openinverter forum

This ended up in my YouTube feed, hadn't read about him on any of the forms before.

http://turbo-electric.org/teslaresearch.html

Read down to the bottom of the page.

Guy is running custom igbt banks with open inverter board. Going for 455 volt @ 1200 amps on LDU.

He also talks about custom cooling of the inverters to get higher Kw through them. I think it ended up in my feed because of all the immersion cooling searching I have been doing.

Anyway his videos have really low counts so I doubt anyone has really seen his work, and he is doing some really unique things. Definitely a mad man, love it.

Bob

Ok. This guy is officially my new favorite channel.

This is relevant to my interests.

I know right? When I saw it I could not belive I had not heard of him before now. I am seriously considering immersion cooling one of my tesla inverters. I am running them remote from the drive, and it would not be that hard.

I am thinking of building a igbt bank too and just seeing what it can do. This guy needs a patron account. I would send him money just to see what he would do with it.

Bob

I need to digest what he’s up to now that a I’ll soon have a spare functioning LDU and a long winter ahead

From first glance its build a big inverter and force feed more current into the motor. Problem will be saturation. At a certain point you are just making heat with increasing current. Of course no one outside Tesla knows those figures so someone has to try.

100 percent agree. But i think we can guess at the heat soak. 35kw continuous, 90kw for 15 min is what most say is what the ldu can handle before heat soak degrades performance.

https://www.diyelectriccar.com/threads/ ... ar.180761/

Best report of real world heat soak. Stator at 150c in 6-8 laps @ 200kw peak power. A
data point 171 stator, inverter 43, water in 35c, water out 37c. Can not keep Temps down with large ice filled cooling system, means the motor is just cooling limited. They talk of wanting to try oil cooling the stator and windings, but I think this is doomed due to the friction with the rotor the oil will cause. Vaccum might be better.

I suspect the heat generated is around 1 degree per 17.5 kw input, with a 175c max temp, and non linear tail to the curve. So if you dump the full 475 kw of a sport unit, you are looking at 27.14 degree per min, if cooling system starts at 35c, you maybe have 5 min till max temp is hit. This tracks with what s and x owners report.

If this mad man can get 546 kw, I bet he heat soaks around 4.2 mins, or worse, mayde he gets 3.Not much different. But on a drag strip, could work. Do your run or 2, then cool it for 75 min.

What it shows is the real weakness of the tesla hardware is cooling. It seems like something almost no one is working on, and I don't think there is any really good solutions without redesigning the windings. Maybe wrapping everything in copper tubing, but I think the heat is all in windings and rotor surface, no much you can do. Pre chilling to 0 might get you 1 more min. Liquid nitrogen injection or something, but nothing is really a good option. Just the hard limit of the system.

Then again there are very few real world places where you can even use that much power for an extended period. Maybe towing or naval applications.

Ive been looking at a co2 based cooling just ahead of the drive unit to chill the coolant before entering. I also map my water pump duty cycle off POT rather than temp. High pot value is going to increase temps, so why wait to react to the temps elevating when you can get a small head start.

*edit, cooling doesn’t totally ignore temp as that modulates a baseline multiplier, but I was speaking more to high power scenarios.

I think the 2 degree per min removal of heat is a hard limit. The guy i link above is pushing 35c coolent in and getting 37 out. Getting the coolent colder is not going to help if I understand it right. If there was something with a better heat capacity the water... But that's moving into scifi.

Are you going to cool the drive with co2 directly or use it to cool the coolent?

I thought about 2 phase cooling the windings and stator with a 3m novac fluid spray. It would keep the fiction near normal, and pull a lot of heat out with the airflow.i need to run some numbers to see if it would work.

I was planning on the cooling the fluid as a first step. Haven’t considered hardware modification on the drive yet.

bobwind wrote: ↑Sun Oct 18, 2020 3:37 pm I think the 2 degree per min removal of heat is a hard limit. The guy i link above is pushing 35c coolent in and getting 37 out. Getting the coolent colder is not going to help if I understand it right. If there was something with a better heat capacity the water... But that's moving into scifi.

No, that's not exactly how the physics works. The general rule of thumb is that temperature conduction varies with the difference in temperature. Of course, it also varies with the speed of the conducting medium, be it air, water, oil, or liquid sodium. So, perhaps with 35C water flowing at the rate it is meant to flow you get a 2C difference. If the water were, say, 20C you might get a 3 or 4C difference. If it were 0C you might get a 10C difference. If you could flow faster it may or may not make any difference. Temperature conduction also takes time so faster flow might actually make it worse, not better. Of course, if you want super conduction you could switch to something that can absorb heat faster. Gallium would probably fit the bill there but it's a smidge thick.

Still, I demand someone replace all the coolant in their drive train with gallium and see what happens. I won't do this for reasons I can't reveal until someone else tries it.

collin80 wrote: ↑Sun Oct 18, 2020 4:55 pm Still, I demand someone replace all the coolant in their drive train with gallium and see what happens. I won't do this for reasons I can't reveal until someone else tries it.

