Repairing a Flooded Tesla Model S : HOW-TO

[JRP3]

you have _______ that makes something work outside it's intended capacity without really knowing what the tolerances are or ______ any safeties put in place by the original manufacturer.

You mean the way hot rodders have been doing since the beginning of ICE's? I'm pretty sure removing the ICE and all associated components from my car and replacement with an electric motor, controller, and lithium batteries was not something intended by the OEM, nor were any of the various more conventional modifications I've done to other ICE's over the years. People are always going to tinker, and sometimes it's not going to turn out well. That's life. We really don't need any more net nannies in this thread.

 

[smac]

I still have >400v worth of 3000F Maxwell boost caps... (I think short circuit is like 5KA)... anyone want to pay me big $ to make the lightest/fastest ESS pack! *Note range is limited to 1 mile, unless I put in a small pack and boost converter in too.

No takers?

Will they fit in the empty hump space of my 60 pack ? :tongue:

On a more serious note, messing around with the cars right now is being done on salvage cars, but as they depreciate I can't help but think they will get down to a price point where people won't care about the warranty and this sort of stuff will be within the tinkering comfort zone of far more people.

The 60 being the cheapest probably will get there first, and I suspect has the most opportunity on a % basis to see performance gains

 

[markwj]

Wait, this thread wasn't about a roadster?

Its about a roadster and my quest to figure out if it can be insured

I'm completely confused. I thought it was about getting the salt out of that thing on the right, and putting the result into that thing on the left. Or is that the Roadster 3.0 thread?

 

[fabi]

Empty...

Could you please make a picture of the empty front Area? :smile:

 

[Btr_ftw]

Soooooo..... If an s85 can be easily converted to a p85. Did Tesla leave any overhead for a p85s performance to be increased?

 

[JER]

Did Tesla leave any overhead for a p85s performance to be increased?

I'd guess that's going to depend on how you feel about reliability.

 

[Btr_ftw]

I'd guess that's going to depend on how you feel about reliability.

Don't care. Want speed.

Ive overboosted a few ICE cars and I was smiling ear to ear as they were being loaded onto flatbeds.

 

[JRP3]

Soooooo..... If an s85 can be easily converted to a p85. Did Tesla leave any overhead for a p85s performance to be increased?

I would guess if you can over ride/rewrite the software you can put as much current through the inverter/motor as you want until the weakest component melts or blows up.

 

[Btr_ftw]

I would guess if you can over ride/rewrite the software you can put as much current through the inverter/motor as you want until the weakest component melts or blows up.

I like where this is headed.

 

[tomas]

Btr-ftw: I've enjoyed this thread… both your adventure and your sense of humor. I would like to see you succeed with this project, but I'm concerned by this recent banter. Someday, at some point, I believe you are going to need help from Tesla, and you will probably want to get your rebuilt car re-certified by them. If your intention is to restore the car, they ought to cooperate, and you will have a truckload of angry followers if they waffle. If your intention (or perceived intention based on posts in public forum) is to boost it until something melts and it gets loaded on a flatbed, I suspect you will be hard pressed to get their cooperation. Humor aside, I hope your primary objective is just get it running and re-certified.

 

[spaceballs]

And importantly A or B pack

- - - Updated - - -

Hmmm.... 60 pack. Can you confirm the contents of the front hump area on the 60 pack?

Now with picture, I find it strange that certain locations cells are missing (circled in red, I assume since it's a smaller pack it needs less cells), but what gets me it's in a different location in every module. Why not make it the same location in every one.

 

[kennybobby]

@spaceballs, Did it have the same 630A fuse as the 85 packs? Thanks for the pictures.

p.s. what are you planning to do with those nodules?

p.s.s. It looks like each module is identical in the layout of the empty cell positions, each of the six bricks that make up a module have a unique shape and polarity, but the layouts are the same module to module.

 

[Btr_ftw]

Btr-ftw: I've enjoyed this thread… both your adventure and your sense of humor. I would like to see you succeed with this project, but I'm concerned by this recent banter. Someday, at some point, I believe you are going to need help from Tesla, and you will probably want to get your rebuilt car re-certified by them. If your intention is to restore the car, they ought to cooperate, and you will have a truckload of angry followers if they waffle. If your intention (or perceived intention based on posts in public forum) is to boost it until something melts and it gets loaded on a flatbed, I suspect you will be hard pressed to get their cooperation. Humor aside, I hope your primary objective is just get it running and re-certified.

No I was just kidding about overboosting, but still curious as to the upgrade potential.

