[mudder]

Unfortunately I didn't make it very far. Early on the test asks for verification that V is 0V. I'm assuming this is confirming that the grid charger is disabled. I'm not quite sure where to make this measurement (at least not with any confidence).

These voltage locations are labelled as written ('V', 'A', etc) on the daughterboard's lower screw terminal, and are referenced to 'HVDC-' on the daughterboard:

Can this test be done with PCB and modules in the car, or will I have to remove both?

If your DMM probes are long enough, you indirectly measure these points at the 9p connector. Find the 3-conductor wire in the 9p wire harness, which has RED, WHT, & BLK wires, with RED, WHT, & GRN Anderson connectors, respectively. Then:
-RED wire is 'V', and;
-WHT wire is 'A' (Note: the ASM prompt will say 'I' instead of 'A'... it's the current limiting control wire), and;
-BLK wire (GRN Anderson) is HVDC- (i.e. voltage meter negative).

I agree wholeheartedly, it will get working eventually. The limiting factor is likely to be my fingers/wrists at this point, I've taken the modules and PCB in and out of the car enough times now I'm starting to hurt!

If you haven't taken the passenger seat out yet, I feel your pain.

 

[Tynow]

are referenced to 'HVDC-' on the daughterboard:

Dangit! I suspected these might be the pins I was looking for, but was feeling sheepish last night due to a) fatigue from work b) overcast, gloomy weather and c) a nagging lack of confidence.

It's amazing how much all of the little things can add up to affect your mood. In this case I decided to listen and back off for the night. I had plenty of other things to work on while waiting for a reply (see below).

For the grid charger test the battery should be disconnected. I want to be absolutely sure that I've disconnected it properly. Aside from turning the pack switch off, what other steps should I take?

If your DMM probes are long enough, you indirectly measure these points at the 9p connector.

Sounds good to me!

I currently have the top two modules and PCB removed so I could check the fuses and look at all of the wires going to the grid charger (all of them look fine, btw). I'll re-install the middle tray, PCB, and top two modules. I'm getting pretty good at it by now 🙃.

I should have more results tonight.

If you haven't taken the passenger seat out yet, I feel your pain.

I have nearly all of the interior removed. For some reason, I had the thought "well, if I take one seat out for the LiBCM install, I might as well clean/freshen up everything else in the interior!". Foolish. Outside of installing LiBCM, my task list from this weekend stands at:

• Remove both seats, their covers, and foam forms (done)
• Remove interior carpet and any plastic trim panels (done)
• Wash seat covers in the washing machine and allow to airdry (done)
• Wirewheel and paint the rusted seat rail mount points (done)
• Thoroughly clean the now-exposed floor pans and apply sound deadening material (done)
• Give the interior carpet a bath and vacuum (done, but will need another round)
• Apply sound deadening material to the panel behind the seats (done)
• Install LiBCM FoMoCo kit (almost there!)
• Re-attach the seat foam and seat covers (driver's side is done, so many zip ties!)
• Add additional foam/support material to seat backs/bases
• Install seat mount spacers to raise front of seats
• Locate a Civic arm rest, paint and install
• Re-install all interior components

Yes, I've bitten off way more than I could chew for one weekend...

 

[mudder]

For the grid charger test the battery should be disconnected. I want to be absolutely sure that I've disconnected it properly. Aside from turning the pack switch off, what other steps should I take?

Turning the pack switch off (and then back on when requested) is all you need to do.

 

[Tynow]

Alright, we have an update.

The first part (up until battery is connected) passes without issue.

Spoiler: Step-by-step

Unplug charger from wall, disconnect battery, then press 'enter' to continue. Result: pass
Good
Plug charger into wall, then press 'enter' to continue. Result: pass
Good
Verify:
-fans off
-Vout ~= 0
Good, fans off and Vout = 0V
Verify:
-fans on
-Vout ~= 250
Good, fans come on and Vout = 250VDC
Verify:
-fans on
-Vout ~= 105
-Daughterboard 'V' is ~0.7 volts
-Daughterboard 'I' is ~0.7 volts
Good, Vout drops to 105 after ~30 seconds. Fans on, both control wires at 0.7VDC

Connect battery, then press 'enter' to continue.