That's so mean! LOL

Gallium is a no go, what about other liquid metals? They use a sodium-potassium alloy as coolant in nuclear reactors.
Galinstan (gallium, indium, tin) won't work either, also corrosive.
Mercury?

Isaac96 wrote: ↑Mon Oct 19, 2020 2:35 am Gallium is a no go, what about other liquid metals? They use a sodium-potassium alloy as coolant in nuclear reactors.
Galinstan (gallium, indium, tin) won't work either, also corrosive.
Mercury?

Mercury alloys itself with aluminum. They sometimes used this to sabotage planes during WWII. If you paint mercury onto aluminum aircraft parts then they get brittle and fall apart in midair. Fun times! Isn't at least part of the drive train aluminum? Also, mercury is just a tad poisonous and that's not nice either. But, I do wonder if there exists a substance that is both thin enough and has better heat transfer than water. Maybe fluorinert? I know they used such stuff for liquid cooling Cray super computers. As a bonus, it's non conductive so you can directly submerge all the electronics right in it.

Looks like Fluorinert has been considered before.
https://saemobilus.sae.org/content/980430/

Paper attached.

collin80 wrote: ↑Sun Oct 18, 2020 4:55 pmGallium would probably fit the bill there but it's a smidge thick.

Ha. Hahahaha.

I'd love to see someone try this.

Gallium isn't "just" corrosive.

If you had a prison made out of aluminum, and a pea-sized dot of gallium, you could turn the whole prison into the consistency of cheap chocolate. It'll looks okay, but even a light tap would reveal it to crumble.

Here's a great video showing what gallium does:

bobwind wrote: ↑Mon Oct 19, 2020 3:44 pm Paper attached.

Looks like the gist of it is that evaporative cooling with flourinert is better than using 50/50 water and antifreeze but it requires some different parts and a different cooling strategy. It seems they just didn't think it was economical at the current scale of EV production. However, I think it would very greatly benefit Tesla. With flourinert they could directly bathe the transistors in the fluid as well as the stator windings in the motor. Though, one of the heating problems in a motor is always the rotor. It's hard to cool it because it has to spin and it's really only connected to everything else by bearings. So, no matter what you do that thing is going to get hot. It may be that all the cooling in the world won't get much better performance than they already have because the rotor will just continue to be hot. The only solution there might be super conductors. Anyone up for running liquid helium in their motor?

The "mad man" has a new video out. He has aspirations to create a new type of EV.

jalovick wrote: ↑Tue Oct 20, 2020 10:03 am The "mad man" has a new video out. He has aspirations to create a new type of EV.

Don't they all ....

We'll just have to keep watching...

We have spent 100's of hours on attempting to cool the Model S stator. Simply lowering the coolant inlet temperature does not work. No matter what the coolant temperature is, the stator cooling circuit cannot remove heat fast enough from the stator. As well as ice bath for the cooling water, we have used a chiller for testing. Even 10 degC inlet water doesn't help for continuous use. However, if we are talking about intermittent use for a drag run, or maybe even a short autocross run, pre-cooling the stator does work.

We have also tried injecting coolant into the stator/rotor air gap. Pumping fluid (ATF) into the gap did not work as the added friction actually increased the heat production! We had some moderate success with spraying a fine water mist into the air gap and using evaporative cooling. However there were issues with using water inside the drive unit! We contemplated spraying liquid CO2 into the air gap as well, but never did try it as we felt the logistics of carrying pressurized CO2 cannisters on a race car were not good. Jon: If you do any CO2 testing, this is what I would recommend trying. You get the benefits of the cold liquid directly impinging on the stator/rotor surface as well as the heat absorbed as the CO2 flashes to gas.

jalovick wrote: ↑Tue Oct 20, 2020 10:03 amHe has aspirations to create a new type of EV.

...

...

... by using lead-acid batteries.

That's his innovation. He has a 400-mile range, lead-acid powered car, he thinks will last 30 years.

Well now I don't know what to think.

We have spent 100's of hours on attempting to cool the Model S stator. Simply lowering the coolant inlet temperature does not work. No matter what the coolant temperature is, the stator cooling circuit cannot remove heat fast enough from the stator. As well as ice bath for the cooling water, we have used a chiller for testing. Even 10 degC inlet water doesn't help for continuous use. However, if we are talking about intermittent use for a drag run, or maybe even a short autocross run, pre-cooling the stator does work.

We have also tried injecting coolant into the stator/rotor air gap. Pumping fluid (ATF) into the gap did not work as the added friction actually increased the heat production! We had some moderate success with spraying a fine water mist into the air gap and using evaporative cooling. However there were issues with using water inside the drive unit! We contemplated spraying liquid CO2 into the air gap as well, but never did try it as we felt the logistics of carrying pressurized CO2 cannisters on a race car were not good. Jon: If you do any CO2 testing, this is what I would recommend trying. You get the benefits of the cold liquid directly impinging on the stator/rotor surface as well as the heat absorbed as the CO2 flashes to gas.

Doug,
Do you have any photos of where you were injecting ATF and water mist?