Paul and Sabrina's Cheap 3 Phase Inverter (AC Controller) with Field Oriented Control

[thingstodo]

Originally Posted by MPaulHolmes

I've finally got all motors code ported over to the Leaf software, which was much more fancy. Now I just need to test it. Also I was going to add the can streaming. Is there a certain rate can is supposed to stream (I mean how often the data is sent)? And does it have to be at 250 or 500kbps?

There seems to be no standard rate. Some sensors use 250K, some use 500K.

If you re-use the same flags and update rate that you use for the serial stream, perhaps a command message could change that update rate and the flags, just like the command line interface does? It's a lot of overhead, but may be worth it.

I don't have anything running right now. My AC controller is still waiting for some attention, to pop apart the blown IGBTs and replace them with functional ones. And I don't have anything that speaks CANbus ... so I'm not much help.

I'm sure that @jackbauer will give your latest control board a good testing!

[cajunfj40]

Hello MPaulHolmes,

I've been following the development of your drop-in "brain boards" for the Leaf and Volt controllers with much interest. The Leaf board basically turns a wrecked Leaf into an EV Conversion Kit in a fairly straightforward fashion, which is *very* nice. Having a controller for the Volt board means having a heftier controller to pair with a GM Two-Mode Hybrid 2ML70 transmission, too, also quite nice. This means that I have a reasonable path forward to use either system, though on the 2ML70 I'd still need to figure out how to shift it as it is computer controlled. It can't be *that* different than a 4L80E, considering it has the same number of clutches, only lacking the lockup torque converter.

That said, I have a few questions that I can't seem to dig up the info for.

What voltage and current can your brain boards safely control through the Leaf and Volt controllers? I know the Leaf was rated at 80kW, and ORNL testing showed that to be essentially a continuous rating - Temperature leveled off without getting too high to be a problem when running at speed/full power. I know from following other developments that the Leaf motor is capable of quite a bit more intermittent power/torque, but I am not familiar with the limits of the components in the OEM controller.

The 2ML70 has two motors rated at 60kW each with the voltage and current limits placed on them by the OEM controller. They are basically HVH 250 Remy cores. I've posted somewhere else on here a few times about it. The Volt controller already is specified to a higher voltage, so that's a boost to about 80 or 90 kw peak from each motor right there, though probably still thermally limited to about 60kw each, maybe a bit less. How much can the Volt controller reliably control?

Can you (or other's who have played around with the boards) shed some light on this? If it won't cause you problems with potential customers, it might be a worthwhile addition to the product descriptions on your webstore to dangle how much power they can reasonably expect to control when using your brain boards to harness these OEM controllers. There are always time limits based on how well the OEM cooling systems can dissipate heat, so users must be cautious of that, but it would be very useful to me to be able to figure out how much "launch" power I can get. I've already figured out that 80kW continuous is "enough" for my needs, now I'm trying to figure out how much "I want more torque ***now***!" instant fun I can command.

Thanks for all your hard work, and to everyone else who's helped along the way. This is starting to get to the point where I can conceivably actually put a real budget together to maybe do a conversion! (Yes, buying a used OEM hybrid or BEV would be far easier and likely less expensive - but they don't make what I want at a price point I find reasonable yet.)

[MPaulHolmes]

For the chevy volt board and leaf board, the volt uses 350amp/volt current sensors, and the leaf uses 400amp/volt current sensors. The full range from 0.5 to 4.5v would mean 700amp peak and 800amp peak for what the controller can actually read through the IGBTs. I can't tell what the leaf igbts are. They are 3 half bridges though, and are really big, but have insulating paper under them rather than metal. I honestly have no idea what their true rating is. The volt has 2 6 packs, that I think are rated for either 400amp or 600amp. The volt 6 pack is quite a bit smaller surface area than the 3 half bridges in the leaf. I do know the volt driver board fails at around 405 or 410 volts, almost like it's a crowbar circuit. Teh driver board makes the 18v or so from the high voltage. I've tested it down to around 50vDC, and stepped up by like 24v at a time (I was using solar panels,like an idiot, rather than a power supply). It failed at about 410vdc. my son is asking for a cookie and wants me to go with him.