Verify P_in ~= 1200|1500 watts @ Vin ~= 120|240 volts

At this point the test fails. Power draw is not ~1200W, maximum recorded on the kill-a-watt is 900W. It stays at this power level for about 15 seconds, then drops to 7W. During this time the voltage and current control wires stay at 4.42V and 4.79V respectively.

For the brief time power draw is at 900W, the charger output is at ~172VDC. This seems about right compared to the pack voltage.

Once the power drops to 7W, both control wires slowly fall to ~0.5VDC. This happens well after the power drops.

Verify P_in ~= 950|950 watts @ Vin ~= 120|240 volts

I see this would be the next step in the test. Is it a coincidence that we are seeing this power level before the charger turns off? It seems like the charger is not reaching full power output and goes into a protect (overload?) state...

 

[mudder]

Based on your complete test data, LiBCM's control signals appear to be working correctly, and the grid charger is unhappy for some reason.

I'm assuming you're using Vin=120 volts and have a long and/or thin gauge extension cord. LiBCM limits input current as the input voltage drops. Can you please report the input voltage and current parameters (using Kill-A-Clone meter)?

When the charger input power drops to 7 watts, do the two grid charger fans stay on? Note that this test code will time out after 60 seconds to prevent over-charging the modules (e.g. if I get distracted mid-test).

Note that the actual power input is based on a 60S pack, whereas you're using 48S. As such, even best case you should measure only 80% of the specified wattage. At Vpack=172, you should expect to measure around 900 watts. This is even lower than the 80% value I just mentioned, because the maximum allowed power is current-limited as the input voltage drops below 120 volts (e.g. at 100 Vin the input power is limited to around 650 watts).

Do you have access to a 240 volt plug? If so, please repeat the test with that input voltage to remove any ambiguity.

 

[Tynow]

I'm assuming you're using Vin=120 volts and have a long and/or thin gauge extension cord.

Correct. Vin=120V. I'm using a ~20' 12AWG extension cord, I figured a 14AWG cheapie wouldn't cut it for this.

Can you please report the input voltage and current parameters (using Kill-A-Clone meter)?

Vin unloaded = 123.1 Vrms
Vin loaded (900W) = 116.3 Vrms
Iin loaded = 8.068 Arms

Just in case the Kill-A-Clone was being slow, I also measured Vin at the 4-pole Anderson connector. Both meters agree.

We had noticed that the black Anderson connector in the 4-pole set is missing on the AC line side. From what we can find in schematics this seems correct, but can you please confirm this so we stop thinking about it?

When the charger input power drops to 7 watts, do the two grid charger fans stay on?

I did this again, paying special attention to the fans. They do come on at the point of "Verify fans on and 250V out" but when the battery is connected ("Verify P_in ~= 1200|1500 watts @ Vin ~= 120|240 volts") the fans stop.

Out of curiosity I re-flashed the default firmware (gridcharger test line commented out) and confirmed that the fans are not turning on at any point when connecting the grid charger.

Interestingly enough, they were spinning when I got home today. I tried re-seating the 9p Anderson connector, and the fans have stopped coming on when I plug the grid charger in...

I swear I remember the fans running initially, but there's no doubt that they are not running at the moment.

As such, even best case you should measure only 80% of the specified wattage.

Thank you for confirming this. I had completely overlooked that difference regarding grid charging. It's good to know that the power levels are as expected!

Do you have access to a 240 volt plug? If so, please repeat the test with that input voltage to remove any ambiguity.

Based on the troubleshooting above, I don't know that being on 120V or 240V will make much difference, but I'm willing to try. I have access to a 12kVA line voltage generator (LVG) at work so I can easily try 240V there. It was time for a maiden voyage anyways, right...?

