Lexus GS450H VCU Support Thread

[Jack Bauer]

Folks has anyone implemented shifting the high / low gear in the Gs450H gearbox? Would like to implement this in the new vcu project.

[xp677]

I think I've got a working solution but have not tested it in real life yet.

The project code gets a bit more complicated since the system as a whole needs to know what gear you're in. The whole shift sequence has to go something like:

RPM gets to the threshold value (with hysteresis for up/down).
Ensure line pressure is high enough for the shift (from either PB1, 2, or 3)
Set torque request (in HTM data) to zero
Wait a predetermined time (a few ms) to allow the data packet to propagate.
Maybe verify torque response from the MTH data.
Close (or open, depending on shift) both SL1 and SL2
Look for response from PB1, 2, 3 (I forget which two)
Reapply torque gradually over time (a few 10ms)
Record appropriate gear (for speed readout adjustment). I posted the ratio a while ago in here, it's 1.9 vs 3.9.

All the while, SP should be modulating line pressure, with one of the PBs as feedback.

Whether you use MG1 or MG2 for this is up to you. IIRC MG1 reaches its max RPM first, 11000 I think, so maybe set a shift up point at 9000-10000 and a shift down point a thousand or two below the equivalent high gear ratio for shift down.

The transmission defaults to the highest gear (SL1 and 2 not energised).

[Jack Bauer]

Brilliant thanks:)

[Jack Bauer]

Did anyone work out the curves for mg1 and mg2 temp sensors in the gearbox?

[xp677]

Maybe here?

viewtopic.php?f=14&t=271&start=25

[TonyV]

Hi everyone,
started bench test wiring for the 450H transaxle.
I have a question:
from the gas pedal, wire #3 and #5 go to GND pin #20 on the VCU,
Is pin #20 the same as pin #5 GND?
Thanks

[jcwright45]

I’ll not had any issues bench testing when supplying the throttle pedal ground via VCU versus another ground. I’ll assume there is a specific reason for the VCU supplying ground versus it being feed elsewhere

[celeron55]

You want to avoid there being any unwanted current flowing in the pedal ground or 5V wires, as those currents would fluctuate the pedal output voltage. Thus either use a separate ground and 5V pins or split those wires right where they exit the VCU.

[sfk]

I think I've got a working solution but have not tested it in real life yet.

The project code gets a bit more complicated since the system as a whole needs to know what gear you're in. The whole shift sequence has to go something like:

RPM gets to the threshold value (with hysteresis for up/down).
Ensure line pressure is high enough for the shift (from either PB1, 2, or 3)
Set torque request (in HTM data) to zero
Wait a predetermined time (a few ms) to allow the data packet to propagate.
Maybe verify torque response from the MTH data.
Close (or open, depending on shift) both SL1 and SL2
Look for response from PB1, 2, 3 (I forget which two)
Reapply torque gradually over time (a few 10ms)
Record appropriate gear (for speed readout adjustment). I posted the ratio a while ago in here, it's 1.9 vs 3.9.

All the while, SP should be modulating line pressure, with one of the PBs as feedback.

Whether you use MG1 or MG2 for this is up to you. IIRC MG1 reaches its max RPM first, 11000 I think, so maybe set a shift up point at 9000-10000 and a shift down point a thousand or two below the equivalent high gear ratio for shift down.

The transmission defaults to the highest gear (SL1 and 2 not energised).

Here's a logic diagram of the solenoid states

[Jack Bauer]

Many thanks sfk. This is now in the new vcu code.

Folks it's been brought to my attention that there is an issue with the github files for the esp module on the sam3 based vcu. I would like to encourage those who have solved this problem to please contribute the solution.

[nkiernan]

JB, I have a post here that explains what we found and how we corrected the ESP programming issue.

viewtopic.php?f=5&p=21543#p21543

Only needs the "solid-gauge.js" file copied from the ESP repo data folder to the GS450H wifi repo data folder. Had a look to see if I could do that but no access.

[Jack Bauer]

Happy to add you as a colaborator on the repo. What is your github name?

