GS450h Transmission and Inverter Loss of drive

[dadiowe]

Since my last issues I have just completed 100 faultless miles in my car. Unfortunately, I was doing some testing with different transmission settings i.e. Auto, High, Low, MG2 only and disaster has struck.

The problem,
I stopped at some traffic lights and suddenly the gearbox starting to apply torque then drop torque very quickly and loudly. Selecting neutral stopped the harsh oscillation. When the traffic lights changed I selected drive with a bit of throttle and could pull away with no shaking. Every time I stopped the whole process would start again. Pulling over I checked my parameter list. I had set regenramp to 1 and throtramp to 10 earlier and enabled regen to 2. I then re-set these parameters to 0.1, 0.1 and 0 respectively and tried the car. There was absolutely no throttle response. I think the parameter issue is a red herring because after a number of attempts of uploading a blank firmware bin and then trying version 2.05, resetting and retrying nothing changed. Interestingly I then applied the hammer and wriggle wire technique under the bonnet and the car then worked. I drove a mile or so stopped and restarted the car and manged get home thinking it was a loose wire somewhere and I would check everything out. The next morning and since nothing I have tried has worked.

The set up,
GS450h transmission and inverter
Zombieverter with any 2.05v to latest software version
F30 Gear lever
Outlander charger
BMW Sbox
At the time of the fault the Gear shift loom was not connected to the Gbox
Brake pedal input is not connected.
BMW e85 instrument panel is showing as it was before the problem.

My readings,
Spot Values (auto ticked and refresh applied)
When the car has ignition and crank applied the values are UDC 340V, UDC2 340V, RUN, 14.5v, Invst on, ign on, gear selected i.e. Forward or Reverse
Throttle readings show throttle pedal movement and spot value shows Throttle pressed.
With only battery power UDC fluctuates with low HV and UDC2 also fluctuates with up to 440v not sure why. This is possibly due to the Inverter slowly discharging.

Thinking of a wiring issue I have checked the wiring to MG1 & 2 G-box sensors resistances from the pin side of the inverter plug as below,
Pin 1 GRF to pin 12 GRFG 9 ohms
Pin2 GCSG to pin13 GCS 14.4 Ohms
pin 3 GSNG to Pin 14 GSN 14.4 ohms
Pin22 MRFG to Pin 33 MRF 9 Ohms
Pin 23 MSN to Pin 35 MSNG 19 Ohms
Pin 24 MCS to Pin 34 MCSG 14 Ohms
I have also checked the resistances from the pin side of Zombie VCU to the inverter as below,
Pin 16 REQ- to Pin 17 REQ+ 124 Ohms
Pin 18 CLK- to Pin 19 CLK+ 124 Ohms
Pin 20 MTH- to Pin 21 MTH+ 124 Ohms
Pin 22 HTM- to Pin 23 HTM + 124 Ohms

I have a spare G-box and inverter and the readings above are very close on those units as well.
I have also swapped inverters.
I have examined the Zombie and noticed a burn mark on IC10 (MCP25625T-E_ML). As I had an old zombie board I have removed IC10 from it and fitted it to my existing board but I am unsure as to the condition of the original IC10 chip or if my install is ok although it looks good.

To me it seems that the Zombie is trying to send out the signals to drive the inverter but they are not getting there.

Questions,
Could a failure of IC10 cause this problem?

IC9,13,15,18 appear to drive the CLK-, REQ-, TX-, RX- outputs is there anyway to check these outputs? I can get hold of an oscilloscope but I do not know much about using one!!!

Is it possible that Zombie has got a bit confused with all the uploading of firmware versions somehow! if so, how do I get back to a base setting. i.e. no parameters set? I have tried the erase flash function to no avail!
I understand that writing a firmware bin file does not seem to change the parameters. I have powered down the VCU after every firmware upgrade and changed from GS450h to none and back as per the 2.17 update suggestions.

Any other possibilities?

Is there any one near-ish to Andover that has a test set up or car that I could drive to and test my Zombie on it?
WTF have I done now?

So near yet so far!!!!

Thanks in advance for any thoughts.

