Teardown - Tesla LDU - Inverter [SOLVED]

[fredd90]

Hello

Time to reveal Tesla LDU Inner secrets

Unfortunally ive got an IGBT Blowout, New inverter piece is incoming thanx to the kind user "Boxter EV"

As return favor to the forum i decided to teardown my broken inverter piece

The solution i found to deal with the "Elon goo" was Metal cleaner: https://veidec.com/gb/metal-clean
It did not eliminate the goo entirely but its stickyness disapered and behave like gelatine the just blew it away with compressed air.

First preheat to 150-200 degree, then the desolderring begin, ecuipped with 150w soldergun in combination with Vaccum solder station.

Total time from blown inverter piece to this, aprox 4 hours. In this condition it is quite possible to restore with new igbt.

By the way, not the same Igbt as the base unit,
AUPS4067D1-B https://www.amazon.com/10PCS-AUPS4067D1 ... 08KTDJ9HR

And it can not be repeated enough, this is an example of fantastic engineering and problem solving

[jon volk]

Great info and nice to have a pn for the IGBT.

[P.S.Mangelsdorf]

By the way, not the same Igbt as the base unit,
AUPS4067D1-B https://www.amazon.com/10PCS-AUPS4067D1 ... 08KTDJ9HR

It's nice to finally have hard proof confirming this.

Thank you for sharing the teardown!

[Jack Bauer]

I have made this topic a sticky. Thanks to the op for sharing.

[P.S.Mangelsdorf]

I have made this topic a sticky. Thanks to the op for sharing.

I've linked to it from the wiki as well.

[fredd90]

Then it was time to spread some light over all the broken Large drive units, i have against all odds managed to bring life to my P85 drive unit.
After replacing all the igbt in one of the bridges.
As well as all semiconductors in all three drive cards, everything had a short circuit. Not unexpectedly i constantly had overcurrent alarms.

Everything solder together, just the cap left.

All semiconductor on the "isolated" side replaced.
Everything from igbt driver to diodes and n/p transtors.

After final test i decided to add some re-entering silicone.
Couldn't find any transparent unfortunately.

Not a project for everyone but it has been quite educational, spare parts have cost about 4-500eur. Also had to buy a new soldering iron of 200w to cope with the thick laminated copper. The worst part was all the lacquered circuit boards in combination with the extreme lack of semiconductors...

[SRFirefox]

What was the part number of the IGBT driver, out of curiosity? I'm interested to see what choices they made in the design

[fredd90]

What was the part number of the IGBT driver, out of curiosity? I'm interested to see what choices they made in the design

The original one was unavailable, but this one was nearly identical.

https://www.mouser.se/ProductDetail/726-1ED020I12FA2XUM

[rjmcdermott81]

Does this mean someone is going to start refurbishing drive units?

[P.S.Mangelsdorf]

If you replaced just the IGBTs with ones with a higher amperage rating, could you in theory increase power output of the inverter? Or is there something else that would prevent that?

[remy_martian]

My understanding is that's what Tesla did in the Sport inverter...

[fredd90]

If you replaced just the IGBTs with ones with a higher amperage rating, could you in theory increase power output of the inverter? Or is there something else that would prevent that?

That's what they did in the sport version just as remy_martian said, i don't think the design can handle more power, both the capacitors and copper rails are maxed out i guess..

[Skudak]

Did you have a way to visually see what was damaged? My ldu is bricked with overcurrent errors (confirmed the openinverter board and parameters are fine by swapping with another ldu). For now I'm just going to put the motor aside but if i can slowly pick at repairing it that would be great...

[jon volk]

Did you have a way to visually see what was damaged? My ldu is bricked with overcurrent errors (confirmed the openinverter board and parameters are fine by swapping with another ldu). For now I'm just going to put the motor aside but if i can slowly pick at repairing it that would be great...

If you’re “lucky” you’ll see that the goo captures the exploding igbt. It’ll look like impacted ballistic gel with some smoke traces. If it was less dramatic then visuals might be minimal.

[Skudak]

Did you have a way to visually see what was damaged? My ldu is bricked with overcurrent errors (confirmed the openinverter board and parameters are fine by swapping with another ldu). For now I'm just going to put the motor aside but if i can slowly pick at repairing it that would be great...

If you’re “lucky” you’ll see that the goo captures the exploding igbt. It’ll look like impacted ballistic gel with some smoke traces. If it was less dramatic then visuals might be minimal.

Guess I'm a little lucky then, I could mildly see that on a couple igbts on the top phase, was trying to decide if I was seeing failed igbt smoke or if I was seeing the white thermal paste possibly in the goo.