[nkiernan]

nbk1202

[Jack Bauer]

You're all set:)

[nkiernan]

Gave it a go JB, no uploads allowed message popped up, looks like some restrictions. Was trying to upload the gauges java file to the data folder.

[Jackk]

Hello everybody!

I am new to the forum and a gigantic noob so go easy! I have reached a stumbling block with my GS450H vcu and I am crying out for help!!

I have a bench test set up following the schematic v2.4 composed my rrobertjanzen... which is brilliant I might add.

I built Damien's v2 kit and uploaded the gs450h_v7.ino firmware and all the lights are flashing as they should.
I have checked my wiring to the diagram many times and I'm certain its correct.

when I power up VCU the positive contactor closes for maybe 5 seconds then opens... why is this happening it should stay open while the pre charge circuit does its thing... I have changed the pre charge value to 12 in the code also as i'm running at 12v for test.
the pre-charge and negative contactor should close first but they do nothing.

the inverter relay coil is energized for maybe a second before it cuts out so no 12v is getting to the inverter....

It seems like my wiring must be off but I have followed the v2.4 wiring closely... what am I missing? am I not running the most up to date code?

Many thanks in advance!!

Jack

[Jack Bauer]

Jack do you have an isa shunt connected? This is how the vcu reads voltage, current , power and stuff like that.

[TonyV]

Hey Jackk,
I posted this exact question 20 minutes ago on the GS450H Discussion thread.
viewtopic.php?f=14&t=644&start=150
Glad to see I'm not the only one with issues.

I do not have a shunt connected, I would of thought it was only used for monitoring and not part of the operation of the inverter.

Will any shunt with can work?

I remember there was a group buy awhile ago, where did everyone buy thier's?

Thanks

[nkiernan]

JackK and TonyV, just checking, if using V7 of the GS450 code, are you doing anything to accommodate the CAN messages that code is expecting to receive for the BMW E65 it was intended to work with?

Just recently got my GS450H setups running on the bench. Started with V3 and some light mods to get the motors turning and prove the inverter/transmission was ok before worrying about the oil pump or HVJB.

Moved onto the V7 code but it needs to see some CAN from a BMW E65 to fire up and get gears selected etc. Maybe you've addressed that. I Just put together a basic arduino sketch for an UNO with a Seeed Studio CAN shield to mimic these, ignition on, gears.

[nkiernan]

Put up a quick video, might help. Switch for T15 or 'on', pot for rotary gear select - P-R-N-D.

[TonyV]

are you doing anything to accommodate the CAN messages that code is expecting to receive for the BMW E65 it was intended to work with?

That is news to me...

I purchased my VCU from Damien about a month ago, thinking I wouldn't need to deal with programming at all.
But I did have to do your fix to get the MG2 RPM back to zero from constantly reading 9000. (That was a nightmare trying to figure that out)
I personally don't know how to check the version that is on the VCU but I would assume it would of been the latest version.

would you be willing to share your code.
I would be happy just to see the thing turn.

here is my disaster

[nkiernan]

Absolutely, no issue sharing what I have and trying to help. I know there are a few of us at the early stages with the GS450H kit. It'll be tomorrow before I get to the computer, so hope that's ok.

[TonyV]

Sure its OK!
been dick'n with this for weeks what's another couple days.

[Jackk]

I do not have the ISA shunt connected and this will be the one of the first things I will do!!

I too would also be very grateful to nkierman for that code and thankyou so much for your help!.

Its nice to know I'm not the only one with this issue and I'm not a total moron!

[Jack Bauer]

Folks , thanks for posting about these issues. I'm upset to see the ends some of you are going to get this vcu up and running. It is true that the v7 code is specifically tailored to run in my E65 project but the earlier versions on github are much more generic and way more suited to getting things spinning. So before putting in any more hard work please give me a few days to do the following :
1)Adapt one of the earlier code versions to work with simple direction inputs (12v) and an analog throttle.
2)Make a video showing how to set this up from scratch on the bench.

Sound good?