Rear Wheel Transaxle Options (Tesla SDU, Leaf, Prius, etc)

[Techshop]

Hi fellas, this is my first post on the forum after reading/lurking for over a year and signing up for an account maybe six months ago.

I'm very keen on doing an EV conversion and I'd love to get some input on drivetrains as before I begin selecting parts and moving forward.

The car that I'd like to convert is a 1983 Toyota Celica GTS. (5 speed, RWD, independent suspension)
My intention is to convert the car using a rear mounted motor & gearbox with a custom subframe and OE suspension arms.

1) Tesla front SDU (model S / model X)

2) Nissan Leaf motor & gearbox with detached inverter mounted nearby.

3) Nissan Leaf motor & gearbox with Prius inverter mounted nearby.

4) Other option that you recommend for rear wheel drive trans-axle?

For any of these options, my reading suggests that the Chevy Volt battery is a good candidate and that if I can find Tesla batteries for a reasonable price they would also be a viable option.




My thoughts on each option

1) Tesla front SDU (model S / model X)
The tesla SDU seems to be relatively inexpensive and can be "hacked" using a replacement logic board from developers here on the forum.

The tesla SDU offers more than enough performance to replace the 2.4L gasoline four cylinder engine (about 105 horsepower).

I would also need pre-charge stuff, a battery, a BMS and a charger.

2) Nissan Leaf motor & gearbox with detached inverter mounted nearby.

The EM61 appears to be the most compact option for a transaxle if the inverter is located nearby, rather than stacked.
Performance of the EM61 appears similar to the original 2.4L Toyota 22RE engine?

The EM57 seems to be very similar, although I would have to de-stack the inverter to fit it in the rear of the car.
Is the EM57 always equipped with a 110kW inverter, or is that only in the 3rd generation?

Is the Nissan Leaf charger/converter a viable option? Does it perform AC charging?

3) Nissan Leaf motor & gearbox with Prius inverter mounted nearby.
I've seen some discussion of this in various forum threads. Is there any major advantage / disadvantage to running a Prius inverter with the EM61 motor and gearbox? I see that configuring the inverter to play nicely with the Nissan motor is more complex than using a simple ZombieVerter and 100% Nissan stack... Is the Prius inverter worth a little more power and flexibility?



For simplicity (mechanical and electronic) the Tesla front SDU seems to be the best option from what I currently understand. I am very open-minded though, and will be happy to hear your thoughts, even if they are for a totally different approach.

Thanks in advance and sorry for the wall of text.

[johu]

All Nissan motors/inverters can be run by sending CAN commands to them, e.g. with ZombieVerter. The Gen2 (EM57 /w 80 kW inverter) can also be run with a replacement logic board to push 130 kW from it. The Gen2/Gen3 charger (PDM) is also hacked but quite bulky. Also includes DC-DC converter. A 6.6 and 3.3 kW version exists.
Then Gen3 stock inverters exist in 110 and 160 kW flavours, controllable via CAN.

You can run the EM57 and EM61 with a Prius inverter, more or less by putting the parameters of the Gen2 logic board on there. You will be down on power though, as the Prius inverter only pushes about half the current compared to a Gen2 Leaf inverter. People have tried paralleling the two Prius inverter stages with limited success if I remember correctly

[Techshop]

Thanks for clearing that up with the Prius vs Leaf inverter.

The Gen2 (EM57 /w 80 kW inverter) can also be run with a replacement logic board to push 130 kW from it.

I read through a build on a GT86 Toyota with Gen2 Nissan components where he reported some issues with "sudden regen".
Does that mean that the regen curve comes on sharply when the throttle is closed quickly?
Is that regen issue something to worry about if I'm staying below 130kW output?

The compactness and low cost of the Gen2 leaf motor and gearbox is appealing, especially with that logic-board upgrade.

I will do more reading on chargers and BMS next.
It seems that some of the Tesla DC chargers are very easy to setup, but I know very little yet about AC to DC charger options.

[johu]

I read through a build on a GT86 Toyota with Gen2 Nissan components where he reported some issues with "sudden regen".
Does that mean that the regen curve comes on sharply when the throttle is closed quickly?
Is that regen issue something to worry about if I'm staying below 130kW output?