 

[Balto]

Correct. Vin=120V. I'm using a ~20' 12AWG extension cord, I figured a 14AWG cheapie wouldn't cut it for this.



Vin unloaded = 123.1 Vrms
Vin loaded (900W) = 116.3 Vrms
Iin loaded = 8.068 Arms

Just in case the Kill-A-Clone was being slow, I also measured Vin at the 4-pole Anderson connector. Both meters agree.

View attachment 105247 View attachment 105248 View attachment 105249 View attachment 105250

We had noticed that the black Anderson connector in the 4-pole set is missing on the AC line side. From what we can find in schematics this seems correct, but can you please confirm this so we stop thinking about it?

View attachment 105251



I did this again, paying special attention to the fans. They do come on at the point of "Verify fans on and 250V out" but when the battery is connected ("Verify P_in ~= 1200|1500 watts @ Vin ~= 120|240 volts") the fans stop.

Out of curiosity I re-flashed the default firmware (gridcharger test line commented out) and confirmed that the fans are not turning on at any point when connecting the grid charger.



I swear I remember the fans running initially, but there's no doubt that they are not running at the moment.



Thank you for confirming this. I had completely overlooked that difference regarding grid charging. It's good to know that the power levels are as expected!



Based on the troubleshooting above, I don't know that being on 120V or 240V will make much difference, but I'm willing to try. I have access to a 12kVA line voltage generator (LVG) at work so I can easily try 240V there. It was time for a maiden voyage anyways, right...?

You should try re-downloading a fresh firmware from GitHub. I know I mentioned this in a different chat we are in, but I had a corrupted firmware that caused my heartache for 2 weeks with one of the packs I built.

 

[Tynow]

You should try re-downloading a fresh firmware from GitHub. I know I mentioned this in a different chat we are in, but I had a corrupted firmware that caused my heartache for 2 weeks with one of the packs I built.

I'll give it a try when I get home. I actually did try this last night (deleted all firmware form my PC, re-downloaded from Github, re-did config.h, and re-uploaded) but it didn't seem to make a difference. Of course, it won't hurt to try it again, so I'll take a shot at it.

 

[Tynow]

I'll give it a try when I get home. I actually did try this last night (deleted all firmware form my PC, re-downloaded from Github, re-did config.h, and re-uploaded) but it didn't seem to make a difference. Of course, it won't hurt to try it again, so I'll take a shot at it.

Okay, I re-downloaded and installed the firmware again (prerelease 0.9.2a). Config.h set as follows:

Spoiler: Config.h

//Copyright 2021-2023(c) John Sullivan
//github.com/doppelhub/Honda_Insight_LiBCM

//config.h - compile time configuration parameters

#ifndef config_h
#define config_h
#include "libcm.h"

#define FW_VERSION "0.9.2a"
#define BUILD_DATE "2023SEP01"

//////////////////////////////////////////////////////////////////

///////////////////////////////////////
// //
// Hardware Specific Configuration //
// //
///////////////////////////////////////

//'choose' exactly one hardware option from each group below, which must match the actual installed hardware.
//'choose' an option by removing both forward slashes ('//') at the beginning of that line.
//all other options in each group MUST begin with two forward slashes (i.e. they are not chosen).
//there are no default options because this firmware works with all LiBCM variants... you need to specify which hardware you have installed

//choose your battery type:
//#define BATTERY_TYPE_5AhG3 //if you're not sure, you probably have this battery
#define BATTERY_TYPE_47AhFoMoCo

//choose how many cells are in series:
#define STACK_IS_48S //All 5AhG3 Kits & FoMoCo Kits with QTY4 modules
//#define STACK_IS_60S //FoMoCo Kits with QTY5 modules

//choose which grid charger is installed
//#define GRIDCHARGER_IS_NOT_1500W //All 5AhG3 Kits & 'standard' 47Ah FoMoCo Kits
#define GRIDCHARGER_IS_1500W //'faster' 47Ah FoMoCo Kits only