[doobedoobedo]

So this is an epic thread, I've read a fair amount of it - and searched. There is mention of a wiki - but never a link. Can anyone share a link?

[MPaulHolmes]

I don't know how to do a wiki, but I can email you a bunch of stuff if you want. It's pandspowerelectronics@gmail.com
there was a wiki at some point, but I don't know where it is now.

[doobedoobedo]

Thanks Paul, you have incoming mail .

[jackbauer]

Video from today :
https://youtu.be/JO21e1XR6Ds

Inverter installed in the track car project and ready for some testing.

[cajunfj40]

Hello MPaulHolmes,

Thanks for that! You jogged a memory that I'd seen something, so I dug this up: https://www.energy.gov/sites/prod/fi...ess_2013_o.pdf

Teardown of 1st gen Leaf inverter shows 3 IGBT modules per switch, 18 modules total for the full 3-phase bridge. Each IGBT estimated to be ~1000V/300A rated. So that puts the assembly at ~900A peak. The cap is rated at 600V. At 80kw continuous rating, 360V nominal battery voltage, that's about a 222A continuous rating for the tripled modules, seems conservative.

Also found this: https://www.diyelectriccar.com/forum...-179922p2.html

Post #12 links to a 404'd paper, but poking around confirms some of the numbers.

Volt 1st Gen Inverter unit A 220/150 Peak/Cont. Arms, unit B 425/150 Peak/Cont. Arms.
Volt 2nd Gen Inverter 325/150 Peak/Cont. Arms for both units.

With your information that the Volt inverter driver clamps/dies/crashes at 405-410V, probably can't go much higher than the nominal 360V pack. That puts the 2nd Gen Volt inverter at 117/54 Peak/Cont.kW on both inverters, for a combined 234/108kW Peak/Cont. unit with the 2ML70 transmission.

If we use the 2nd Gen Chevy derating curve (325/150 = peak is 217% of continuous) on the Leaf, then we can run the Leaf at 481 Arms peak. 1st gen Inverter B ratings are higher, with peak being 283% of continuous. On the Leaf, that would be ~630A. With a 360V nominal pack that's a 227/80 Peak/Cont. kW rating on that inverter. Voltage may be pushable higher - not sure how close to the 600V cap rating one should get?

So roughly comparable power available from both inverters, though the Volt divvies it up across two outputs, so need 2 motors.

Definitely two quite viable candidates! Thanks again for all your work!

Now I need to see which chargers and DC/DC converters and EPS units and HVAC compressors, etc. have been hacked... I'd love to repurpose a 2nd Gen Leaf heat-pump setup, for example.

[hehehole]

Hi,

I'm new here and have been reading quite a bit already. Still need to do a lot and I definitely need to work on some of my maths as it looks like a lot if the control logic is based on some mathematical functions.

I found the next university paper on an extended back EMF algorithm but my
rusty maths and lack of knowing, what seems to be symbol conventions, make me not understand a lot of it. I will share it anyway.

If there is someone who could point me to the right reading material to educate myself on these things, I would appreciate it.

Extended Back EMF model for permanent magnet synchronous machine with different inductances in d- and q-axis:
publik.tuwien.ac.at/files/PubDat_176266.pdf (couldn't post a link as I'm new, you can also look for the title in google)

Regards,

Geoff

[xstanx]

Thank you very much, Paul, for the documentation. I have already improved the schematic diagram of the entire record. I drew a diagram from photos so there were a lot of mistakes (30). I did not find the markings of two systems: U32 and U31, the converter isolated from 24 V to 5.1 V already have. If someone has symbols of these layouts, please give me here on the forum. I have drawn a schematic diagram in its entirety to get to know and analyze the principle of the entire controller system. I am currently familiarizing myself with the source program. You've been working here on the forum for many years and for others to create such a successful design. Currently on the stock exchanges you can buy engines from EV cars at the price of scrap. These engines can be developed to suit your needs using the Paul project. I'm promoting this project among friends in my country.