Yes it took three years to solve this issue but it has lately been solved for good with the 5.24.R inverter firmware release

[Peter]

Hi Techshop. I think you may need to double check first if a custom subframe is acceptable when it comes to an inspection (or not called for hopefully )

[Bigpie]

Tesla batteries, depending on the voltage you're aiming for, you might need a fair few modules, like a full pack, give or take a couple, to get to the higher voltages usually used.

[MattsAwesomeStuff]

You will be down on power though, as the Prius inverter only pushes about half the current compared to a Gen2 Leaf inverter. People have tried paralleling the two Prius inverter stages with limited success if I remember correctly

... why?

Why wouldn't it be exactly the same? It's just an inverter, and we know the Prius Gen 2 inverter to be capable of some 500hp (if you can keep up with the cooling). Is cooling the limitation, or is there an electrical reason it would yield half the power?

[nkiernan]

Was the GS300H or GS450H option considered? See the smile on Damian's face in the E39! Maybe not the power you're after but wins on simplicity for the RWD?

[Techshop]

Hi Techshop. I think you may need to double check first if a custom subframe is acceptable when it comes to an inspection (or not called for hopefully )

Inspections only apply to a vehicle that has been written off due to a wreck in my state and county.
More importantly, I have the skills and the tools to do high quality work.

[Techshop]

Was the GS300H or GS450H option considered? See the smile on Damian's face in the E39! Maybe not the power you're after but wins on simplicity for the RWD?

I have seen Damien's work on the Lexus running gear, including his test drive. My reasoning for the rear transaxle is to eliminate mechanical components between the power-source and the road surface. This is reduced complexity when working on the vehicle, increased efficiency and reduced weight. The rear suspension design on the old GTS is not the common dual wishbone style, so there is quite a bit of width between the suspension components where a motor can live.

[Peter]

Hi Techshop. Sorry, I wrongly assumed you were in the UK Good luck.

[johu]

... why?

Why wouldn't it be exactly the same? It's just an inverter, and we know the Prius Gen 2 inverter to be capable of some 500hp (if you can keep up with the cooling). Is cooling the limitation, or is there an electrical reason it would yield half the power?

Ah Matt, no rumours and myths sources?

Prius larger power stage is capable of, optimistically 500A (peak!) and Leaf Gen2 of 1000A. The latter manages 140 kW@360V before it desats. Half the current -> half the power. Unless you run a 600V battery or something.

[Techshop]

Hi Techshop. Sorry, I wrongly assumed you were in the UK Good luck.

I'm in Washington State, USA... I'll see if I can navigate the process of adding some info to my profile.

*Edit*
I found a few photos of a complete subframe assembly for the same year Toyota Supra for reference.


The imgur links didn't show correctly, but here they are:

https://imgur.com/a/2WEkQyT

https://imgur.com/a/KZZMYty

https://imgur.com/a/bXaiB3l

[MattsAwesomeStuff]

Ah Matt, no rumours and myths sources?

Prius larger power stage is capable of, optimistically 500A (peak!) and Leaf Gen2 of 1000A. The latter manages 140 kW@360V before it desats. Half the current -> half the power. Unless you run a 600V battery or something.

600A, with both MG1 and MG2 combined, peak, before it Toyota slaps your hands and makes you take a nap.

600A @ 600V = 360kW = 482hp.

If you're running at 360v, then 216kW.

Which is why I said, if you can keep up on the cooling. Obviously that's only good for a few seconds even with liquid cooling, but, outside of a racing, towing, or boating environment, how would you ever sustain 216kW? It's nice to have during acceleration, but otherwise isn't useful.

And yeah, it'll overheat quick, but it'll also deplete our battery in almost no time, so it has a built in duty cycle of "Oh, my battery's empty".

Maybe (probably) I misunderstand the limit. I don't get why, other than cooling, you'd be any more limited by the Prius than the Leaf inverter. If one's 500A peak and the other is 1000A peak, okay, you're down by half, but the Leaf won't even let you pull 1000A, the Prius will. Also are you ever actually pushing either one anywhere near its peak? I don't think you'd be noticeably down on power with either one.

