[WIP] Jeep 2ML70 Conversion

[Eagle411]

I'm going to convert some type of Jeep vehicle from the mid to late 90s or early 2000s using a 2009 Chevy Tahoe Hybrid transmission. I don't have the vehicle picked out yet. I'm trying to develop the powertrain right now.

Powertrain Design: Zombiverter with a Prius Inverter driving the Transmission.
About the 2ML70 vs GS450h/LS600H... I actually picked up an LS600H cheap, but those are pretty rare around here. I got very very lucky. 2ML70s are more common, but not common either. The spec on these is 2 60Kw motors. I suspect the 2ML70 motors are very under rated. The 2ML70 is rediculously larger and heavier than the LS600H (maybe 100lb/45kg heavier). I realize size isn't everything (maybe its just more durable or inefficient), but I've got to beleive some of it is untapped power. I suspect that they could do 80kw and maybe 100kw with propper cooling. Anyway ... I'll find out.

I haven't given much thought to the battery or much of anything else. I did pickup electric vaccume and power steering pumps while I was cruising through the junk yard.

Here's what I know:

The Auxiallary oil pump kicks on between 230 and 280Hz
I can command the transmission gears using 0A2 byte 2.(counting from 0).
0A2.B2=0x20 -> ?, 1F5.B4=0x0B
0A2.B2=0x30 -> Park,
0A2.B2=0x40 -> Reverse,
0A2.B2=0x50 -> Neutral,
0A2.B2=0x60 -> Drive
0A2.B2=0x0D ->?, 1F5.B4= 0x0D

The 1F5 status indicates that it was commanded, but it didn't shift. I am working to enable the auxiallry pump so I can get it to actually shift.

Changing other bits later in the same message causes solenoids to change states. I took the oil pan off and took some solenoid measurements. I'm not sure what's really going on, but the pressure solenoids operate at 3.01185Khz. The shift solenoids appear to be on/off only. I can make the forward most shift solenoid turn on and off by setting various 0A2 bits. I beleive the other solenoid can also be manipulated in this way. Only the PW of the PCS closest to the rear of the tranmssion could be changed. The others were ~static at 25us or ~140us no matter how I manipulate the bits in the message. THere are four messages that could affect change... 0A1, 0A2, 0A3, 0B4

What gears should I use?
I beleive 1st and 3rd gear are the most like candidates providing dual motor output capability.

Software
I'll need to create some custom Zombiverter Software to shift gears.

[P.S.Mangelsdorf]

Sounds cool! Check out the articles I linked in the 2ML70 thread - there's some info there about the gears/shift logic that might be helpful.

I'm considering building a shop truck with the 2ML70 this spring so we might be able to work together on the development. I've got to clean up the wiring diagrams I have so that they'll be useful and not piss GM's lawyers off when I post them (they're service manual screenshots).

Also in terms of power, I know from those transmission repair articles the 2ML70 uses only electric in reverse. My father once used our family's hybrid Yukon to tow a stuck rollback flatbed truck out of deep mud and up a hill in our yard (long story), and it was in reverse. So anecdotally anyways, there's PLENTY of power on tap.

[Eagle411]

Sounds good! I'll keep posting information as I move forward and I'll take whatever help I can get.

[P.S.Mangelsdorf]

Sounds good! I'll keep posting information as I move forward and I'll take whatever help I can get.

I'm hoping to have time to work on the wiring diagrams this weekend. Been a hectic few weeks at work and personally.

[Bratitude]

So your going to get a oi replacement board for the Prius inverter and use zombie as your VCU?

because you can’t just hook up any motor to Prius inverter if it’s controlled by zombie

[Eagle411]

So your going to get a oi replacement board for the Prius inverter and use zombie as your VCU?

because you can’t just hook up any motor to Prius inverter if it’s controlled by zombie

Why not? Is there some motor to inverter communication that's missing?

[johu]

The prius inverter expects a motor/resolver pole pair ratio of 2. That is commonly found on Lexus and Toyota motors but not on most others. With a different ratio the motor will not run at all.

In addition the inverter is probably optimised for the induction values of the Prius motor. So even if the pole pairs match it will have a few percent less efficiency.

