Kia Niro BMS

[inkubate1981]

All working now so thanks for the information. Got Comms with one module with 12 cells so will just tweak for the six modules when I've added what I need. Voltages read through ok so very happy.

[dadez87]

Hi guys

I checked your Slave BMS pictures, I would like to confirm some points with you:
1) Each Slave BMS manages 24 cells (2x12 cells) due to the fact that 2x MAX17845 are mounted
2) There are 24 passive balancing resistors (12 pcs per PCB side), so this seems to confirm the above hypothesis
3) For the daisy chain communication with other Slave BMS boards, I have not noticed any "isolation transformer" (for instance, these are used in Porsche Taycan Slave BMS). In this board, I was not able to find any magnetic transformer (easily identifyable due to their bulky size). Did you figure out how the isolation is achieved? Is it performed in a capacitive way instead of magnetic way (less space needed on the PCB)?
4) In addition to cell voltage monitoring, does the Slave BMS board also have temperature sensing? From the data-sheet of the MAX17845 cell monitoring IC, I understand that each of the 2 IC could in principle acquire 2x thermistors signals, so in total 4x external temperature sensors could be connected. Do you know how many of them are actually managed?
5) Did you find any thermistor directly connected on the PCB of the Slave module? (in my understanding, some manufacturers place it to monitor the temperature of the PCB, which could rise extremely during cell balancing; it is some kind of diagnosis function of the board)

Thank you in advance.

[bexander]

Hi guys

I checked your Slave BMS pictures, I would like to confirm some points with you:
1) Each Slave BMS manages 24 cells (2x12 cells) due to the fact that 2x MAX17845 are mounted
2) There are 24 passive balancing resistors (12 pcs per PCB side), so this seems to confirm the above hypothesis
3) For the daisy chain communication with other Slave BMS boards, I have not noticed any "isolation transformer" (for instance, these are used in Porsche Taycan Slave BMS). In this board, I was not able to find any magnetic transformer (easily identifyable due to their bulky size). Did you figure out how the isolation is achieved? Is it performed in a capacitive way instead of magnetic way (less space needed on the PCB)?
4) In addition to cell voltage monitoring, does the Slave BMS board also have temperature sensing? From the data-sheet of the MAX17845 cell monitoring IC, I understand that each of the 2 IC could in principle acquire 2x thermistors signals, so in total 4x external temperature sensors could be connected. Do you know how many of them are actually managed?
5) Did you find any thermistor directly connected on the PCB of the Slave module? (in my understanding, some manufacturers place it to monitor the temperature of the PCB, which could rise extremely during cell balancing; it is some kind of diagnosis function of the board)

Thank you in advance.

1) On the OEM config it manages 9 or 10 cells but it should be able to handle 12 cells, all HW is there for it, but I have not tested this only 10 cells.
2) See above.
3) There need to be isolation transformers on the main board but not in between slave modules. See MAX17841 and MAX17823 (or MAX17845) datasheets where they explain the daisy-chain.
4) Yes, two temperatures per IC so four temperatures per salve module. See picture in post of connector pinout, https://openinverter.org/forum/viewtopi ... 062#p46062.
5) The external temperature sensors are connected via cables to the PCB. There are one internal die temperature per IC which can be read using my code.

Regarding reverse engineering of this entire slave module board, I would start by reading the HW design recommendations from MAX in the datasheets mentioned above. Kia seem to have used what Maxim recommends i most cases as far as I can tell.

[bexander]

All working now so thanks for the information. Got Comms with one module with 12 cells so will just tweak for the six modules when I've added what I need. Voltages read through ok so very happy.

Great! What was the problem?

[inkubate1981]

Great! What was the problem?

Was the 1.5k resistors. Changed to 470 and the signal edges improved. I'm assuming this is because I'm using the OEM transformers and must be slightly different. Thanks

[woleg]

I'm just about to test a new interface which communicates with Range Rover and BMW slaves. https://i.imgur.com/6QupRET.jpeg

How has development been progressing regarding software, do many unknowns remain?

[EV_Builder]

You got picture of said slaves?

[woleg]

Not as yet, haven’t cracked the lid off the module yet. The owner of those modules (14.7kWh) has already used a board from here to interface with this pack successfully. He just wants to repeat the exercise with more modules.

