Simp BMS

[Dusty4]

Try flashing the display.

Flashed the display yesterday Tom, still nothing mate

[Dusty4]

I found a resolution to my problem today, a new screen got delivered in the mail. Flashed the screen and plugged in. Worked straight away.

[BogdanT]

Jumping aboard this train. Is anyone using Volt/Ampera + SIMP BMS up for some testing? I noticed that the balancing code on the SIMP is not ok, and would like to push a patch but not before testing on some different setups. Shoot me a PM if balancing is not working on your side and would like to test.

[Boxster EV]

Jumping aboard this train. Is anyone using Volt/Ampera + SIMP BMS up for some testing? I noticed that the balancing code on the SIMP is not ok, and would like to push a patch but not before testing on some different setups. Shoot me a PM if balancing is not working on your side and would like to test.

How do you know it’s not balancing (looking at display or debug)?

[BogdanT]

No, neither debug or display will show if it's actually working.

Checking the code I noticed that balancing on Volt/Ampera was never enabled (bug) - I sent it to Tom and he reactivated the balancing commands 2 days ago.
Then, when I tested, it froze my BICMs. I checked the balance code and noticed that it doesn't send the data as per the spec. Further testing revealed some bugs in the voltage decode method.
I'm chatting to Tom but as he doesn't have a test rig I need to try and sort it out before I send the code to him to merge.

On my side it's balancing now, and I'm just waiting to see if the BICMs can discharge 'all' the cells, or if I need to trigger that for odd/even cells sequentially. Testing the balancing takes some time (12hours to discharge ~30mV)

[mdrobnak]

I have the Volt BICMs but have only wired up module 4. I have the Chrysler Pacifica / LG cells that I will being using it with. I hope to be further along with battery testing within a month. Need to get my Rust ISA shunt library written first.

[BogdanT]

So far so good on my side. Balanced to 15mV now back to charging.
I do see some modules that report a constant temperature (17C)... Maybe the sensors are broken?

[Boxster EV]

So far so good on my side. Balanced to 15mV now back to charging.
I do see some modules that report a constant temperature (17C)... Maybe the sensors are broken?

Good work. Care to share the changes you’ve made to the code?

[BogdanT]

I'll ping you in 1-2 days, it's a bit of a mess now as I got carried away and updated a bunch on the program logic + the display. it'll probably require SIMP & display update.

[Dusty4]

I'll ping you in 1-2 days, it's a bit of a mess now as I got carried away and updated a bunch on the program logic + the display. it'll probably require SIMP & display update.

Hi BogdanT, how did you get on with this update? I have noticed that mine is not balancing, I can test on my setup if you like.

[tom91]

There is a Chevy volt update to the firmware to correct balancing issues.

It can be found here: https://github.com/tomdebree/AmperaBatt ... /VoltBMSV2

[Dusty4]

Cool, thanks Tom.

[BogdanT]

I'm currently testing some updates on the balance process (only Volt/Ampera version) with Paul (speed in particular). Can you share your results with the main branch software?
PS: if you want to test the updates send me a PM

[jon volk]

I'm currently testing some updates on the balance process (only Volt/Ampera version) with Paul (speed in particular). Can you share your results with the main branch software?
PS: if you want to test the updates send me a PM

I pulled your latest last night and after a couple mods for my hardware am happy to report the balancing is working quite well. One pack has always had a 165mV delta. Its down to 75 after around 22hours.

[BogdanT]

Awesome! Could you also look out for something I noticed with Paul?

Balancing process seems to drain the other cells as well (as of now assumed to be due to constant voltage sampling @200ms during balancing). At 100mV discharged (by balance) from the top cells you should also see around 30mV discharged from all other cells.

If that is consistent across various setups I'm thinking of starting the balance process ONLY on AC present, and fully stop voltage sampling when BMS is in "Ready" state. This is because even in "Ready" without balancing, the current code is still sampling voltages every second, and that might be risky if the car sits for a couple of months with the BMS on.