[Eagle411]

Thanks. You saved me from a lot of future frustration. I beleive the ratio is 1 to 1. I see a couple of options:
1. Use a Gen 3 Prius Controller Board
2. Hack the Tahoe's Inverter, such that I can control it.

#1
Cons:Costs ~$350
Pros: Known setup and control process. Still need to VCU, which I already purchased and received, to control the transmission. Potential reduction in time spent vs #2
#2
Cons: Mostly unknown interfaces and control functionality. There might be predefined limits on power output and control.
Pros: 0 initial cost. The Tech2win software provides a readout of all motor control parameters and some transmission control overrides.

I may do both. More to think about. Thanks again!

[P.S.Mangelsdorf]

The third option would be to get a Volt inverter, and have some place like JLC PCB build one of Damien's controllers for it.
https://github.com/damienmaguire/Ampera-Volt-Inverter

And the 2b option would be to try to create a control board for the Tahoe inverter. I am wondering how close it is to the Volt inverter. There's a junkyard near me with a couple Volts, and one further away that has apparently had a Tahoe hybrid in the yard for a few months. I may venture out to see if the good parts are still there, but likely not for another week or two.

[Eagle411]

Yes! The Volt Inverter was my original plan, but I switched. The toyota inverters are very common in the junk yards around here, so I'm a little disappointed. Its not a huge deal, because I have all three inverters (Tahoe, Volt, and Prius) sitting around here.

The Tahoe and Volt Inverters are made by Hitachi. I was hoping that I could google some interface documentation on these inverters. The volt inverter is on Hitachi's web page. The tahoe and volt boards are very similar including the connectors. I beleive they're the same, but white instead of black. Using the new board with the Tahoe inverter would probably work with different connector locations and pinouts.

Regarding the volt boards... I actually priced having the Damien board mostly build by JLCPCB last year. I had to have 5 made and they came to about $400. I figured that I could make some of it back by selling the boards. Some of the parts are now obsolete, so I had to find some subs. I might try doing it again.

Here are some pics:

[P.S.Mangelsdorf]

Ok I started a github repo for the 2ML70 project.
https://github.com/PSMangelsdorf/2ML70
Only thing there right now is an excel file with pinouts, and that only contains the main inverter connector pinouts. I'll try to add more this week.

Can you take a couple photos of the Volt and Tahoe inverter guts next to each other? I'm interested to see how similar/different they are.

[Eagle411]

Here is a better photo of the Tahoe board. I would prefer not to open the volt inverter right now, but those are already on the internet.
Here are some of the parts

HITACHI board GD219561

MPC565MVR56
1MByte Flash
36Kbytes of Static RAM

ST M58BW0160
16MB Flash memory
25LC640:64K SPI Bus Serial EEPROM
CY7C1021CV26

IC SRAM 1MBIT PARALLEL 44SOJ

ADW71205YSTZ-ND
Complete monolithic resolver-to-digital converter (RDC)

TI 2444AQ: Advanced LinCMOS™ RAIL-TO-RAIL OUTPUT WIDE-INPUT-VOLTAGE OPERATIONAL AMPLIFIERS

ACPL-M46T:Automotive Intelligent Power Module with R2 Coupler™ Isolation and Small Outline

Altera ACEX EP1K10TI100-2N – 1K FPGA (56000 gates, 12288 bit memory
576 LE
66I/O
TQFP-100
72 Logic Array Blocks

SN74LVC245A Octal Bus Transceiver With 3-State Outputs

MC74HCT541A
Octal 3-State Non-Inverting Buffer/Line Driver/ Line Receiver With LSTTL-Compatible Inputs
LC245AQ:SN74LVC245A Octal Bus Transceiver With 3-State Outputs
APIC-S03 is a specific type of photoMOS relay, which is used in various circuits where reliable and fast switching of signals is required. Here are some common applications and circuits that might use the APIC-S03

[madude]

Great project! following for progress. I have always been interested in doing something like this but didn't have the technical knowledge to tackle.