//choose ONE of the following
//must match actual "current hack" hardware configuration:
#define SET_CURRENT_HACK_40 //actually +45.8% //most LiBCM users installed this hardware option
//#define SET_CURRENT_HACK_20 //actually +25.0%
//#define SET_CURRENT_HACK_00 //OEM configuration (no current hack installed inside MCM)

//choose which display to use
//using both displays simultaneously could cause timing issues (FYI: the Serial Monitor prints '*' each time the loop period is violated)
#define LCD_4X20_CONNECTED //display included with all LiBCM Kits
//#define LIDISPLAY_CONNECTED //optional color touch screen display //JTS2doLater: mudder has not yet tested this code. Use at your own risk.

//////////////////////////////////////////////////////////////////

////////////////////////
// //
// Voltage Spoofing //
// //
////////////////////////

//'choose' exactly one option from each group below.
//the default values below should work in all cars.
//modify these parameters if you want more power during heavy assist and/or regen.

//48S ONLY: choose ONE of the following
//60S MUST use 'VOLTAGE_SPOOFING_DISABLE':
#define VOLTAGE_SPOOFING_DISABLE //spoof maximum possible pack voltage at all times //closest to OEM behavior
//#define VOLTAGE_SPOOFING_ASSIST_ONLY_VARIABLE //increase assist power by variably spoofing pack voltage during assist
//#define VOLTAGE_SPOOFING_ASSIST_ONLY_BINARY //increase assist power by statically spoofing pack voltage during heavy assist
//#define VOLTAGE_SPOOFING_ASSIST_AND_REGEN //increase assist and regen power by variably spoofing pack voltage //DEPRECATED (regen too strong)

//48S ignores this parameter (choose any value)
//60S ONLY: to increase assist power, choose the lowest spoofed voltage that doesn't cause p-codes during heavy assist (e.g. P1440)
//#define MIN_SPOOFED_VOLTAGE_60S 180 //voltage spoofing related p-codes won't occur in any car
//#define MIN_SPOOFED_VOLTAGE_60S 175
#define MIN_SPOOFED_VOLTAGE_60S 170 //recommended starting value //choose higher voltage if p-codes occur during heavy assist
//#define MIN_SPOOFED_VOLTAGE_60S 165
//#define MIN_SPOOFED_VOLTAGE_60S 160
//#define MIN_SPOOFED_VOLTAGE_60S 155
//#define MIN_SPOOFED_VOLTAGE_60S 150 //voltage spoofing related p-codes will occur in most cars during heavy assist

//////////////////////////////////////////////////////////////////

/////////////////////////
// //
// Firmware Settings //
// //
/////////////////////////

//the default values below will work in any car.
//you only need to modify these parameters if you don't like the default behavior.

#define STACK_SoC_MAX 85 //maximum state of charge before regen is disabled
#define STACK_SoC_MIN 10 //minimum state of charge before assist is disabled

#define CELL_VMAX_REGEN 43000 //43000 = 4.3000 volts
#define CELL_VMIN_ASSIST 31900
#define CELL_VMAX_GRIDCHARGER 39600 //3.9 volts is 75% SoC //other values: See SoC.cpp //MUST be less than 'CELL_VREST_85_PERCENT_SoC'
#define CELL_VMIN_GRIDCHARGER 30000 //grid charger will not charge severely empty cells
#define CELL_VMIN_KEYOFF CELL_VREST_10_PERCENT_SoC //when car is off, LiBCM turns off below this voltage
#define CELL_BALANCE_MIN_SoC 65 //when car is off, cell balancing is disabled when battery is less than this percent charged
#define CELL_BALANCE_TO_WITHIN_COUNTS_LOOSE 32 //'32' = 3.2 mV //CANNOT exceed 255 counts (25.5 mV)
#define CELL_BALANCE_TO_WITHIN_COUNTS_TIGHT 22 //'22' = 2.2 mV //LTC6804 total measurement error is 2.2 mV //MUST be less than CELL_BALANCE_TO_WITHIN_COUNTS_LOOSE
#define CELL_BALANCE_MAX_TEMP_C 40
//#define ONLY_BALANCE_CELLS_WHEN_GRID_CHARGER_PLUGGED_IN //uncomment to disable keyOFF cell balancing (unless the grid charger is plugged in)

