openinverter forum

FYI, it won’t take much of an oil pump since there are no clutch packs. It only needs lube flow (for all the bushings/bearings). I suspect 15psi and 1GPM should be plenty.

A little earlier I took the MG1 end (bell housing) off of my L210. It was actually very straightforward. The only tricky (messy) part was that you have to take the oil pan off to access three of the bolts. Everything is the same as the GS450h CVT as far as I can tell from looking at the pics in the OSTI doc, except, that the ring gear has teeth on the outside and there is a second gear attached to the oil pump which is driven by the ring gear. So, it looks like MG1 or the ICE can drive the oil pump. Those little gears on the oil pump have some sort of one-way mechanism. They only catch going forwards (counter-clockwise in that pic).
I took a quick little video to demonstrate as well.

So, I want to check (and re-check) everything, but, it looks like you could use this CVT in the same way as the GS540h one. Lock the input shaft. MG1 and MG2 will both drive the oil pump. Or, if you want to just use MG2, that will work as well.

Anyone see anything wrong with this?

MG1 will spin backwards if you lock the input shaft. When you do this does the one way clutch spin the oil pump? Check the output shaft to make sure you are spinning the correct direction.

Excellent work.

slow67 wrote: ↑Thu Oct 01, 2020 12:19 am MG1 will spin backwards if you lock the input shaft. When you do this does the one way clutch spin the oil pump? Check the output shaft to make sure you are spinning the correct direction.

Yep, if I spin MG2 (by turning the output shaft) clockwise (looking at it from the input shaft end) and hold the input shaft still, MG1 spins counter-clockwise. And yes, the oil pump still turns, because the MG2 oil pump gear is still turning in the correct direction. Note the little red mark I put on the MG1 gear.



Here's some detail on the clutches on the oil gears so you can see how they work.


So, I've been puzzling through this a little more today to make sure I'm not being dense. When the ICE turns the input shaft (clockwise), then the MG1 gear also turns clockwise (at a higher speed).



When the ICE is turning MG1 clockwise it's acting as a generator. So I guess clockwise is backwards for MG1.

Then I asked myself, if I want MG1 to contribute to traction, which way do I need to spin it? So, I pulled the ring gear out a bit to make it easier to turn. Then turned the MG1 gear. The test here is backwards as I want the ring gear to turn clockwise, but you get the idea. This confirms that counter-clockwise is forward for MG1 and I want to spin MG1 counter-clockwise to have it contribute to traction.



So, the only remaining concern is the speed that MG1 will end up spinning - as others have noted - when the input shaft is locked. I'll try and get a rough measure of the ratio between MG1 and MG2 when the input shaft is locked tomorrow.

The alternative is to not lock the input shaft and do the Prius transaxle weld job on the planet carrier. Then MG1 and MG2 would be locked together and spin at the same speed. But I think I would do some bench testing before doing that. Plus, you'd have to spin MG1 backwards if you do the weld job.

That’s great news on the oil pump! That means that it is purely mechanical and we don’t need an external pump. We are only lubricating bushings/bearings, so no oil pressure at stand still is fine. I’m sure Toyota planned it this way.

If Toyota has stuck to their typical ratios, the PSD is about 2.6:1. If you count the teeth on the ring and sun gears I’ll do the math.

The preferred way to lock the input shaft, so you get more shaft torque. If anything, MG2 should be your speed limiting factor, as I believe the new MG2 planet ratio is 3.333:1. Further disassembly should tell us.

slow67 wrote: ↑Sat Oct 03, 2020 5:55 pm That’s great news on the oil pump! That means that it is purely mechanical and we don’t need an external pump. We are only lubricating bushings/bearings, so no oil pressure at stand still is fine. I’m sure Toyota planned it this way.

If Toyota has stuck to their typical ratios, the PSD is about 2.6:1. If you count the teeth on the ring and sun gears I’ll do the math.

The preferred way to lock the input shaft, so you get more shaft torque. If anything, MG2 should be your speed limiting factor, as I believe the new MG2 planet ratio is 3.333:1. Further disassembly should tell us.

I did this little test earlier before seeing your response. MG1 turns approx 2.5 times (it points roughly down when it stops) for 1 turn of MG2. So, this lines up with your 2.6:1 ratio suggestion. But I can count the gears to be sure.



