Lexus GS450H VCU Support Thread

[Bryson]

Looks good, but for me, the Lexus based dc_bus_voltage is unreliable. It reads around 90v most of the time for me now, even though I run a 370v nominal pack. Occasionally it did read the correct voltage near 400v, but typically not. The ISA reads right for me so you might take that input instead going forward if you find the same symptoms.

[Bryson]

Also, you might want to use throttle pot 2 for your regen input. Just a thought

[arturk]

... the Lexus based dc_bus_voltage is unreliable. It reads around 90v most of the time for me now, even though I run a 370v nominal pack. Occasionally it did read the correct voltage near 400v, but typically not....

Works the same way for me. Actually, when inverter is powered on, it reads correct value. When motors exceed 3000 rpms it drops to around 90V (every time). Perhaps there is a bug in our protocol interpretation. To my knowledge most if not all users experience this issue.

[Bryson]

Have you been able to get current or power to show in your interface? It looks like I might need to convert current and power to integers within the olimex code. Curious if you’re using the ISA shunt or not. Thanks Artur

[arturk]

Have you been able to get current or power to show in your interface? It looks like I might need to convert current and power to integers within the olimex code. Curious if you’re using the ISA shunt or not. Thanks Artur

I got everything showing correctly but I remember tweaking code little bit on VCU side, nothing major though. On web interface part I only change styling slightly and ranges for certain gauges.
Keeping in mind web interface is parsing values from CSV format, perhaps you have something misaligned.
Yes, I am using ISA shunt.

[Bryson]

Would you mind posting code for both VCU and WiFi interface in their entirety? Or send to me and I can post to GitHub. At this point I don’t have anything to compare to that’s known working

[arturk]

No problem, here it is.

[Bryson]

Thanks! Looks like it should be working in my car based on what yours is outputting. Time for more debugging I guess, neither CAN0 nor CAN1 report correct current, but both report correct voltage.

[arturk]

Thanks! Looks like it should be working in my car based on what yours is outputting. Time for more debugging I guess, neither CAN0 nor CAN1 report correct current, but both report correct voltage.

Try to simulate some expected values within VCU first to verify WiFi interface shows correct values.
Also, do you have ISA initialized correctly to output all the parameters? You can also try ISA test/initialization sketch to verify if it is reporting correct values.

[Bryson]

I believe it’s initialized correctly based on all the examples I’ve seen and comparing my code to yours. I’ll try one of the example sketches next as another check, along with passing some test numbers through to the WiFi interface. Both great suggestions, thanks

[86elcamino]

post deleted

[86elcamino]

Sooo, came across this tonight and was ready to throw in the towel. Anyone know the best solution to this. I was going to clean things up and get ready to wire this out and looks like im missing SL2 + , SL1+ and SL1-. I gentle pried the connector up and there is almost zero play to get to the wires and I would need to figure which is which as it not in line with the pins. Anyone have suggestions?

0618212332~2.jpg
0618212332b.jpg

I had the exact same problem. It's an easy fix, just pull the pan, then the filter, and it's all right there. Supposedly I just bought the only one of those connectors in the states, about $45. The Ebay yard I bought the trans from even reimbursed me.

Will try to remember to get part # and post tomorrow.

[Gigas]

I'm sure i'm doing something wrong here but why do I get this everytime I try and upload the firmware?

[bexander]

Looks like you need to install the Metro library.

[Gigas]

Well, it's been so long since i've done programming I forgot about all the libraries. brain fart day. I added all the libraries and it fixed it

Looks like you need to install the Metro library.

[86elcamino]

Sooo, came across this tonight and was ready to throw in the towel. Anyone know the best solution to this. I was going to clean things up and get ready to wire this out and looks like im missing SL2 + , SL1+ and SL1-. I gentle pried the connector up and there is almost zero play to get to the wires and I would need to figure which is which as it not in line with the pins. Anyone have suggestions?

0618212332~2.jpg
0618212332b.jpg

I had the exact same problem. It's an easy fix, just pull the pan, then the filter, and it's all right there. Supposedly I just bought the only one of those connectors in the states, about $45. The Ebay yard I bought the trans from even reimbursed me.

Will try to remember to get part # and post tomorrow.

I started hooking individual wires from the vcu to this connector today, and realized that the new part I just installed has the same pins missing Someone may correct me if I'm wrong, but I think you are looking at the connector wrong(the diagram shows the wiring side), and the missing pins don't actually have anything that's supposed to go to them.

I did get the part number for the connector going into the trans case: 82125-50100

[Gigas]

You may be right - I think the blank SP pin is what threw me. If your right(Which I think you are) it should be this at the transmission

If someone can confirm this then we are both good! I may order the new connector anyway as mine is in pretty rough shape.

[PatrcioEV-ATX]

You may be right - I think the blank SP pin is what threw me. If your right(Which I think you are) it should be this at the transmission


0618212332_2.jpg

If someone can confirm this then we are both good! I may order the new connector anyway as mine is in pretty rough shape.

Correct, there is nothing on those three pins. I have an original harness and those are plugs on the wire side of the harness (and missing in the plug itself).

[Bryson]

Is the motor stator thermistor temp routine working for anyone? My MG1/2 stator temps always read 109.69 no matter the ambient condition

[Gigas]

Is the motor stator thermistor temp routine working for anyone? My MG1/2 stator temps always read 109.69 no matter the ambient condition

I was wondering that too - in Damien's video it says that temp too - I just uploaded the latest code and mine says the same temp.

[Bryson]

Do we know what type of thermistors are in there off hand before I start digging around for info?

[Bryson]

For those that are having IVT-S issues with the VCU: when someone asks you if you’ve initialized it properly, they’re referring to this process!



My problem with my sensor was solved by googling until I found out that I had more work to do before the sensor was ready to be useful. Oops.

[Bassmobile]

No problem, here it is.

Bryson, did you get these into a "Arturk" branch yet? If not, I will do it today.

I did get the part number for the connector going into the trans case: 82125-50100

Thanks for following up with the part number. It has been added to the GS450H wiki.

[Bryson]

Haven’t made a new branch. If you could, that would be awesome.

[Bryson]

So I've been digging into the stator temp situation - it looks like the voltage divider resistance is an order of magnitude too weak to deal with the thermistors in the transmission. They appear to be 33k ohm NTC thermistors, which means we'd want 20-40k resistors on the board to divide 5V down into a readable range for the input signal. The board has 1k8 resistors which are railing the input high. Next up is to order some suitable replacements and test from there.

It is so helpful to have schematics available! Open source rocks.

edit: it could also be a 47k thermistor. I've got to let the car settle for a few hours to see what it converges on - it was 20C outside when I measured 33kohm about an hour after driving the car

edit2: looks like I'm late to the party and this was discussed earlier in the thread, they're 47k thermistors. With a 1k8 resistor divider you get more fidelity in higher temp ranges, but the ADC can't measure anything under ~100 degrees C