[tom91]

Zombieverter with any 2.05v to latest software version

Please confirm the EXACT firmware version

Regen is 2.15.A introduction.

Also provide your setting .json

[dadiowe]

Hi Tom,

Thanks for the reply.

I have tried versions 2.05a, 2.15a and 2.17a since the problem occurred all of which had worked, all be it with various parameter changes to suit the version and no regen enabled.
I was on 2.17a when the problem occurred as it was my first test run with regen. Now any version of the software does not work. I have included the json file as it is now. The only settings that are different from when the issue started were, regenramp to 1 and throtramp to 10 and regen to 2. The interesting thing was that suddenly everything started to work correctly and I got the car home. But the next day the car would not move.

I hope that helps.

[tom91]

Okay so the issue clearly is not software or settings then.

Do you get any reported value from the inverter like InvUDC?

The issue sounds like failing of inverter comms, there is a push request to have the speed set to 0.

As regen uses speed to determine off pedal regen, this can cause the behaviour you mention flakey speed reporting from the inverter.

[dadiowe]

I will check tomorrow but my concern is that if I install an earlier firmware bin i.e. 2.05a and use parameters that do not worry about regen why does the car not respond to throttle input?
I agree that it seems to be a failure of comms to the inverter because all the spot values indicate that the VCU is happy. It seems like the VCU thinks it is sending the right messages but they don't get through. Hence my question as to how to test the Clk, Req and the other outputs to the inverter.
I am unsure as to regen causing the issue or if a component failure has caused the problem and if so which one.

[robshephard]

Hi INVudc is 345v. I have checked the resistances on the output pins (6,7) of the 4 Ic's mentioned. All are at 60 ohms & 120 ohms with the socket unplugged from the inverter. They all have 3.3v on pin 2 and earths on pin 3.

It seems like the vcu is inhibited some how.

Is there any way to check the inputs from the STM 32 to the ic's.
Thanks again

[tom91]

If you are getting a spot value back from the inverter I would not doubt the comms just yet.

Please post a screenshot of all spot values when you are expecting the motor to turn.

[dadiowe]

Here we go.

[tom91]

You have 0 torque requested, thats your problem.

why is your pedal being reported as pressed?

There is a potnom calculated and reverse selected but no torque filled in so the zombie is not happy.

your vehicle speed reporting a negative value, no idea how this happens. Please have throttramp at 1 and set regen max to 0 for now.

When you back date to the older V2.05 you are on about does the same thing happen?

[tom91]

Increase your throtramp to 1 or higher to get drive back.

I checked code and with Damien yes we did encounter this too (mentioned in a video even), but the decimal places are for fine tuning. Will push a change to force minimum value of 1

[dadiowe]

My foot was on the throttle to show throttle position was working.

Vehicle speed always shows this when I am connected to the VCU but I am unsure as to why.

Using v2.05, or 2.17 has had the same effect.

I have just set the parameters as you suggested and the car is working again. I will try version 2.17 tomorrow, with regen turned off and let you know how I get on.

I have not watched the most recent videos it seems like I need to catch up.

I am also happy to test any settings you might need to try in the future.

Thank you so much for taking the time to help out I really appreciate your efforts.

[uhi22]

Maybe it would be helpful to create a time plot of the interesting spot values, to understand what is happening first and what next. For example to see the throttle increasing during slowly pushing it, and observe the torque and speed during this.

[joeflickers]

My foot was on the throttle to show throttle position was working.

Vehicle speed always shows this when I am connected to the VCU but I am unsure as to why.

Using v2.05, or 2.17 has had the same effect.

I have just set the parameters as you suggested and the car is working again. I will try version 2.17 tomorrow, with regen turned off and let you know how I get on.

I have not watched the most recent videos it seems like I need to catch up.

I am also happy to test any settings you might need to try in the future.

Thank you so much for taking the time to help out I really appreciate your efforts.

Did you resolve this?

[dadiowe]

Hi Joe. I think the problem with regen was that my brake input signal was incorrect so torque was not limited at a standstill.
Also if the throttramp is set to less than 1 i.e. the minimum recommended setting of 0.09 then you have no drive.
I have not as jet fixed the brake input signal and tested regen.
Hope that helps.