[marcexec]

Not sure what power you are looking for but the Outlander rear drive unit is a very neat package in my opinion: https://openinverter.org/wiki/Mitsubish ... Drive_Unit

[johu]

600A, with both MG1 and MG2 combined, peak, before it Toyota slaps your hands and makes you take a nap.

600A @ 600V = 360kW = 482hp.

If you're running at 360v, then 216kW.

Ah yes, I remember that, and it resembles my findings. Thing is, you have to do some more conversions to arrive at the power this can actually deliver to a motor. So first off paralleling the two inverters doesn't just add their power, unfortunately. You'd have to wire it in a way that pulls exactly 1/3 the current from MG1 and the other 2/3 from MG2 inverter. Failing to do so means one of them trips out and no longer helps.

So optimistically assuming 500A peak current from MG2, you have to first go to RMS current to arrive at something usable. So I_rms=I_peak/sqrt(2)=354A. Something similar happens to DC voltage, modulation can only carry a factor about 0.6 from DC to AC (as discovered on the IPM thread). So 360V DC becomes 216V AC. Now you can start multiplying and arrive at 216*354=76 kW. Back to a bit more than half a Leaf inverter. In practise, Jamie got 70 kW with MG1 and MG2 combined.

I'd say it's up to more experiments how much this package can actually deliver at practical voltages but thats the ballpark.

[Pete9008]

Ah yes, I remember that, and it resembles my findings. Thing is, you have to do some more conversions to arrive at the power this can actually deliver to a motor. So first off paralleling the two inverters doesn't just add their power, unfortunately. You'd have to wire it in a way that pulls exactly 1/3 the current from MG1 and the other 2/3 from MG2 inverter. Failing to do so means one of them trips out and no longer helps.

So optimistically assuming 500A peak current from MG2, you have to first go to RMS current to arrive at something usable. So I_rms=I_peak/sqrt(2)=354A. Something similar happens to DC voltage, modulation can only carry a factor about 0.6 from DC to AC (as discovered on the IPM thread). So 360V DC becomes 216V AC. Now you can start multiplying and arrive at 216*354=76 kW. Back to a bit more than half a Leaf inverter. In practise, Jamie got 70 kW with MG1 and MG2 combined.

I'd say it's up to more experiments how much this package can actually deliver at practical voltages but thats the ballpark.

Not sure this is quote right (and I'm really hoping it isn't as this is the setup I'm using, Outlander rear plus Prius Gen3, and I'd like to achieve 100kW).

The rms calculation is right for single phase motors, where it is there to allow for the variation of voltage/current/power over the supply cycle, it's not relevant to 3phase power which is continuous/constant throughout the cycle. I coulcn't find anything online to calculate power from Q and D voltages and current so derived something from the three phase power equations and park/clarke to get from the Q and D components to power. It's only calculated for one motor angle (I'm too lazy to go though the full sin/cos algebra) but it should apply throughout. It simplified down to:

P = 3/2*(VqIq + VdId)

Ignoring VdId for now just the Iq for a 360V battery and 300A would give Vq=360 * 1.15 / 2 = 200V (allowing for SVM and a few volt drop) so

P = 3/2*(200*300) = 90kW

On a high saliency motor the VdId terms can also contribute to torque and power so it could actually go a bit higher. This agrees with the 100kW power result the simulator gives for this motor and is well within the 500A limit for the Gen3 inverter

Regarding current sharing on MG1 and MG2 this largely depends on the IGBTs used in the inverter. The IGBTs used for MG1 and MG2 look to be the same device and if they are they should share fairly well just by running them in parallel. If they are not the same devices then there is no knowing how they will share without trying it! (I plan on running them in parallel and then using the MG1 and MG2 current sensors to see how well the do share). If the same and given that MG1 has a single IGBT per leg while MG2 has two I would expect running them in parallel to give up to a 50% increase in current (25% would probably be a more realistic though when allowing for device to device variation).

Edit - Having said all that I think the Outlander rear would still be a little low on power for the Celica.

[Techshop]

Not sure this is quote right (and I'm really hoping it isn't as this is the setup I'm using, Outlander rear plus Prius Gen3, and I'd like to achieve 100kW).

Edit - Having said all that I think the Outlander rear would still be a little low on power for the Celica.