//temp setpoints
#define COOL_BATTERY_ABOVE_TEMP_C_KEYOFF 36 //cabin air cooling
#define COOL_BATTERY_ABOVE_TEMP_C_GRIDCHARGING 30
#define COOL_BATTERY_ABOVE_TEMP_C_KEYON 30
#define HEAT_BATTERY_BELOW_TEMP_C_KEYON 16 //cabin air heating, or heater PCB (if installed)
#define HEAT_BATTERY_BELOW_TEMP_C_GRIDCHARGING 16
#define HEAT_BATTERY_BELOW_TEMP_C_KEYOFF 10
//other temp settings
#define KEYOFF_DISABLE_THERMAL_MANAGEMENT_BELOW_SoC 50 //when keyOFF (unless grid charger plugged in) //set to 100 to disable when keyOFF

#define LTC68042_ENABLE_C19_VOLTAGE_CORRECTION //uncomment if using stock Honda 5AhG3 lithium modules

#define KEYOFF_DELAY_LIBCM_TURNOFF_MINUTES 10 //Even with low SoC, LiBCM will remain on for this many minutes after keyOFF.
//to turn LiBCM back on: turn ignition 'ON', or turn IMA switch off and on, or plug in USB cable

//Choose which sign (±) the LCD displays when the battery is discharging
#define DISPLAY_POSITIVE_SIGN_DURING_ASSIST //current is positive when battery is discharging
//#define DISPLAY_NEGATIVE_SIGN_DURING_ASSIST //current is negative when battery is discharging

//////////////////////////////////////////////////////////////////

////////////////////////
// //
// Debug Parameters //
// //
////////////////////////

//don't modify these parameters unless you know what you're doing. They are primary for mudder's internal testing

//#define RUN_BRINGUP_TESTER_MOTHERBOARD //requires external test PCB (that you don't have)
//#define RUN_BRINGUP_TESTER_GRIDCHARGER //requires external test equipment

#define CHECK_FOR_SAFETY_COVER //comment if testing LiBCM without the cover

#define DEBUG_USB_UPDATE_PERIOD_GRIDCHARGE_mS 1000 //JTS2doLater: Model after "debugUSB_printLatestData"

//#define DISABLE_ASSIST //uncomment to (always) disable assist
//#define DISABLE_REGEN //uncomment to (always) disable regen
//#define REDUCE_BACKGROUND_REGEN_UNLESS_BRAKING //EXPERIMENTAL! //JTS2doLater: Make this work (for Balto)

//choose which functions control the LEDs
#define LED_NORMAL //enable "LED()" functions (see debug.c)
//#define LED_DEBUG //enable "debugLED()" functions (FYI: blinkLED functions won't work)

/*
JTS2doLater:
#define SERIAL_H_LINE_CONNECTED NO //H-Line wire connected to OEM BCM connector pin B01
#define KEYOFF_TURNOFF_LIBCM_AFTER_HOURS 48 //LiBCM turns off this many hours after keyOFF.

Change these #define statements so that all they do is reconfigure EEPROM values.
Also, change this file so that all #define statements are commented out by default.
With these two changes, a user only needs to edit config.h if they want to change a previously sent parameter.
If user doesn't uncomment anything, then the previously uploaded value remains in EEPROM
*/

//JTS2doNow: Implement this feature
//if using 1500 watt charger with 120 volt extension cord, choose input current limit
//#define CHARGER_INPUT_CURRENT__15A_MAX //select this option if using 12 AWG extension cord up to 100 feet, or 14 AWG up to 50 feet**, else if;
//#define CHARGER_INPUT_CURRENT__13A_MAX //select this option if using 14 AWG extension cord up to 100 feet, or 16 AWG up to 50 feet**, else if;
//#define CHARGER_INPUT_CURRENT__10A_MAX //select this option if using 16 AWG extension cord up to 100 feet, or 18 AWG up to 50 feet**.
//**please verify maximum continuous current rating for your specific extension cord

//JTS2doNow: change tabs to QTY4 spaces everywhere
#endif

No change in charger behavior, except that the fans now run all the time while plugged in. I have no idea why the fans would have changed from simply re-downloading and uploading the firmware, but they are working now.