[jon volk]

I would have to make it a point to log over a set period with a 12v trickle charger plugged in. My 12v battery is just about done so my car kicks the dc/dc on a bunch of times throughout that day lately. So there will be some drain present from that.

[Boxster EV]

My pack was horribly out of balance to the tune of like 300mV at full charge.

After some testing with Bogdan over Christmas the new code is balancing at >20mV per day. Delta is now tiny. He’s also introduced some refined logic so that charging will not be allowed until balancing criteria has been met.

I’ll report back as further testing progresses.

Top work, even if I do now have a Trabant as a splash screen upon starting the display.

Would be great to figure out how to get the Teensy to sleep when it’s not doing anything. I’m running three of the buggers in my car and the 12v battery is not impressed.

Cheers,
Paul

[jon volk]

Pretty happy with the results. This pack was 165mV out last week.

[tom91]

Pretty happy with the results. This pack was 165mV out last week.

Jon, thanks for the feedback. The balancing code has been rolled into my main branch already.

If anyone else finds quirks in the code / missing items they have a fix for please let me know and I can implement these in the main branches for everyone to enjoy.

Example: I have support for the Isa Shunt running stock canbus configuration, but not pushed to all branches yet until someone wants it and can test it.

[jon volk]

I could probably do some ISA testing. I was going to roll the ISA into the next iteration of my main VCU, but could utilize it from the BMS level.

The current BMS code that Im running stripped out all but the necessities so Ill likely need to adapt it but can validate the functionality.

[rstevens81]

Hi tom,

I have 2x VW (passat) battery packs for my project, I have now managed to get the modules to talk to your teensy (3.2/3.6) quite happily.
[I don't want to digress about the issues of eddies with dual pack setups, am trying to find ways to be productive on the project without spending any money in the current climate ]
I understand that Simp BMS can only (currently) talk to a maximum of 1 full pack due to the limitation of CAN ID's.
My Intention was to use a Arduino Due / ESP32 to act as a CAN Man in the middle to basic offset the ID's from the second pack to overcome the ID Limit (I know i would require to mod the firmware), the idea being that if i got it work it could be useful to others without messing with the Simp BMS hardware (Which i would do if i based it around a teensy 3.6).
So i started to look at your code for get a better understanding of things, noticed something interesting ...
you seem to have already added code for extenders? there seems to be an entire extra battery packs worth of id's is this a previous attempt to use a second battery pack using the serial expansion? (there seems to be a lot of serial expansion code commented out).



For anyone reading this this only seems to apply to the VWBMS code

Code: Select all

    ///////////////// one extender increment//////////
    
    case (0x1D0):
      CMU = 9;
      Id = 0;
      break;
    case (0x1D1):
      CMU = 9;
      Id = 1;
      break;
    case (0x1D2):
      CMU = 9;
      Id = 2;
      break;

    case (0x1D4):
      CMU = 10;
      Id = 0;
      break;
    case (0x1D5):
      CMU = 10;
      Id = 1;
      break;
    case (0x1D6):
      CMU = 10;
      Id = 2;
      break;
    case (0x1D8):
      CMU = 11;
      Id = 0;
      break;
    case (0x1D9):
      CMU = 11;
      Id = 1;
      break;
    case (0x1DA):
      CMU = 11;
      Id = 2;
      break;
    case (0x1DC):
      CMU = 12;
      Id = 0;
      break;
    case (0x1DD):
      CMU = 12;
      Id = 1;
      break;
    case (0x1DE):
      CMU = 12;
      Id = 2;
      break;

    case (0x1E0):
      CMU = 13;
      Id = 0;
      break;
    case (0x1E1):
      CMU = 13;
      Id = 1;
      break;
    case (0x1E2):
      CMU = 13;
      Id = 2;
      break;

    case (0x1E4):
      CMU = 14;
      Id = 0;
      break;
    case (0x1E5):
      CMU = 14;
      Id = 1;
      break;
    case (0x1E6):
      CMU = 14;
      Id = 2;
      break;

    case (0x1E8):
      CMU = 15;
      Id = 0;
      break;
    case (0x1E9):
      CMU = 15;
      Id = 1;
      break;
    case (0x1EA):
      CMU = 15;
      Id = 2;
      break;

    case (0x1EC):
      CMU = 16;
      Id = 0;
      break;
    case (0x1ED):
      CMU = 16;
      Id = 1;
      break;
    case (0x1EE):
      CMU = 16;
      Id = 2;
      break;