[HerbP]

Pardon the intrusion. I'm in the 'read everything you can' mode of my project and someone suggested looking at the 2ML70 due to local availability (as opposed to LS600H that I initially wanted to use). I found this thread but before I did, I stumbled on this: https://www.transmissiondigest.com/gms- ... mission-2/ ... which contains some useful (to me) information including pinouts. Maybe you guys already knew about it.

[P.S.Mangelsdorf]

I found this thread but before I did, I stumbled on this: https://www.transmissiondigest.com/gms- ... mission-2/ ... which contains some useful (to me) information including pinouts. Maybe you guys already knew about it.

Yep, that's linked over on the development thread: viewtopic.php?t=5833

[HerbP]

Yep, that's linked over on the development thread: viewtopic.php?t=5833

heh oops. Sorry, I hadn't seen that thread yet.

[P.S.Mangelsdorf]

heh oops. Sorry, I hadn't seen that thread yet.

No problem, welcome to the forum! If there's more of us thinking about using the 2ML70, it should make things easier.

[HerbP]

No problem, welcome to the forum! If there's more of us thinking about using the 2ML70, it should make things easier.

My project is still in the sheet metal phase. Cutting out rust and rebuilding panels, gathering chassis parts, etc. I'll follow along for now.

[Eagle411]

A little bit of information.
The HVH has 5 pole pairs and it apears that it may have mutliple resolver poles as well. Still investigating.
THe HVH250 uses delphi connectors for the resolver connections. This isn't typically exposed outside the transmission, but I have a motor sitting on the bench. 12064769 female 1-way Metri-Pack 150 connector is used. There are 2 variants with different keying between MG1 and MG2.

[Eagle411]

I was able to get the thing spinning on the bench, but I've got an issue with the input shaft spinning. I need a way to stop the input shaft from spinning and to seperate it from the MG1 motor. It apears that they are locked together by a sun gear. I have an idea, but I feel like others might know better. I'll probably ask this other places.

[P.S.Mangelsdorf]

I guess I'm missing why you want to separate them. Its been a while since I looked at the diagrams but I don't recall seeing a need to separate them like that.

[Eagle411]

I'll have to make another video because I don't think people really understand what I'm asking. There will be a torque dampener attached to the input shaft. The dampener will be secured in place so it can't rotate. If it's always attached to MG1 then I can't use MG1 at all. I need to put the input shaft back on without the sun gear, so they are no longer attached.

Adding some helpful links for my volt battery packs and BMS

https://github.com/Tom-evnut/AmperaBattery

[P.S.Mangelsdorf]

Why keep the torque damper? My understanding is that there's no fluid flow to/from the transmission, so why not just remove it and put a blockoff plate over the bellhousing?

[cajunfj40]

I've been poking around at this one a while, but all files, no actual parts, so this is just internet info dumps. There's some videos on youtube of disassembly. Apparently there's a particular clutch that goes bad on them, as well as the usual automatic transmission "gritty fluid messes up solenoids" issues. There's some big O-rings in them that seal the motor housings to the transmission housing that apparently swell a LOT, so when they come out they need to be replaced - and you'll need to figure out what size they are because I've not found anything about it, nor a part number. Transmission supply places list some of the components, but not the motors or o-rings.

From the layout in the patent (US 6953409), it looks like you'd lock clutches 1 and 4 for 1st gear (3.76:1) dual motor output, and lock clutches 2 and 4 for 3rd gear (1:1) dual motor output. 2nd gear is one motor at 1.76:1, 4th gear is the other motor at 0.74:1. Normally those "fixed gear" modes in the Tahoe don't use the motors at all, but they are splined to the shafts anyway and basically locked together 1:1 so I thought that's how I'd use one if I had the time/space/tools/budget to finagle it. Input shaft not to be locked, just let spin in the air, or cut it short and cover over the stub. https://patents.google.com/patent/US6953409B2/ It's more complex than the GS450H/LS600H, but standard GM bolt patterns so lots of adapters available for the output. I've attached a pdf of one of the articles linked in the other thread, but with the addition of a few more pictures and a nicely done set of drawings showing the clutch/gear/motor layout. The patent has a truth table for the clutches that appears to match the articles and the clutch drawings in the attachment.