openinverter forum

The logs provided by Solition1DC and P.S.Mangelsdorf contain no PCS messages so you have a hardware / wiring problem. Please see my description of setup here and my log for comparison : viewtopic.php?p=44146#p44146

Jack Bauer wrote: ↑Tue Aug 23, 2022 7:51 am The logs provided by Solition1DC and P.S.Mangelsdorf contain no PCS messages so you have a hardware / wiring problem.

Is there any indicator in the wifi portal that will confidently confirm the PCS is powered up if the HVDC isn't connected? The only value I see changing on mine during first powerup is the 12VDC voltage, which I can manipulate using my battery charger, but I don't know if this signal is measured by the PCS controller board or published by the PCS over CAN.

Easiest way is to load up the software I posted a few days ago. You should then see the error bits update as the pcs boots. If they remain at zero the pcs is either not powered or not communicating.

I tried to update the inverter controller firmware using the web portal, and it appears I temporarily bricked my controller by not updating the ESP-32 firmware first:/ upon refreshing the web portal after uploading the new firmware, the parameters section is empty and the activity light on the controller pcb isn't flashing anymore...

I found this thread in the general software section of the forum (viewtopic.php?t=1756) that describes a problem the ESP-32 software used to have with debug messages being sent between inverter commands= causing bad software updates... I am fairly confident this is my problem, as my board was ordered and delivered in mid 2021 before this problem was discovered.

I am following the process described in this thread to confirm this is in fact the problem, and I have gotten to the point where I am listening on the serial port between the ESP-32 and the main controller micro, and I am seeing a repeating 2 character message from the main controller micro, but it isn't the " 2D2D2D2D2D2D" described, instead it is 3 and another unrecognized character ( 33h 00h 17h 00h repeating) with some framing errors thown in. I suspect I don't have the serial port settings correct, although I am using the only settings I could find, described in the openinverter main board v3 documentation (115200, 8, none, 2, none).

Can someone with more understanding of the serial connection between the wifi module and the controller PCB on our PCS controllers tell me if these characters are the correct ones I should be seeing? If so I am going to try to proceed with updating the main board directly using the python process.

EDIT: My ossciliscope decodes the message coming from the controller pcb as "2D", so i think there might be a problem with my usb serial adapter. invstigating that next.

Jack Bauer wrote: ↑Tue Aug 23, 2022 7:51 am The logs provided by Solition1DC and P.S.Mangelsdorf contain no PCS messages so you have a hardware / wiring problem. Please see my description of setup here and my log for comparison : viewtopic.php?p=44146#p44146

I just rechecked my wiring again, its correct. So I'm assuming either a bad pcs or a hardware change between versions. I'm going to work on getting it out of the car and opened up to check, will report back.

Edit to add a clarification question: Pin 2 of the 12 pin board connector (v2) doesn't need to be connected anywhere, correct? Its marked as 12V but also "not used"

Very very unlikely to be a hardware change. Do you have the pcs case connected to 12v negative?

Yes the case was grounded with a dedicated wire to 12V neg.

Update: I pulled the PCS out of the car so that I could test with the cover off. Appears to still not have any PCS messages in the log, the only thing I could see was a green and red LED alternating on one end of the board. I'll upload logs, photos, and video later today

P.S.Mangelsdorf wrote: ↑Sat Aug 27, 2022 4:15 pm Edit to add a clarification question: Pin 2 of the 12 pin board connector (v2) doesn't need to be connected anywhere, correct? Its marked as 12V but also "not used"

I've not connected pin 2 of the 12 pin connector and I have the V2 also. Poor photo below, I can get a clearer one if needed. All I needed to connect was pin 1 (Gnd/12V-), pin 2 (12V+), and pin 5 (Input 1 connected to 12V+ also) on the 20 way connector, then ground the PCS case. Have a 5A fuse off the battery 12V terminal too. Firmware Damien posted earlier uploaded via the web interface ok, restarted, and could see the error bits info in the web interface.

