EV contactor control logic. Although the original issue does sound encoder related if the motor was just oscillating before it fully let go.

Either way, sounds like you need to remove the cover and test the IGBTs.

I have some spare inverter parts if you get stuck.
Thank you very much, I'll test each leg of the inverter ASAP, and will be a lot more careful when it comes to fuses and contactors in future! Fortunately this isn't actually my vehicle, but I'm consulting on it, so really hoping I can get it fixed without requiring a full new LDU!
User avatar
Boxster EV
Posts: 434
Joined: Fri Jul 26, 2019 9:32 pm
Location: UK
Has thanked: 35 times
Been thanked: 26 times

Re: Tesla Large Drive Unit Support Thread

Post by Boxster EV »

The drive unit will turn with an ocurlim of -300, so it’s good to adjust that down for testing purposes before hitting the accelerator too hard at full voltage.
Porsche 986 powered by a Tesla large drive unit. Backwards. Build documented here and Instagram @tesla_porsche here.
User avatar
catphish
Posts: 954
Joined: Fri Oct 08, 2021 11:02 pm
Location: Dorset, UK
Has thanked: 93 times
Been thanked: 179 times

Re: Tesla Large Drive Unit Support Thread

Post by catphish »

Boxster EV wrote: Tue Feb 08, 2022 1:13 pm Either way, sounds like you need to remove the cover and test the IGBTs.
Testing reveals the following:
1) There appears to be a dead short between all 3 motor windings, and HV- (and hence between the 3 motor windings to each other)
2) The OVERCURRENT error occurs as soon as any throttle is applied, even if HV is not present

I'd appreciate any possible explanation for either or both of these faults. My assumption up to this point has been a dead power stage, but *3* dead power stages seems odd, and why does the error occur when HV is not present?
User avatar
johu
Site Admin
Posts: 5684
Joined: Thu Nov 08, 2018 10:52 pm
Location: Kassel/Germany
Has thanked: 153 times
Been thanked: 960 times
Contact:

Re: Tesla Large Drive Unit Support Thread

Post by johu »

the motor windings are still connected? So even one blown phase leg would show up on all 3 phases because the motor essentially is a dead short.
Support R/D and forum on Patreon: https://patreon.com/openinverter - Subscribe on odysee: https://odysee.com/@openinverter:9
User avatar
catphish
Posts: 954
Joined: Fri Oct 08, 2021 11:02 pm
Location: Dorset, UK
Has thanked: 93 times
Been thanked: 179 times

Re: Tesla Large Drive Unit Support Thread

Post by catphish »

johu wrote: Wed Feb 09, 2022 5:23 pm the motor windings are still connected? So even one blown phase leg would show up on all 3 phases because the motor essentially is a dead short.
Yes they are, I hadn't realized the DC resistance of the motor was so low. I'll test again without the motor thanks!

I'm also interested to know what is triggering OVERCURRENT when no DC voltage is applied at all. Is this more involved than just a current sensor?

Thanks again for all the pointers!
User avatar
johu
Site Admin
Posts: 5684
Joined: Thu Nov 08, 2018 10:52 pm
Location: Kassel/Germany
Has thanked: 153 times
Been thanked: 960 times
Contact:

Re: Tesla Large Drive Unit Support Thread

Post by johu »

Yes the LDU board is not very precise in what caused the error. So actual over current as detected by the comparator is thrown into one pot with desat errors from the gate drivers. I reckon the gate driver detects the shorted IGBT and throws the error
Support R/D and forum on Patreon: https://patreon.com/openinverter - Subscribe on odysee: https://odysee.com/@openinverter:9
jon volk
Posts: 572
Joined: Wed Apr 10, 2019 7:47 pm
Location: Connecticut
Been thanked: 2 times

Re: Tesla Large Drive Unit Support Thread

Post by jon volk »

There’s a light on each Tesla phase gate driver board that will illuminate during a desat fault. That’ll tell you for sure it’s a hardware issue. I don’t have a photo immediately handy.

EDIT. Found my old pic from Instagram.
Attachments
77769FEE-A72B-4CAA-98AD-9C2ED9347F06.jpeg
Formerly 92 E30 BMW Cabrio with Tesla power
User avatar
catphish
Posts: 954
Joined: Fri Oct 08, 2021 11:02 pm
Location: Dorset, UK
Has thanked: 93 times
Been thanked: 179 times

Re: Tesla Large Drive Unit Support Thread

Post by catphish »

jon volk wrote: Wed Feb 09, 2022 10:23 pm There’s a light on each Tesla phase gate driver board that will illuminate during a desat fault. That’ll tell you for sure it’s a hardware issue. I don’t have a photo immediately handy.

EDIT. Found my old pic from Instagram.
This is extremely useful, thanks!

EDIT: LOVE THE PHOTO!
P.S.Mangelsdorf
Posts: 753
Joined: Tue Sep 17, 2019 8:33 pm
Location: Raleigh, NC, USA
Has thanked: 89 times
Been thanked: 92 times

Re: Tesla Large Drive Unit Support Thread

Post by P.S.Mangelsdorf »

I had a new and disturbing fault show up and I'm hoping someone else has seen it.