[Robertotesla]

Hello
We have an Tesla P85 inverter with some damaged IGBT in one phase of the inverter, we change all the IGBT of this phase and test the 3 power phase and i think all is ok, we resold all the parts in the phase control card of the phase have the damaged IGBT, but the inverter not work, the message in the car is the same, gate drive fault counter, i think have a problem in one of phase card control…

[fredd90]

Hello
We have an Tesla P85 inverter with some damaged IGBT in one phase of the inverter, we change all the IGBT of this phase and test the 3 power phase and i think all is ok, we resold all the parts in the phase control card of the phase have the damaged IGBT, but the inverter not work, the message in the car is the same, gate drive fault counter, i think have a problem in one of phase card control…

Hello

When my repaired my inverter i got to change all components on the small pcbs, when the igbts fail the igbt driver chip get the full HVDC across the terminals.... you can easily do a quick measure np/pnp mosfet and pwm driver. Mine was shorted/open. Please check the datasheet.

/Best regards Fredrik

[Robertotesla]

Where is the datasheet?

[fredd90]

Where is the datasheet?

Check datasheet to each components, there you vill be able to get a hint of valid measurements regarding operation.

Look at mouser or similar

[Kovansky]

Good evening everyone.

Sorry if my English is not very good, since it is not my mother tongue, in some parts I have had to use translator.

I have been reading through some of the topics on this forum and also Damien Maguire's youtube channel.

To summarize.
I'm about to start a project of converting a combustion vehicle to EV, in pricipio my idea is to take a medium SUV (to have the highest ground clearance and to be able to put as much of the battery pack under the chassis (keeping the center of gravity as even and centered as possible) at the moment I plan to use the engine of a model X as powertrain, but if you have any suggestions of any other TESLA or another engine I would love to hear.

For my part, I have taken several courses in electronics at an institute in my country and I have done several small projects using brushless motors, I consider that I know the logic of how it works and I have a little knowledge beyond the basics, I have also worked with lithium-ion batteries and I know that they must be treated with great care and the points related to them (BMS, minimum discharge, maximum, internal resistance, voltages, etc.).

I have seen several topics on the forum about the tesla motor control unit and I found several topics that I am curious about:
1) "Tesla LDU - Inverter" (I understand for Model S/X).
2) "Tesla Model 3 Rear Drive Unit Controller" (I understand for Model 3/Y).
3) Zombieverter (I still don't fully understand what this system does or how it is used).
4) The controllers for the Tesla motors ("OpenSource Board") have some functions disabled compared to the original Tesla?
I have seen some Tesla controller solutions, but some functions are lost, such as:
Cruise control.
Regenerative braked.
Reduced top speed.
Less power.
Reverse
Etc.

I ask

1. Depending on the motor, should a different controller be used?
2. With the Tesla LDU you can control a Tesla Model 3/Y motor?
3. Which motor has the best power/consumption ratio (which one gives more miles per kilowatt consumed) [ignoring external factors such as wind resistance, vehicle weight, rain, etc.].
4. Which engine is the best of all.
5. What exactly does the Zombieverter do? Is it necessary to control a tesla motor (my conversion will not be with a tesla vehicle, I will only use the motors).
6. Do you have several models of controllers for the same type of motor (if yes, what would they be)?
7. If I want to run a Tesla motor (in a lab environment, like Damien Maguire in his youtube channel), which topics can I use as reference to assemble and program the controller?

Note: I apologize in advance if this post was not in this topic, I'm new to this.
Thanks for all beforehand

[P.S.Mangelsdorf]

Note: I apologize in advance if this post was not in this topic, I'm new to this.
Thanks for all beforehand

Your questions should be asked in a new thread under the Tesla subforum. Once you do so, I and others would be happy to answer.

I'd also suggest perusing the wiki https://openinverter.org/wiki/Category:Tesla

[Kovansky]

Thanks i will check that link

[Kovansky]

Your questions should be asked in a new thread under the Tesla subforum. Once you do so, I and others would be happy to answer.

I'd also suggest perusing the wiki https://openinverter.org/wiki/Category:Tesla

Thank you very much Mangelsdorf for the Wiki that you have sent, practically all my doubts, for the moment, have been solved.

I will continue reading in the forum, as I see that you have many interesting topics.

[P.S.Mangelsdorf]

I just picked up another LDU, which I knew had a mechanical failure. Upon opening it up, I can see some signs of IGBT failure. Where/what do I need to measure to confirm a blown power stage? I know it's been discussed before but I could not find it.

[fredd90]

I just picked up another LDU, which I knew had a mechanical failure. Upon opening it up, I can see some signs of IGBT failure. Where/what do I need to measure to confirm a blown power stage? I know it's been discussed before but I could not find it.

Did answer you on PM also, put this is answer here to help other people,

I did measure in diode mode, the data sheet say that forward bias should be around 0,5v the ones that was in worst condition was totally shorted, the other around 0,1-0,2v.
Then the low ohm(1ohm) power resistors for "gate" control was in Mohm range. probably therefor the igbts blown, not had the possibility to open/close correctly...
Both +/- side engaged at the same time... My final verdict is that the problem started with bad resistors...