This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

BQ27441EVM-G1A: Corrupt State of Charge during polling and high load changes

Part Number: BQ27441EVM-G1A
Other Parts Discussed in Thread: BQSTUDIO, GPCCHEM, BQ27426, BQ27427

Using the bqStdio to poll the BQ27441 we observed large jumps in State of Charge during periods of load changes. One was from 57% to 43% and the other 56% to 86%

Any thoughts why this happens and how to avoid?

  • Hello James, 

    Can you please provide the log file for this? 

    Regards, 

    Jonny. 

  • here are the log files

    try2.log

    try3.log
    Fullscreen
    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    12
    13
    14
    15
    16
    17
    18
    19
    Sun Mar 17 14:46:12 EDT 2024
    Device Version Info = 0421_1_09_40
    BQZ Device Name = bq27441G1A
    BQZ Firmware Version = 1_09
    Sample,DateTime,ElapsedTime,Control,CtrlStatus,Temperature,Voltage,Flags,NomAvailCap,FullAvailCap,RemCap,FullChgCap,AvgCurrent,StbyCurrent,MaxLoadCurr,Debug1,AvgPower,Debug2,StateOfChg,Int Temp,SOHStat,SOH,RemCap UnFlt,RemCap Flt,FullChgCap UnFlt,FullChgCap Flt,True StateOfChg,Debug3,Debug4,Debug5,Debug6,Debug7,LogRowTime(ms),LogStatus
    1,2024-03-17 14:46:16,4.013,0x008E,0x008E,19.95,3994,0x01A8,612,1233,557,1178,386,-3,-570,1340,1542,8697,48,19.95,1,88,557,557,1178,1178,48,9406,1104,557,-58,679,792,SUCCESS
    2,2024-03-17 14:46:20,8.016,0x008E,0x008E,19.95,3994,0x01A8,612,1233,557,1178,386,-3,-570,1340,1542,8697,48,19.95,1,88,557,557,1178,1178,48,9406,1104,557,-58,679,805,SUCCESS
    3,2024-03-17 14:46:24,12.020,0x008E,0x008E,19.95,3994,0x01A8,613,1233,558,1178,386,-3,-570,1340,1542,8685,48,19.95,1,88,558,558,1178,1178,48,9406,1104,558,-59,679,799,SUCCESS
    4,2024-03-17 14:46:28,16.034,0x008E,0x008E,19.95,3994,0x01A8,613,1233,558,1178,386,-3,-570,1340,1542,8685,48,19.95,1,88,558,558,1178,1178,48,9406,1104,558,-59,679,802,SUCCESS
    5,2024-03-17 14:46:32,20.053,0x008E,0x008E,19.95,3994,0x01A8,614,1233,559,1178,386,-3,-570,1340,1542,8673,48,19.95,1,88,559,559,1178,1178,48,9406,1104,559,-60,679,798,SUCCESS
    6,2024-03-17 14:46:36,24.055,0x008E,0x008E,19.95,4158,0x01A8,615,1233,560,1178,1026,-3,-570,1340,4266,8661,48,19.95,1,88,560,560,1178,1178,48,9406,1104,560,-61,679,796,SUCCESS
    7,2024-03-17 14:46:40,28.066,0x008E,0x008E,19.95,3998,0x01A8,615,1233,560,1178,392,-3,-570,1340,1567,8661,48,19.95,1,88,560,560,1178,1178,48,9406,1104,560,-61,679,801,SUCCESS
    8,2024-03-17 14:46:44,32.080,0x008E,0x008E,19.95,4152,0x01A8,616,1233,561,1178,1026,-3,-570,1340,4260,8648,48,19.95,1,88,561,561,1178,1178,48,9406,1104,561,-62,679,787,SUCCESS
    9,2024-03-17 14:46:48,36.083,0x008E,0x008E,19.95,4142,0x01A8,616,1233,561,1178,1026,-3,-570,1340,4250,8648,48,19.95,1,88,561,561,1178,1178,48,9406,1104,561,-62,679,832,SUCCESS
    10,2024-03-17 14:46:52,40.094,0x008E,0x008E,20.05,4150,0x01A8,617,1233,562,1178,1026,-3,-570,1340,4258,8636,48,20.05,1,88,562,562,1178,1178,48,9406,1104,562,-63,679,789,SUCCESS
    11,2024-03-17 14:46:56,44.096,0x009E,0x008E,20.05,3901,0x01A8,617,1233,562,1178,0,-3,-570,1340,0,8636,48,19.95,1,88,562,562,1178,1178,48,9406,1104,562,-63,679,786,SUCCESS
    XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