When plugged in, power is 900W for a brief time then drops back to 7W.

 

[mudder]

We had noticed that the black Anderson connector in the 4-pole set is missing on the AC line side. From what we can find in schematics this seems correct, but can you please confirm this so we stop thinking about it?

The black anderson connector isn't populated with the 1500 watt grid charger option. All other grid chargers populate the black anderson connector.

I did this again, paying special attention to the fans. They do come on at the point of "Verify fans on and 250V out" but when the battery is connected ("Verify P_in ~= 1200|1500 watts @ Vin ~= 120|240 volts") the fans stop.

If the fans are stopping, then that means the grid charger itself is unhappy (not LiBCM).
I suspect there's something wrong with the grid charger itself.
Please send me your middle tray assembly so I can troubleshoot.

Out of curiosity I re-flashed the default firmware (gridcharger test line commented out) and confirmed that the fans are not turning on at any point when connecting the grid charger.

Quick sanity check: have you uncommented the GRIDCHARGER_IS_1500W option in config.h?

Based on the troubleshooting above, I don't know that being on 120V or 240V will make much difference

I agree... no need to test 240 mains.

 

[Tynow]

The black anderson connector isn't populated with the 1500 watt grid charger option. All other grid chargers populate the black anderson connector.

Thanks for confirming this. Looking at schematics we figured this was the case, but of course couldn't stop second-guessing ourselves once we saw it.

If the fans are stopping, then that means the grid charger itself is unhappy (not LiBCM).
I suspect there's something wrong with the grid charger itself.

Ah phooey. Alright, I'll plan on pulling the top modules and middle tray tonight.
Unless you need it attached for troubleshooting, I'll probably just disconnect the L, N, G from the supply itself so I can leave the power cord installed in the car. I'd rather not pull the bottom modules and tray if possible.

Knowing that the issue is most likely on the charger gives me some scope focus. While I have it out I'll take a deep dive into everything attached and see if I can spot something that would cause trouble.

Quick sanity check: have you uncommented the GRIDCHARGER_IS_1500W option in config.h?

Indeed I have. You can see it in my last post (#2,069) where I copy-pasted my config.h.

 

[mudder]

If you'd rather, I can just send you a replacement grid charger and you can install it. Of course, if that doesn't work then you'll need to send me everything anyway.

 

[Tynow]

I'll plan on looking it over and seeing if anything obvious sticks out that I missed in my first inspections. If I don't see anything I'll plan on sending it to you.

This may be a terrible idea (and feel free to tell me if it is), but could I remove the charger from the middle tray and send just the charger and cables to you? It would allow me to use the car in the meantime, I'd just install the middle tray without the MW power supply.

To be clear, I don't need to have the Insight up and running right now. I'm just excited to start driving with LiBCM, even if I don't have grid charging capabilities. Again, if this is likely to cause problems you are welcome to be blunt about it.

 

[mudder]

You can send me just the cables and grid charger. There's a small bit of silicon caulk you'll need to separate from the 8p connector plugged into the grid charger, but otherwise it's all just plug-and-play. FYI: If I do another batch of FoMoCo kits, I'm going to make a small PCB to house the few inline components in the grid charger wire harness.

 

[Tynow]

In case anyone is following from home, here's an update on where we are at with this:

I removed the grid charger from the car and inspected all of the wires going to/from it, both visually and with a meter for continuity. Everything looked good. Unfortunately I am not equipped to do much more investigation here, so the grid charger got packed up and is being returned to Mudder. In the meantime, I am able to run LiBCM without the grid charger and have been very impressed so far. It's amazing how much drive-ability has improved with the more robust pack.