    ///////////////////////standard ids////////////////
    

    case (0x1B0):
      CMU = 1;
      Id = 0;
      break;
    case (0x1B1):
      CMU = 1;
      Id = 1;
      break;
    case (0x1B2):
      CMU = 1;
      Id = 2;
      break;

    case (0x1B4):
      CMU = 2;
      Id = 0;
      break;
    case (0x1B5):
      CMU = 2;
      Id = 1;
      break;
    case (0x1B6):
      CMU = 2;
      Id = 2;
      break;

    case (0x1B8):
      CMU = 3;
      Id = 0;
      break;
    case (0x1B9):
      CMU = 3;
      Id = 1;
      break;
    case (0x1BA):
      CMU = 3;
      Id = 2;
      break;

    case (0x1BC):
      CMU = 4;
      Id = 0;
      break;
    case (0x1BD):
      CMU = 4;
      Id = 1;
      break;
    case (0x1BE):
      CMU = 4;
      Id = 2;
      break;

    case (0x1C0):
      CMU = 5;
      Id = 0;
      break;
    case (0x1C1):
      CMU = 5;
      Id = 1;
      break;
    case (0x1C2):
      CMU = 5;
      Id = 2;
      break;

    case (0x1C4):
      CMU = 6;
      Id = 0;
      break;
    case (0x1C5):
      CMU = 6;
      Id = 1;
      break;
    case (0x1C6):
      CMU = 6;
      Id = 2;
      break;

    case (0x1C8):
      CMU = 7;
      Id = 0;
      break;
    case (0x1C9):
      CMU = 7;
      Id = 1;
      break;
    case (0x1CA):
      CMU = 7;
      Id = 2;
      break;

    case (0x1CC):
      CMU = 8;
      Id = 0;
      break;
    case (0x1CD):
      CMU = 8;
      Id = 1;
      break;
    case (0x1CE):
      CMU = 8;
      Id = 2;
      break;

[tom91]

So i started to look at your code for get a better understanding of things, noticed something interesting ...
you seem to have already added code for extenders? there seems to be an entire extra battery packs worth of id's is this a previous attempt to use a second battery pack using the serial expansion? (there seems to be a lot of serial expansion code commented out).

Correct this is in the VW code, however the canbridge used in testing did not give consistent enough results. So if you want to do some testing you can enable the sections relation to the additional Ids.

The biggest down side of the VW code right now is the fact balancing has not been figured out yet.

[boekel]

about multiple VW packs: I've experienced heavily unbalanced modules when using a CANbus brigde to offset addresses. Not sure what happened but might have to do with the modules seeing those messages and doing strange things...

[arber333]

about multiple VW packs: I've experienced heavily unbalanced modules when using a CANbus bridge to offset addresses. Not sure what happened but might have to do with the modules seeing those messages and doing strange things...

Hm... what if you would use and connect those cells in paralel modules with some 2.5mm2 wires? That should stabilise individual cells inside parallel modules. Then you only need a single BMS circuit? Can you get to the top(sides) of the cells? There should be bolts to connect to.

[Ev8]

I’m trying to do the same thing with bmw phev packs, I currently have 3 whole packs and am going to run into address limits, I had thought about paralleling cell taps but I would rather use all 15 slaves if possible, I suppose the other way would be run 3 complete simp bms. CAN is still new to me so creating a way of offsetting addresses is beyond my skill set!