Zombieverter developemental SW version testing on full Leaf setup

[J0hannes]

Maybe this is worth a separate topic and to minimize confusion what comes to official releases

Ok, so I have noticed this behavior with several developemental versions (can't say if this is true to released versions as current ones have their issues for which I can't drive them.)

Background:
I have "full" Leaf setup meaning full drive stack and an unmodified Leaf battery.
I updated to 2.33TH and I took a couple of CAN traces related to this, short one when I started the charge, a longer one with few captures and then I put this in runmode afterwards.

The issue:
The charge current command in 1F2 doesn't seem to follow LB_Charge_Power_Limit given in 1DC.
Neither does it seem to follow LB_max_power_for_charger

Looking at the trace, LBC is asking for 0kW, but OBC pushes what comes out of the EVSE

[tom91]

Correct, barely any of the BMS messages are ingested and used.

[jerrykco]

barely any of the BMS messages are ingested and used.
[/]
HI Tom, is there a list so us new to this can know what is covered and what is not consumed?

I.E. Is there a way in Params to specify these maximums?

[vitecd]

released versions as current ones have their issues for which I can't drive them

Can you share the issues, please?

[J0hannes]

Can you share the issues, please?

viewtopic.php?p=84499

The no-torque-reversing problem

[J0hannes]

Sooo

After changing some fresh contactors, I found myself facing a precharge fail -issue.
I synced my vehicle_testing with the main but did some tweaking to get precharging to work.
https://github.com/Romukeisari/Stm32-vc ... le_testing

To me it seems that we have a DCDC ON command in one of the VCU messages, which should be off status during precharge.

This results in a situation where the DCDC starts to draw power before precharge is complete. This is also visible when plotting

Any ideas, what to try

[tom91]

Do you know where the DCDC is turned on via CAN?

Possibly the main HV system relay bit should toggle. in 0x1D4

[tom91]

Added an update that should now toggle the HV status as per leaf logs. Pushed into the testing branch, so please try this and provide some feedback.
https://github.com/damienmaguire/Stm32- ... 811d0b12e9

[J0hannes]

Tomorrow, on Sunday I have a test and trace day. I really need this to get back on road as there's an driving event next weekend I want to attend to

I'll try the toggle, but then again I have this gut feel that there is a discrete command message for DCDC ON/OFF

Do you know where the DCDC is turned on via CAN?

Possibly the main HV system relay bit should toggle. in 0x1D4

I've been struggling to find a right CAN log from a leaf with startup sequence logged. Then I had troubles to read an EV CAN Log from a 2016 Leaf to check what the VCU sends and how the PDM/OBC responds.
I'm not sure where I found these two PCAP format logs, but they have the clues needed. Savvycan doesn't open this and Wireshark doesn't interpret the messages with CAN DBC, so this was a bit of manual work. I've been looking at this with goggles that something yet unidentified, but documented MUST contain the control command...

Here we go:

In x1F2 there is this unknown bit in byte 7

image.png (5.98 KiB) Viewed 1360 times

I stared the hex values in Wireshark and saw that x390 byte 3 has values B2, B3 or 13 when DCDC is on

image.png (5.84 KiB) Viewed 1360 times

This information was also "told" in this google sheets Dala has linked in his LeafLogs

So if x1F2 byte 7 starts with 8, unknown bit 7 is 1. If it starts with 0, bit 7 is 0. -->
Cue filtered log after learning how to use Wireshark field modifiers
Looks like there is approx 400ms delay/buffer before DCDC turns on after the start of byte 7 in x1F2 changes from 0 to 8

So this is also what I want to test out tomorrow

[J0hannes]

So I did some testing on the weekend. Params attached, though I made my settings default in paramprj in my branch to reduce fudge factor of forgetting the param save to flash.

A side note, LEDs indicating which contactors get pulled is a really good debuging feature


Can't identify which version on, but there seems to be a bug in the precharge exit condition. This first test was done without HV, but shows that precharge gets skipped. This is probably what had killed my contactors.

030825_233TH_Prech_skip.csv

(630.33 KiB) Downloaded 20 times

I modified and then got functioning precharge

030825_233TH_Prech_fail.csv

(863.19 KiB) Downloaded 18 times

Several iterations and various tweaking later, I still couldn't get the suspect DCDC ON command to do anything else but get the DCDC in failsafe mode?

030825_2343JN_dcdc_cmd_working_With_HV_INV_volts_issue.csv

(1.69 MiB) Downloaded 17 times

Then went back some steps and started to look at precharge timing as it seems to be way too slow to pull the contactors before the DCDC kicks in.
Reduced precharge min time and the rlydly so the contactor pulling sequence (precharge) is complete in less than 1200ms which seems to be some magic time after which the DCDC starts chooching and inverter voltage sags if HV main+ is not pulled in time.

030825_234.5JN_Prech_success_UDC_OK.csv

(826.47 KiB) Downloaded 23 times

030825_234.5JN_Prech_success_again_also_chg.csv

(851.91 KiB) Downloaded 18 times