I was working on cleaning up some wiring, and so I tested the car on jack stands. For some reason, when I come off throttle, not only does regen not kick in, but the motor continues to accelerate.

I've checked the parameters, and cruise/idle is off.

I watched the spot values, and the inverter is definitely seeing both the throttle position (pot value matches what I'm doing with the throttle) and its seeing the revolutions (rpm value changes).

This may be related to a different error where I don't seem to get regen until the pack is at 60%ish. Pack is currently at 95%.

EDIT TO ADD: the motor will come to a stop when the brake pedal is pressed.

Attached are current parameters and screen shot of spot values without the pedal pressed.
Attachments
Screenshot (50).png
params.json
(1.36 KiB) Downloaded 78 times
If at first you don't succeed, buy a bigger hammer.

1940 Chevrolet w/ Tesla LDU - "Shocking Chevy" - Completed 2023 Hot Rod Drag Week
User avatar
johu
Site Admin
Posts: 5684
Joined: Thu Nov 08, 2018 10:52 pm
Location: Kassel/Germany
Has thanked: 153 times
Been thanked: 960 times
Contact:

Re: Tesla Large Drive Unit Support Thread

Post by johu »

udc is 468 and therefor above udcmax of 426V. Probably udcgain off. 0 throttle can result in minimal propulsion when there is no resistance
Support R/D and forum on Patreon: https://patreon.com/openinverter - Subscribe on odysee: https://odysee.com/@openinverter:9
P.S.Mangelsdorf
Posts: 753
Joined: Tue Sep 17, 2019 8:33 pm
Location: Raleigh, NC, USA
Has thanked: 89 times
Been thanked: 92 times

Re: Tesla Large Drive Unit Support Thread

Post by P.S.Mangelsdorf »

Ah that makes sense. Voltage is definitely not that high.
If at first you don't succeed, buy a bigger hammer.

1940 Chevrolet w/ Tesla LDU - "Shocking Chevy" - Completed 2023 Hot Rod Drag Week
Jackk
Posts: 57
Joined: Thu Dec 17, 2020 6:29 pm
Has thanked: 3 times
Been thanked: 2 times

Re: Tesla Large Drive Unit Support Thread

Post by Jackk »

jon volk wrote: Wed Feb 09, 2022 10:23 pm There’s a light on each Tesla phase gate driver board that will illuminate during a desat fault. That’ll tell you for sure it’s a hardware issue. I don’t have a photo immediately handy.

EDIT. Found my old pic from Instagram.
Haha out of interest does this light illuminate the time ? Or does HV need to be connected?
User avatar
catphish
Posts: 954
Joined: Fri Oct 08, 2021 11:02 pm
Location: Dorset, UK
Has thanked: 93 times
Been thanked: 179 times

Re: Tesla Large Drive Unit Support Thread

Post by catphish »

Jackk wrote: Fri Feb 11, 2022 9:35 am Haha out of interest does this light illuminate the time ? Or does HV need to be connected?
I re-tested *without* HV. The light did *not* illuminate, but something did generate an OVERCURRENT fault on openinverter. I didn't try with HV connected.

By disassembling the inverter, I have now identified my fault. One phase output is shorted to battery negative, so there must be a failed low side IGBT on that one phase. I will need a single replacement power stage.

Thank you all for the assistance in debugging! I believe I will be able to source a replacement.
User avatar
gesso
Posts: 8
Joined: Sat Feb 05, 2022 5:12 pm
Has thanked: 1 time
Been thanked: 4 times

Re: Tesla Large Drive Unit Support Thread

Post by gesso »

I've received my V5 LDU logic board from the OpenInverter shop and as I like to know how things work I went to look at the schematics and found that the V5 stuff is on Johannes branch https://github.com/jsphuebner/Tesla-Dri ... master/V5/ rather than Damiens, no big surprise there. But when I went to look at the schematic https://github.com/jsphuebner/Tesla-Dri ... sla_V5.bmp it's pixelated as all get out! :(

Any tips on where to look for V5 schematic that's legible?

Image
User avatar
johu
Site Admin
Posts: 5684
Joined: Thu Nov 08, 2018 10:52 pm
Location: Kassel/Germany
Has thanked: 153 times
Been thanked: 960 times
Contact:

Re: Tesla Large Drive Unit Support Thread

Post by johu »

Isaac, we need you :)
I always rolled with the V4 schematic as there's not much difference. But I agree it should be there.
Support R/D and forum on Patreon: https://patreon.com/openinverter - Subscribe on odysee: https://odysee.com/@openinverter:9
Isaac96
Posts: 656
Joined: Sat Oct 05, 2019 6:50 pm
Location: Northern California, USA
Been thanked: 1 time
Contact:

Re: Tesla Large Drive Unit Support Thread

Post by Isaac96 »

Ah ha!
Printing to PDF instead of exporting as BMP is much better.
This should be correct.
Attachments
Tesla_V5 - Schematic.pdf
(390.94 KiB) Downloaded 135 times
User avatar
gesso
Posts: 8
Joined: Sat Feb 05, 2022 5:12 pm
Has thanked: 1 time
Been thanked: 4 times

Re: Tesla Large Drive Unit Support Thread

Post by gesso »

Thanks for the clear schematic :)
I made a pull request to update the repo
Locked