Is the Outlander the same as the Highlander and Rav-4 rear drive unit? The Q211 or something along those lines?
I see they are listed as 50kW units, but they don't include any liquid cooling provision.
I agree that those are less than what I'd want out of my little Celica.
*EDIT* I see the Outlander is a Mitsubishi, not a Toyota, so I crossed out my incorrect assumption. I'll be reading more about that Mits transaxle.


So far, the Nissan motor & inverter with the Chevy Volt Battery seems like a good place for me to start.

I have an opportunity to get a 2016 Tesla front SDU from a P90D car for $1000 locally.
Is there any reason to choose that over the Nissan Leaf parts?

Is there a better choice for a battery than the Chevy Volt?

The mechanical fitment and fabrication are not major concerns. I'm a machinist and regularly build gearboxes, splined shafts and suspension parts for my customers. I'm most concerned about easy compatibility of electronic components before I begin collecting parts.

[Pete9008]

For the SDU have a look at this thread https://openinverter.org/forum/viewtopic.php?t=2202 by @catphish. It covers the use of the SDU pretty comprehensively in a very nicely executed install.

I believe the Outlander rear can do around 100kW but have seen little real evidence of anyone achieving that yet (early days though). The SDU can do a lot more, another thread by @catphish is worth a look here https://openinverter.org/forum/viewtopic.php?t=2558. It's about tuning but it also gives a very good idea of the kind of power levels that can be achieved.

On batteries decide what capacity/range you need and have a look here https://openinverter.org/wiki/Batteries.

[marcexec]

[...]

I believe the Outlander rear can do around 100kW but have seen little real evidence of anyone achieving that yet (early days though).

[...]

70kW stock @80S (~300V nominal) means @96S would be 84kW, which the OEM inverter can take.
(BTW, Techshop - it's water cooled, too.)
so 100kW wouldn't be much of a stretch

195nm isn't too bad either and it comes in a neat package with a 7:1 reduction.

[Techshop]

I did a little searching around and I don't see those outlander drive units for sale near my location, while Nissan, Prius, Tesla and other parts are very common. It seems that the Outlander PHEV model was not available in the USA until 2017, so there are fewer years of vehicles available here currently.

[TorontoTurbo]

In Canada we have the Lexus RX400h AWD with a rear motor trans axle the sells very cheap at the scrap yards and looks like it has 65hp.
I wondered why this was not being used more?

[Bratitude]

In Canada we have the Lexus RX400h AWD with a rear motor trans axle the sells very cheap at the scrap yards and looks like it has 65hp.
I wondered why this was not being used more?

The MGR, a few threads on here covering it. people have not had great success with them. Good for an around town vehicle but struggle with hwy speed. they operate at much higher voltages In the Lexus (800v or so) compared to 400v systems

[asavage]

Tesla batteries, depending on the voltage you're aiming for, you might need a fair few modules, like a full pack, give or take a couple, to get to the higher voltages usually used.

Techshop is planning on using Volt cells, but if using Tesla Model S modules (~25v), they can be DIY modified or purchased already modified to reconfigure to 50v each, making them more useful for smaller builds.

http://edisonmotors.net/shop/product/12 ... category=7

Tesla Model S battery Module converted to 12 series (12s) from 6 series (6s)

From my notes a couple of years back:

Height: 3.1 Inches
Width: 11.9 Inches
Length: 26.2 Inches
Weight: 55 Pounds

[justinlinh]

All Nissan motors/inverters can be run by sending CAN commands to them, e.g. with ZombieVerter. The Gen2 (EM57 /w 80 kW inverter) can also be run with a replacement logic board to push 130 kW from it. The Gen2/Gen3 charger (PDM) is also hacked but quite bulky. Also includes DC-DC converter. A 6.6 and 3.3 kW version exists.
Then Gen3 stock inverters exist in 110 and 160 kW flavours, controllable via CAN.

You can run the EM57 and EM61 with a Prius inverter, more or less by putting the parameters of the Gen2 logic board on there. You will be down on power though, as the Prius inverter only pushes about half the current compared to a Gen2 Leaf inverter. People have tried paralleling the two Prius inverter stages with limited success if I remember correctly

Which logic board to push 130 kW for the leaf Gen2? Please provide me a link, i don't see Leaf logic board on the shop section.