  • i do not see them so i will try again

    0143.try2.log

    7485.try3.log
    Fullscreen
    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    12
    13
    14
    15
    16
    17
    18
    19
    Sun Mar 17 14:46:12 EDT 2024
    Device Version Info = 0421_1_09_40
    BQZ Device Name = bq27441G1A
    BQZ Firmware Version = 1_09
    Sample,DateTime,ElapsedTime,Control,CtrlStatus,Temperature,Voltage,Flags,NomAvailCap,FullAvailCap,RemCap,FullChgCap,AvgCurrent,StbyCurrent,MaxLoadCurr,Debug1,AvgPower,Debug2,StateOfChg,Int Temp,SOHStat,SOH,RemCap UnFlt,RemCap Flt,FullChgCap UnFlt,FullChgCap Flt,True StateOfChg,Debug3,Debug4,Debug5,Debug6,Debug7,LogRowTime(ms),LogStatus
    1,2024-03-17 14:46:16,4.013,0x008E,0x008E,19.95,3994,0x01A8,612,1233,557,1178,386,-3,-570,1340,1542,8697,48,19.95,1,88,557,557,1178,1178,48,9406,1104,557,-58,679,792,SUCCESS
    2,2024-03-17 14:46:20,8.016,0x008E,0x008E,19.95,3994,0x01A8,612,1233,557,1178,386,-3,-570,1340,1542,8697,48,19.95,1,88,557,557,1178,1178,48,9406,1104,557,-58,679,805,SUCCESS
    3,2024-03-17 14:46:24,12.020,0x008E,0x008E,19.95,3994,0x01A8,613,1233,558,1178,386,-3,-570,1340,1542,8685,48,19.95,1,88,558,558,1178,1178,48,9406,1104,558,-59,679,799,SUCCESS
    4,2024-03-17 14:46:28,16.034,0x008E,0x008E,19.95,3994,0x01A8,613,1233,558,1178,386,-3,-570,1340,1542,8685,48,19.95,1,88,558,558,1178,1178,48,9406,1104,558,-59,679,802,SUCCESS
    5,2024-03-17 14:46:32,20.053,0x008E,0x008E,19.95,3994,0x01A8,614,1233,559,1178,386,-3,-570,1340,1542,8673,48,19.95,1,88,559,559,1178,1178,48,9406,1104,559,-60,679,798,SUCCESS
    6,2024-03-17 14:46:36,24.055,0x008E,0x008E,19.95,4158,0x01A8,615,1233,560,1178,1026,-3,-570,1340,4266,8661,48,19.95,1,88,560,560,1178,1178,48,9406,1104,560,-61,679,796,SUCCESS
    7,2024-03-17 14:46:40,28.066,0x008E,0x008E,19.95,3998,0x01A8,615,1233,560,1178,392,-3,-570,1340,1567,8661,48,19.95,1,88,560,560,1178,1178,48,9406,1104,560,-61,679,801,SUCCESS
    8,2024-03-17 14:46:44,32.080,0x008E,0x008E,19.95,4152,0x01A8,616,1233,561,1178,1026,-3,-570,1340,4260,8648,48,19.95,1,88,561,561,1178,1178,48,9406,1104,561,-62,679,787,SUCCESS
    9,2024-03-17 14:46:48,36.083,0x008E,0x008E,19.95,4142,0x01A8,616,1233,561,1178,1026,-3,-570,1340,4250,8648,48,19.95,1,88,561,561,1178,1178,48,9406,1104,561,-62,679,832,SUCCESS
    10,2024-03-17 14:46:52,40.094,0x008E,0x008E,20.05,4150,0x01A8,617,1233,562,1178,1026,-3,-570,1340,4258,8636,48,20.05,1,88,562,562,1178,1178,48,9406,1104,562,-63,679,789,SUCCESS
    11,2024-03-17 14:46:56,44.096,0x009E,0x008E,20.05,3901,0x01A8,617,1233,562,1178,0,-3,-570,1340,0,8636,48,19.95,1,88,562,562,1178,1178,48,9406,1104,562,-63,679,786,SUCCESS
    XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

  • Hello James, 

    Thank you for providing me with the log file for this. Can you also please provide me with the gg file for this, exported from bqStudio? I suspect this could be due to your configuration for load mode. Additionally, was a learning cycle completed for this, and how was the cell matched? 

    Regards, 

    Jonny. 

