[joeaax1j]

Hi Mudder,

This evening I opened up the battery compartment on my 2nd Insight that has a 5AhG3 install. I've been using this system for several years without issue. The reason I opened it up was to change my current hack in the MCM from 40% down to 20%. So I did that, switching the jumpers on Bull Dog's board. I then put everything back together, modified the firmware's config.h for the 20% hack, and then compiled the firmware. After getting that uploaded to LiBCM, it is malfunctioning. I'm attaching a photo of my 4x20 display and I'll also upload my ASM output. Any ideas what I've messed up?

Thanks,
Bryan

 

[mudder]

Your ASM output is all kinds of wonky. Maybe the program memory didn't upload correctly. I hesitate to even troubleshoot that data, because LiBCM isn't showing the keyOn event, and also thinks the key is on, and also thinks it's time to grid charge, and also thinks it can't read serial data from the MCM... it's too many wonky things to trust that the firmware installed successfully. You even get a watchdog reboot at the end.

Try uploading the firmware again with both "compilation" and "upload" checkboxes ticked in the Arduino preferences (in the "show verbose output during:" field). Once the firmware upload completes, see if there are any warnings reported in the Arduino upload status window. I bet the upload verification step fails.

I'm certain you're getting a P1648... but if you're getting other P-codes, too, please note them in your reply.

 

[joeaax1j]

Thanks, John! Your reply made me really focus on what had changed in my firmware update process. I forgot that a few weeks ago, I upgraded my computer from Debian 12 to Debian 13. The version of the Arduino IDE included in Debian hasn't really changed (both the old/new versions of Debian include Arduino version 1.8.19), but clearly something about that upgrade has broken the firmware compilation. I'll dig deeper into that. In the meantime, I threw the Arduino IDE on my wife's Windows laptop and was able to successfully compile/upload.

Thanks again,
Bryan

 

[Everett99]

So I picked up the battery pre-built by greentec maybe 8 months ago or so and shortly after I was hearing a noise when the fan came on (it was making contact with the black wire on the bay 3 BMS ribbon) I was able to just push it out of the way a bit and it was no longer making the sound of it contacting anymore. A few months goes by (today) and I head to work and had probably around 65-70% SOC and when I get in after work this is what I see and as I drove the high cell just kept climbing. Ended up pulling over and flipped the switch off. Not sure if this was the result of that fan making contact or something else. For reference I also added the 40% current hack about a month ago and have also noticed 4.3-4.5v during regen

 

[Everett99]

So I picked up the battery pre-built by greentec maybe 8 months ago or so and shortly after I was hearing a noise when the fan came on (it was making contact with the black wire on the bay 3 BMS ribbon) I was able to just push it out of the way a bit and it was no longer making the sound of it contacting anymore. A few months goes by (today) and I head to work and had probably around 65-70% SOC and when I get in after work this is what I see and as I drove the high cell just kept climbing. Ended up pulling over and flipped the switch off. Not sure if this was the result of that fan making contact or something else. For reference I also added the 40% current hack about a month ago and have also noticed 4.3-4.5v during regen
View attachment 118215

So a little update (about 4 hours after flipping ima switch off) I turned it back on and this is now what I’m seeing on the 4x20. Interesting to see it past 85% SOC but this is probably due to the cell imbalance from earlier. Not sure where I should go from here now that all seems fine

 

[Balto]

Something there is NOT right. Unfortunately I don't know enough to add anything further, but I'd avoid driving the car until someone with more knowledge than me responds, or be very careful driving it.

 

[Everett99]

Something there is NOT right. Unfortunately I don't know enough to add anything further, but I'd avoid driving the car until someone with more knowledge than me responds, or be very careful driving it.

Just drove it a couple miles around the neighborhood and it threw a couple codes, cleared them, drove and then rescanned after the drive and they were no longer there. Worse case I just send the battery back to greentec as I don’t really have the time to fully diagnose the issue and with it being under warranty anyway I may as well. Their customer service has been very good but my only concern is I was the one who (installed) it in the car however the BMS cable getting worn through definitely wasn’t on my end as all I did was install the grid charger and ran the wires for the 4x20 and usb cable for firmware updates

 

[mudder]

My guess is LiBCM was (incorrectly) reading one cell's voltage lower than it actually was (for some reason TBD).
Between drive cycles, whatever was causing said low voltage cleared (i.e. all voltages were read correctly), at which point the pack was now charged slightly higher than normal² (e.g. the 88.7% shown in your picture). 89% SoC doesn't pose any safety concerns, but obviously we need to figure out the root cause.