Log is attached, link to video below of the only thing I could see, and photos of the inside of the PCS.

Just to give an update I now have a non functioning PCS on the bench and am deep into it. The problem is twofold : Some bugs in the STM32 controller firmware AND as suspected, the differences in firmware versions on the PCS. I don't want to jump the gun in announcing a solution but my gut feel is the STM32 controller will read the firmware rev over CAN from the PCS and adjust its own CAN message format to suit. This is where CAN logs from both working and non working controller and PCS combos will be vital. So do please keep them coming.

@P.S.Mangelsdorf thanks for your efforts. Sadly the log still has no pcs messages. In your video the pcs is awake so the only thing I can think of that has given me that behaviour in the past is failure of the digital isolator chip on the IPC can line. Did your unit come from a working car do you know? Also that is the first 32A PCS I have seen:)

Jack Bauer wrote: ↑Sun Aug 28, 2022 9:04 am @P.S.Mangelsdorf thanks for your efforts. Sadly the log still has no pcs messages. In your video the pcs is awake so the only thing I can think of that has given me that behaviour in the past is failure of the digital isolator chip on the IPC can line. Did your unit come from a working car do you know? Also that is the first 32A PCS I have seen:)

Unfortunately this one came from evil-bay so I don't know if the car was working or not. If I remember correctly you did a video on fixing that issue correct? Or was that something else?

Jack Bauer wrote: ↑Sun Aug 28, 2022 8:02 am This is where CAN logs from both working and non working controller and PCS combos will be vital. So do please keep them coming.

PCS CAN log attached. First few seconds with PCS controller powered up and no enable at Input 1, then input 1 connected to 12V. No HV or AC connected.

Thank you very much indeed.

So thanks to the log from nkiernan we now have 3 know firmwares in play that effect the PCS can messaging.
1)My know working unit in the goose.
2)My Non working unit on bench
3)nkiernan unit.

Unit 3 is complaining of "cpmia" meaning one of the CAN messages its expecting so see originating from the charge port controller is missing. Now the good news is I'm learning as I go. Just sorry i've been away for so long.

We are turnin' and burnin'

Jack Bauer wrote: ↑Mon Aug 29, 2022 3:29 pm We are turnin' and burnin'

So just to flesh this out a bit better. The issues with the non working PCS is indeed down to firmware revisions messing with the CAN message ids and content. Nothing to do with whether its an EU 3 phase or US single phase model. The good news is I know where the messages are being adjusted now and am into a rewrite of the STM32 PCS firmware to accommodate this and add a few features such as auto detecting EU/US , 3 phase or single phase connected and most importantly identifying the PCS firmware revision and adjusting the CAN to suit. This will be clunky at first but the more can logs and info I can obtain the better. I will also open source this firmware in the hopes of encouraging collaboration.

Great news !

Jack Bauer wrote: ↑Sun Aug 28, 2022 9:04 am In your video the pcs is awake so the only thing I can think of that has given me that behaviour in the past is failure of the digital isolator chip on the IPC can line.

Damien could you mark on a photo which component is the isolator? I can do the soldering but identifying components is still just a bit out of my depth.

Also, do you have the part number? It appears there are variations of the ISO7741

Jack Bauer wrote: ↑Sun Aug 28, 2022 9:04 am In your video the pcs is awake so the only thing I can think of that has given me that behaviour in the past is failure of the digital isolator chip on the IPC can line.

P.S.Mangelsdorf wrote: ↑Tue Aug 30, 2022 10:49 pm Damien could you mark on a photo which component is the isolator

The one Damien blew was at U22 on the PCS board.

This video from the first page of this thread shows the ISO7741 removed from U22, and the location from which it was removed (about 30 seconds past where I've bookmarked it below, but for context it's best to start the vid where I've bookmarked it):



". . . about a fiver from Farnell . . . "


I can't help with any variations, though.

Thanks for that!. I'll take a few pictures tomorrow.

Thank you! I had searched the videos but kept missing it.