Yeah, I'll do more investigation to determine the ratio between MG2 and the output shaft. While I have it open

Yeah, I'll do more investigation to determine the ratio between MG2 and the output shaft. While I have it open

According to Toyota, the ratio is 3.333 with a final drive of 2.764 for the GS300h and 2.475 for the IS300h

https://media.toyota.co.uk/wp-content/f ... ATIONS.pdf
https://media.toyota.co.uk/wp-content/f ... ATIONS.pdf

That is a damn clever system:) So even with the ice stopped the oil pump can run when in ev mode. I guess it might be time to start loking at porting the new dual motor design to the IS300H inverter:)

Everyone was thinking Toyota downgraded their system when going from the L110 to the L210, while in reality they made it more reliable and cheaper at the same time. It looks like guessing just doesn't work.

So would you say the 2-stage gearset at the back of the L110 is useful or important? In relation to road speed/energy consumption/road noise etc.

sfk wrote: ↑Sun Oct 04, 2020 7:32 pm So would you say the 2-stage gearset at the back of the L110 is useful or important? In relation to road speed/energy consumption/road noise etc.

Toyota probably dropped it for simplicity/reliability. It is important if you need the performance though.

I am watching like a hawk for the L310 transmission. They moved the output of the PSD directly to MG2. So a welded planetary would make them spin at the same speed. Then has a 4 speed gearset behind it (so both MG1 and MG2 get the advantages of the additional gearing).

Had a few minutes at lunch there and popped open the output end to see what's going on.

Here is the output end. The planet carrier is directly connected to the output flange. If you look closely, you can also see splines on the inside there. These mate with the output shaft from the ICE/MG1 stuff up front. Also, there's the mechanical parking lock down the bottom.
On the other end we have the output shaft from the ICE/MG1 as I say, driving the output flange directly. Then you have the MG2 gear sitting around that. The ring gear is locked stationary with 6 stud things.
It takes roughly 3-and-a-bit turns of MG2 for one turn of the output shaft. So, that lines up with the 3.333 ratio suggested.

If someone wants to double-check:
- Ring gear has 70 teeth
- Planets have 20 teeth
- MG2 gear has 30 teeth

So, at a max rpm of 14000 on MG2, the output shaft would be spinning at 4242 rpm, right? And at that speed MG1 would be spinning around 11000 rpm. Is that right?

Confirmed then, ratio is 3.3333333...

the math is:
ratio = 1/(S / (R+S))
R = Ring gear teeth
S = Sun gear teeth

FYI this is for a simple planetary gearset. The older L110 used a compound ravigneaux gearset.

As this thread contains very important information on the L210 I have made it a sticky.

Jack Bauer wrote: ↑Sun Oct 11, 2020 9:22 am As this thread contains very important information on the L210 I have made it a sticky.

Cool, I'm planning to write all of this up on the wiki as well.

Started a page for it here : https://openinverter.org/wiki/Toyota/Lexus_GS300h_CVT

i think the one way bearing is called a Sprag bearing.
https://en.wikipedia.org/wiki/Sprag_clutch

Disappointing that this unit isn't available in the US. Looks like a perfect fit for most EV conversions, especially with the deletion of the external oil pump

chrskly wrote: ↑Sun Oct 11, 2020 8:31 pm Started a page for it here : https://openinverter.org/wiki/Toyota/Lexus_GS300h_CVT

I've added a link from the Toyota main page as well. The matching IS300h/GS300h inverter model is G9200-30132, sometimes labelled as G92A0-30052.

I'm tempted to grab one myself now.

SRFirefox wrote: ↑Mon Oct 12, 2020 2:50 am Disappointing that this unit isn't available in the US. Looks like a perfect fit for most EV conversions, especially with the deletion of the external oil pump

If you know of a good importer who can freight forward to the US from Japan, you can get them fairly easily, both the L210 and L210f (AWD) variants...

https://auctions.yahoo.co.jp/search/sea ... 10&x=0&y=0

Great information in this thread!
Was about to get a L110 (GS450H) transmission, yet the L210 discussed here seems to fit even better for my project (intending to convert a classic Mercedes with RWD).

But how to get the inverter to work? On the EVBMW webshop I can only find a bare / unpopulated PCB board without components for the IS300h: https://www.evbmw.com/index.php/evbmw-w ... board-copy.
I addition, it says that as of March 2019 the design is not yet tested.

I have basic soldering skills, yet I am not confident soldering SMD parts and sourcing all the needed components.

Are there populated boards available in the meantime?

There are two ways to get the IS300H inverter working :

1)someone with a working and driving IS300H captures the serial data between the hybrid ecu and the inverter and then writes the code needed for the GS450H VCU.

2)someone designs, prototypes, tests and releases a logic board for the IS300H inverter enabling it to work with the openinverter system.

I have removed that bare board from the webshop as it was only ever intended for testing not use in a vehicle and is causing confusion.