If you're up for gathering data via the ASM, I'd love to see the $TESTC results when the issue is present. This of course assumes you can reproduce the issue.

²LiBCM sets SoC based on the lowest cell voltage. There are additional safeguards to disable regen above a fixed cell voltage, and also to disable assist below a fixed cell voltage.

 

[Everett99]

My guess is LiBCM was (incorrectly) reading one cell's voltage lower than it actually was (for some reason TBD).
Between drive cycles, whatever was causing said low voltage cleared (i.e. all voltages were read correctly), at which point the pack was now charged slightly higher than normal² (e.g. the 88.7% shown in your picture). 89% SoC doesn't pose any safety concerns, but obviously we need to figure out the root cause.

If you're up for gathering data via the ASM, I'd love to see the $TESTC results when the issue is present. This of course assumes you can reproduce the issue.

²LiBCM sets SoC based on the lowest cell voltage. There are additional safeguards to disable regen above a fixed cell voltage, and also to disable assist below a fixed cell voltage.

Oh perfect! Yeah I wish I had my laptop in the car earlier haha. I’ll keep it with me now just in case the problem persists and then I can send the logged data over

 

[Everett99]

I tried to send a video of the 4x20 while I was driving during the event but it didn’t seem like it wanted to send unfortunately

 

[mudder]

Was this a subsequent event (after post#4589), or had you tried to film a video during the single event (reported in post#4584)?

 

[Everett99]

Was this a subsequent event (after post#4589), or had you tried to film a video during the single event (reported in post#4584)?

It was during the single event

 

[Everett99]

Tried to send it in here but doesn’t want to go through. I’ll try to send it via PM

Edit: can’t send videos through PM’s 😓

 

[mudder]

You'll need to post the video you Youtube and then send a link.

 

[Everett99]

You'll need to post the video you Youtube and then send a link.

 

[mudder]

I suspect one of the BMS sense wires had/has an open in it, which can cause voltage measurement error. LiBCM hardware can detect this, but I haven't written the firmware yet to do so; will exist prior to exiting beta.

My first recommendation is to keep your laptop handy so that when this issue occurs again we can figure out which cell has the loose wire (e.g. by typing $TESTC into the ASM window when the issue is present).

Once we know which cell is acting up, we can focus on fixing it (e.g. by replacing one of the three BMS harnesses, etc.

 

[Everett99]

I suspect one of the BMS sense wires had/has an open in it, which can cause voltage measurement error. LiBCM hardware can detect this, but I haven't written the firmware yet to do so; will exist prior to exiting beta.

My first recommendation is to keep your laptop handy so that when this issue occurs again we can figure out which cell has the loose wire (e.g. by typing $TESTC into the ASM window when the issue is present).

Once we know which cell is acting up, we can focus on fixing it (e.g. by replacing one of the three BMS harnesses, etc.

Black on Bay 3

 

[mudder]

Wow, I'm impressed the fan was able to cut through the BMS wire! I certainly wish there was more room between the PCB and lithium modules... but the design is beholden to Honda's original NiMH mechanical enclosure. Obviously not ideal that the cable wasn't tucked far enough back during the installation process.

My recommendation is to contact Greentec for a replacement "18S BMS wire harness adapter". Greentec actually sources these cables from a 3rd party (i.e. I do not make the cables for their LiBCM product). If you have any difficulty getting one through them, send me a PM and I'll send you one I've made (for my own LiBCM Kits, which I do make all the cables for).

 

[retepsnikrep]

The bms wires I have from John have some sort of kevlar/nylon wrap on them to negate this issue.
I have seen fan damage a couple of times now so I use additional cable ties to secure the wires.
Greentech need to make sure the cables they source have the same extra protection.

Are protective grilles available for fans this small?
Maybe something plastic perforated cut to size and glued on the back of the fans.

 

[mudder]

I considered just adding holes to the PCB itself, but didn't think it was necessary at the time... plus, doing so would have obstructed airflow. The problem with adding a discrete grille is it takes up even more room. Without going custom, the thinnest grille I found was 4 mm thick.

With enough care, you can tuck the wires well out of harm's way. I agree Greentec should add TechFlex F6Z1.00-50 nylon wrap around their 18S adapter harnesses. I'll let them know.