I have no doubt that we will get my install up and running 100%. My only fear is that he finds something that I missed/broke and I will be embarrassed...

Thank you Mudder for providing this amazing upgrade to the community, and thank you for your support on my install!

 

[Calpod]

My only fear is that he finds something that I missed/broke and I will be embarrassed...

You have done an amazing job troubleshooting the issue, I'm glad mine got running without any issues as I usually work alone on the installs and the testing equipment I have is very minimal.

 

[Silver2000]

@mudder
Finally got my FoMoCo 60s installed, just not covered up yet. But 4x20 is trying to induce a seizure and not showing real data beyond the timer, temp, SoC and I suspect a non-zero error count. Connecting the gridcharger spreads the pain to the ears but does log that charging is disabled due to a low cell. HVDC- to HVDC+ is 214V and the upper modules are 41.6, 44.7 and 42.4 (back to front). I've gradually reduced CELL_VMIN_GRIDCHARGER to 1000 and still get the low cell error. I flashed the main firmware and had commented out LTC68042_ENABLE_C19_VOLTAGE_CORRECTION since I've got FoMoCo. Any hope that's default uncommented for good cause or is it the E counter on the 4x20? Or am I just lucky to have multiple issues?

Now I'm also realizing OEM_FAN_INSTALLED should be commented out. Comparing config.h in prerelease, I see that's gone entirely and there are some 60s specific notes. Should I flash prerelease instead?

 

[mudder]

Should I flash prerelease instead?

Right now FoMoCo requires Prerelease.

 

[Silver2000]

Indeed, PreRelease is loads better. 4x20 still strobes, but now it's showing data with zero errors, and I'm grateful I'm not sensitive to refresh rates. Which brings us to the data itself... High cell around 4.5V and low around 2.5V which $TESTC says are on IC4 (and IC4 is the lowest V of all 5). The $TESTC IC4 sum matches my measurement over module 5, but it's cell 8,9,10 do not agree with my reading of those cells. I saw 3.46V for each cell in module 5, not the 2V delta the 4x20 and $TESTC report. Curiously, the average of IC4 cells 8,9,10 does agree with my reading. I confirmed my config.h has LCT68042 commented out, but even leaving it dflt did not reconcile the $TESTC values.
I went to bed expecting to spend today gingerly, manually discharging the high cell and charging the low cell to please the machine. Instead, I think I'll exorcise Arduino IDE v2.2.1 and install v1.8.19 that allows copy-paste from serial monitor. If I then return to the car, besides more closely reviewing the module 5 balance wires and PCB traces, should I swap module 5 with another and see where the problem goes? Other advice?

 

[mudder]

1: Yes, please install Aduino 1.x IDE so you can copy paste the cell voltages into this thread.

2: The LCD is flashing to get your attention, because LiBCM thinks a cell is overcharged.

3: I understand your cells aren't actually at 4.5 & 2.5 volts, but if they were then I don't recommend manually balancing them because those (hypothetical) cells are going to be trashed and unsafe to use.

4: Yes, let's swap the BMS cables on modules 4 & 5 and see if the behavior follows the module. Specifically: connect the 37:48 BMS cable to module 5 and the 49:60 BMS cable to module 4. You do not need to remove LiBCM from the car or remove any other cables.
WARNING: DO NOT SWAP BMS CABLES 1:12, 13:24, or 25:36, AS THIS WILL DESTROY THE LiBCM PCB. Swapping BMS cables for cells 37:48 & 49:60 is allowed because those two BMS ICs are electronically isolated (whereas the three monitoring cells 1:36 are not).

5: After performing the above test, if the issue doesn't follow the modules, swap the BMS cables at the PCB level, to see if the issue is caused by the BMS cable itself. This test is based on the fact that you mentioned adjacent (incorrect) cell voltages average out to the actual cell voltages, which suggests one of the BMS sense leads is floating.