Understanding ZV VCU Pre-Charge contactor operation at restart

[nkiernan]

I want to understand what could appear to be quirky pre-charge contactor operation with the Zombieverter VCU. Video links below show what's happening and this is with a 50V test battery only (so may be part of the reason). I know the pre-charge contactor comes on before the positive contactor and stays on during vehicle operation. But it appears there is a scenario when the pre-charge contactor stays off during operation, and my concern is it may be happening even if HV is below the pre-charge set point?


Key thing to watch in videos below is the LED indicators out in the engine bay just over the top of the steering wheel in the video. There are three LEDs connected to the auxiliary circuits of the negative (red LED), pre-charge (orange LED), and positive (green LED) contactors.


Video below shows normal expected error state with no HV connected, so pre-charge voltage set point can't be reached. Negative and pre-charge come on, after a period where voltage doesn't rise, pre-charge contactor switches off and I turn ignition off




Now with HV connected (50V DC test voltage), negative and pre-charge come on, and once pre-charge set point is reached, positive contactor closes and we're in run mode. Move into drive and wheels spin. Then back to Park and switch off ignition and remove key




Here I insert the key and restart the car within one minute of the previous 'drive'. Watching the LEDs now, the red (negative) and green (positive) come on pretty instantaneously, but no orange (pre-charge)! Now this seems to happen whether HV has dropped down below the pre-charge set point or not. But, as the test voltage is so low here, this may be relevant? Car can go into drive and spin wheels, but pre-charge contactor remains open.

Either way, should the pre-charge contactor come back on as per normal operation?

[tom91]

Precharge should be happening.

Can you generate a plot of your UDCs?

Will review logic to see how this could be possible. Please confirm your software version number.

[nkiernan]

I got a little time to do some logging this evening. I am running a GS450H setup and observe the same behavior (only negative and positive contactors come on after a quick restart) with both the current V2.30A firmware and a modified version of this in which I added a Jaguar Land Rover vehicle class to control my dash which was derived from the E39 vehicle class.


Plots shown in the PDF's below. One observation is that when the ignition is switched off between the first normal run and the restart, UDC and InvUDC remain constant and don't taper off (fair enough, the vcu isn't logging when ignition off). I remember discussions previously about possibly forcing these to zero on restarts, but maybe not relevant here anyway. To get back to normal operation, I have to power cycle the ZV VCU, and the next start up behaves. Without a power cycle, that same neg/pos only contactor behaviour remains. I need to check this after a longer ignition off cycle to ensure DC voltage does drop below pre-charge set point as another test.


I'll spend some time thinking through what possible errors I might have introduced into my setup to cause this.

Edit: The dips shown in UDC and InvUDC after DIN_Start is wheels spinning in drive mode, so vcu goes into Run mode each time

params_V230.json

(2.12 KiB) Downloaded 3 times

NormalRun_Then_Restart_V230.pdf

(213.28 KiB) Downloaded 4 times

params_V231_JLR.json

(2.12 KiB) Downloaded 3 times

NormalRun_Then_Restart_V231_JLR.pdf

(205.78 KiB) Downloaded 3 times

[tom91]

Where does you UDC come from, an ISA shunt right. I would suspect it be a true reading. The ISA shunt is always on (if wired to be so) and zombie always reading it.

The its possible the code just by passes pre charge because of the UDC being over the UDCsw

[nkiernan]

Hi Tom,
Yes, using an ISA shunt for UDC (powered on at ignition on). I'll do some more tests restarting both above and below UDCsw for more data points.

[tom91]

It could be indeed that the way you currently have it setup it holds the last voltage reading from the ISA and does not catch a new one in time. Should look at having ISA shunt readings go "Stale".

The software was written with the idea of having the ISA shunt always powered, not the most efficient way from a current draw point of view.

I may find time to look at refactoring the whole handling of VCU states and HV Bus states to create more read-able code.

[nkiernan]

Ok, I can easily swap the shunt power setup for a comparison test, hopefully later this evening.

[nkiernan]

Few more tests this evening. UDCsw is 30V for each test, pack voltage is 50V. GS450H setup. Multimeter is showing inverter input voltage and LED's show which contactors close (red = negative, orange = precharge, green = positive. Attaching a video and PDF plot for each test below.


Test 1

ISA shunt powering up with ignition on. This test does a normal first time start up and run, then a restart with InvUDC above 30V/UDCsw, and then a restart with InvUDC below 30V/UDCsw.

Results: First run behaves as expected. Neg and Positive contactors only on restarts for both above and below UDCsw, but car drives normal

030326_Test1_V230.pdf

(206.11 KiB) Downloaded 3 times

Test 2

ISA shunt powering up with ignition on. This test does a normal first time start up and run, then a restart with InvUDC below 30V/UDCsw. VCU power cycled after previous test.

Results: First run behaves as expected. Neg and Positive contactors only on restart below UDCsw, but car drives normal

030326_Test2_V230.pdf

(191.91 KiB) Downloaded 3 times

Test 3

For this test, the ISA shunt power was changed to permanent 12V. This test does a normal first time start up and run, then a restart with InvUDC above 30V/UDCsw. VCU power cycled after previous test.

Results: This test had a quirky result where at the restart, no contactors closed. Graph shows T15 and start signals but VCU doens't respond. Required a VCU power cycle and things were back to normal.

030326_Test3_V230_ISA_Perm.pdf

(219.1 KiB) Downloaded 4 times

Test 4

For this test, the ISA shunt power was changed to permanent 12V. This test does a normal first time start up and run, then a restart with InvUDC below 30V/UDCsw. VCU power cycled after previous test.

Results: This test gave expected results for first run and then restart where all contactors came on in correct sequence in both cases, once you don't try an restart before UDC drops below UDCsw.

030326_Test4_V230_ISA_Perm.pdf

(203.72 KiB) Downloaded 3 times

Thoughts:

• Review my setup in case issue is self inflicted!

• ISA shunt operation affected by power choice, permanent 12V vs ignition on. Benefit in voltage values being cleared if no live readings?

• In all tests, restarting before UDC drops below UDCsw seems to affect the operation of the precharge contactor. In theory may not be an issue for normal use cases

• Would the VCU recover itself after a time and not require power cycling to get it back to normal operation after a restart that left the precharge contactor off for a run?

[tom91]

Key take away. ISA shunt ignition powered means the old bus voltage hangs.

Will add a force to zero when zombie goes off as minimum change.

[nkiernan]

Thanks Tom

[nkiernan]

Note now added to the Zombieverter wiki regarding permanent vs ignition on power for the ISA shunt.

[nkiernan]

Just to close this out, I ran a couple more tests to be clear on how the pre-charge resistor operates on the ZV.

When restarting after sufficient time to let InvUDC drop below UDCsw, pre-charge brings up InvUDC to UDCsw by closing the pre-charge contactor before closing HV+ contactor. In this case, the pre-charge contactor remains closed during operation as per normal operation.

When restarting before InvUDC has time to drop below UDCsw, there is no requirement for pre-charge and the pre-charge contactor is not closed. HV+ contactor closes without pre-charge contactor closing. In this case, the pre-charge contactor remains open during operation.