 

[satoshi]

I would guess if you can over ride/rewrite the software you can put as much current through the inverter/motor as you want until the weakest component melts or blows up.

I like where this is headed.

Speaking from experience... components melting/blowing up gets really expensive really quickly...

 

[spaceballs]

I'm completely confused. I thought it was about getting the salt out of that thing on the right, and putting the result into that thing on the left. Or is that the Roadster 3.0 thread?

The pack I pictured next to the roadster is not the same pack that Btr_ftw is working on. I posted it in case Btr_ftw wanted the ESS enclosure in restoring his car.

- - - Updated - - -

@spaceballs, Did it have the same 630A fuse as the 85 packs? Thanks for the pictures.

I think it's 550 Amp, I'll check later today.

p.s. what are you planning to do with those nodules?

Undecided at the moment, we got a few ideas.

p.s.s. It looks like each module is identical in the layout of the empty cell positions, each of the six bricks that make up a module have a unique shape and polarity, but the layouts are the same module to module.

I'll take a closer peek then.

 

[Ingineer]

I actually believed this too until I couldn't confirm that with my data. It kind of doesn't make much sense, in hindsight, since the IGBTs on the S85 motor I've seen a tear down of were already the best available in that line of ICs at the time that Tesla designed these things.

<snip>

Bigger IGBTs, slower temp rise. More/bigger capacitors, slower temp rise. Better heat sinking, slower temp rise. I can go on. The temperature rises during full throttle acceleration in both is virtually identical, as would be expected from.... well, identical hardware.

IGBTs generally have a fixed voltage drop. Upgraded parts with more peak current handling capability does not change the efficiency, so the heat production is still the same. However it greatly alters the reliability picture.

Your assumption has one major flaw; if they were identical, which many other parts are across P/non-P, they would have the same part number and they wouldn't bother stocking 2 variants for the reasons you explain.

I have been told on the very early cars, all units were built to "P" specs, and they software-limited non-P. As a cost-saving measure and to improve reliability (early DUs pretty much all failed), they created an improved P unit and a standard unit.

Until one you upgrade blows up, I guess it's all unproven.

 

[tom66]

IGBT current rating is dominated by hotspots on the die. If the die has imperfections then it may fail at a higher operating current, even if operating at the same temperature. And failure of the IGBT often destroys the gate driver and any other IGBTs in the same half-bridge (looks like 9-12 IGBTs per phase in Model S so they would all go pop.)

A device used for plasma TV displays may be rated at peak 180A (for example https://www.fairchildsemi.com/datasheets/FG/FGA180N33AT.pdf), but there is no way on earth you could use it for a motor controller with an average operating current of 180A because the device would not be reliable in such an application.

 

[spaceballs]

IGBT current rating is dominated by hotspots on the die. .... A device used for plasma TV displays may be rated at peak 180A (for example https://www.fairchildsemi.com/datasheets/FG/FGA180N33AT.pdf), but there is no way on earth you could use it for a motor controller with an average operating current of 180A because the device would not be reliable in such an application.

How would these do in a motor controller? IXGX72N60B3H1 IXYS | Discrete Semiconductor Products | DigiKey

 

[tom66]

72A continuous current might be a limiting factor, you'd have to parallel them up to get near Model S output power (~400kW / 1400A.) Thing is with IGBTs is that they have a negative temperature coefficient. As they get hotter the forward voltage of the device decreases, which causes more current to flow through the device and increases the device's temperature. This can quickly lead to a thermal runaway condition where a minority of devices hog the full operating current which can lead to IGBT failure and destruction of the inverter. (And by destruction I mean craters in PCBs and components due to the fault energy being significant.)

To avoid it the devices should be thermally coupled to minimise temperature differences, and they should also should be well matched (minimal forward voltage difference between devices.) I'm sure avoiding this thermal runaway condition was a critical concern for Tesla in the design of the drive inverters for their vehicles and will be one of the factors limiting continuous operating current (higher peak current, provided it remains within the ratings of the parts, both current and temperature, is usually acceptable.)

MOSFETs are a lot easier to put in parallel due to their positive tempco, which means as they get hotter the forward voltage (~RdsON) increases, reducing the current flowing through paralleled devices. Because of this, MOSFETs are a lot easier to connect in parallel, although typically the on-state losses for high drain voltages (dollar-to-dollar with IGBTs) make them impractical or too expensive to use for most high-voltage (>150V) inverters.

 

[spaceballs]

72A continuous current might be a limiting factor, you'd have to parallel them up to get near Model S output power (~400kW / 1400A.) ...

These are in Roaster motor controllers, ~84 of them, with output of 240kw.