[Gregordoc]

Hello, I'm new here,my name ist Grzegorz and I come from Poland, thank you very much for allowing me to talk to you. I have a kia ev6 battery, I also tried to read the master bms there, unfortunately without success. Then I also had the idea with the slave bms, so I found the forum here. Is there a possibility of finished circuit boards from the ones presented at the beginning to acquire ? the software for the circuits can be downloaded. thanks in advance for your work

[bexander]

Hello, I'm new here,my name ist Grzegorz and I come from Poland, thank you very much for allowing me to talk to you. I have a kia ev6 battery, I also tried to read the master bms there, unfortunately without success. Then I also had the idea with the slave bms, so I found the forum here. Is there a possibility of finished circuit boards from the ones presented at the beginning to acquire ? the software for the circuits can be downloaded. thanks in advance for your work

No, sorry, I do not have finished PCB:s for sale.

[kochajdam]

Dear bexander
Could you please help me where is power pins on slave board and what is the voltage level.
Thank you Marek

[bexander]

Dear bexander
Could you please help me where is power pins on slave board and what is the voltage level.
Thank you Marek

There are no external power pins, they are powered from the battery cells.

[maarek2]

Dear bexander

Thank you for the reply.
I have 12pc of slave modules as you. I want to read voltage with BMS2_7_3.ino (your latest BMS SW)
Thank you once again.
Marek

[maarek2]

Dear bexander

I tested BMS2_7_3.ino only with 6pcs slaves. I haven´t no reply from slaves. Could you please write me if I need all 12pcs for test or not.
I don´t know if old slaves are damaged or MAX17841B is wrong.

[maarek2]

This is my Arduino uno log file.

[EV_Builder]

Biggest chance is the software needs adaptation for amount of slaves and or slave differences.

[maarek2]

Dear EV_Builder
I think BMS is same V2.9

[maarek2]

I have also Battery management unit.
Worse way is find Can on management unit and decode cells voltage but first i want to use code from bexander.
Marek

[bexander]

Dear bexander

I tested BMS2_7_3.ino only with 6pcs slaves. I haven´t no reply from slaves. Could you please write me if I need all 12pcs for test or not.
I don´t know if old slaves are damaged or MAX17841B is wrong.

Biggest chance is the software needs adaptation for amount of slaves and or slave differences.

EV_Builder is correct, the SW needs to be adopted for the number of slaves used.

[maarek2]

Dear bexander.
Cells voltage and temperatures are already working. Now I using 12pcs slaves.
Now I have a little problem with cells ballancing. I edited the code with small error I think.

[bexander]

What errors are you getting and what did you change?
Please post more details.

[maarek2]

Dear bexander
changes in your code:
const uint8_t lowCurrentPIN = 17;
const uint8_t highCurrentPIN = 18;
in loop disabled //readCAN(); because i don't have can

Please see my log. If i change startShuntVoltage from 4050 to 3600 / for Start cell balance I received:
3 2 2 0 2 0 0 0 2 0 2 0 2 0 2 0 2 0 2 0 2 0 2 0 2 0 40 3D C
STATUS cleared
4

Thank you very much for help.

[bexander]

Seem like STATUS register bit ALRTFMEA2 (0x02) is set for all slaves exept no 10??
I don't know why this is set but the problem is that the specific FMEA2 register needs cleering once set. Clearing just STATUS register is not enough. And that is not handled in my code.

Cell #4 in slave #10 is the lowest cell, close to 100mV lower than all the other cells?? Please, verify this with a multimeter and/or check your wiring for this perticular cell.

1: 3610 3633 3633 3633 3630 3634 3633 3633 26
2: 3609 3633 3633 3633 3632 3632 3632 3633 26
3: 3610 3634 3634 3633 3631 3633 3633 3633 26
4: 3608 3632 3632 3632 3632 3631 3630 3632 27
5: 3611 3632 3634 3632 3631 3632 3632 3633 26
6: 3611 3633 3633 3633 3631 3631 3633 3632 27
7: 3632 3637 3638 3637 3636 3637 3637 3637 25
8: 3632 3638 3639 3637 3637 3637 3638 3638 27
9: 3633 3639 3639 3640 3636 3638 3638 3638 27
10: 3632 3638 3638 3508 3637 3637 3637 3637 26
11: 3631 3638 3639 3638 3636 3638 3638 3638 27
12: 3631 3639 3639 3638 3638 3638 3639 3638 26

[bexander]

What happends if you start the cell balancing at a higher voltage level and then charge the pack a bit? Do you still get the same error?
Asking cause the FMEA2 indicated something with slave supply voltage and the internal charge pump of the slaves.

[maarek2]

Dear bexander
The current was causing the problem i think. My test charger make only 0,5A /350V.
Now i split one baterry to 6small in parallel connection. I using charger with 60V and 10A and balansig working 10 min without error.
Die temperatures are also higher.
Thank you very much for your help!!!