  • I ran the setup again and captured a jump of 72% to 82%. The batteries are programmed with the golden image developed awhile back. However, this particular battery was from engineering stock so I will reprogram with golden image and then rerun tests. In the mean time here are the log and gg files for the last jump spotted. It's at the end of the file and I stopped logging shortly after.

    try4.log

    try4.gg.csv

  • Hello James, 

    Thank you for providing this. It looks like your load mode is configured correctly, but I would recommend changing the load select to be 6 (user rate) and then configure the user rate for the max load. This should help with this jump you are seeing, because your current configuration is for present average discharge power. Please refer to this document as I believe it will help explain what is occurring here. 

    Regards, 

    Jonny. 

  • I made the recommended changes but still have the same problem. I can reproduce a corrupt State of Charge at any level of battery charge. The changes you suggest were specific to the lower level of charge when using a high dynamic load causing the predict algorithm to be inaccurate.

    I can associate a jump in the state of charge if it happens close to a I2C event. The following scope plots show the relay signal in yellow and the SCK line in blue. When this happens is when I also see the jump. Note that the SCK is on a 1V scale and only dropping a volt which is also strange. The jumps were 61%-75% and 67% to 57%

      

  • Hello James, 

    Are you using the actual battery here or are you using a battery simulator? I have consulted with the team on this and the behavior of the voltage in the previous logs is very strange. 

    Regards, 

    Jonny. 

  • The actual battery is being used. It is the RRC1130.

  • Hello James, 

    What is the load of the system here? It is just very strange that the voltage changes almost instantaneous. A typical cell will accumulate charge and increase/decrease its voltage more slowly. The behavior of the voltage would likely cause this SOC jump in the middle of discharging.  

    Regards, 

    Jonny. 

  • I measure 7.1 ohm across the resistors. On the test stand its just power resistors controlled by a relay. The battery is 3.8V nominal, 3810mAh

  • Hello James, 

    I am not really sure what you mean here, is the load just power resistors and a relay that opens and closes the circuit to the power resistors? Additionally, this still doesn't explain the voltage values you are getting here. The battery voltage should not be behaving like this, the measurements you are getting for the voltage are strange. How was the ChemID matched? 

    Regards, 

    Jonny. 

  • Yes I am using the bqStudio to control a relay to switch in/out a pure resistive load using power resistors.

    The ChemID was matched by the selection of the part. However, we are using a G1A gauge with a 4.35V single cell battery. Could this be part of the issue with the "jumps" we are seeing?

  • Hello James, 

    An improperly matched chemID certainly can affect the SOC. I recommend using the GPCCHEM tool to ensure that the chemID you are using is a good match for your cells. 

    Regards, 

    Jonny. 

  • Since the ChemID is not programable on the bq27441 I will have to disqualify using the -G1A version and test with the -G1B version to see if the problem goes away. Is this correct?

  • More data that perhaps will help resolve this SOC issue:

    I start with programming the BMS with a Golden Image

    I am at 60% SOC, battery parameters are correct in the five boxes

    I apply a LOAD, and see a jump in SOC and the scope shows abnormality on I2C SCK line (may not be the scope plot for this exact event but I've captured many of these now) The SCK line drops to mid-voltage for ~7 msec (always the same each time I see this)

    And, the SOC jumps from 60% to now 77%

    And, all 5 battery parameters are at their default values. 

    It's like the BMS has been reset

    I reload the golden image

    And, now SOC is 29%

    QUESTION:

    Should a LOAD change also change memory contents?

    How do I stop this from happening?

    Do you still believe this is related to battery chemistry?

  • Hello James, 

    Since the ChemID is not programable on the bq27441 I will have to disqualify using the -G1A version and test with the -G1B version to see if the problem goes away. Is this correct?

    For the BQ27441, there are a total of three chemIDs to choose from. If you run the GPCCHEM tool, the report will tell you which chemID (from the 3 to choose from) is the best match for your cell. The reason I am suggesting this is just because the full charge voltage of the chemID matches your cell, does not necessarily mean that the chemID is a good match for the cell. 

    Additionally, you could also try to change the load select to be the user rate, then configure the user rate to be the maximum load that the gauge will see. This should help with this SoC jump as well.

    A load change should not cause the memory contents to be reset unless the gauge has been power cycled or reset. 

    Regards, 

    Jonny. 

  • I now have the results from the chemID study. The part is only an OK match with a 12% accuracy to the battery being used. So, with those results, and the unreliable SOC reporting, we have disqualified using the bq27441.

    Thanks for helping me understand how the BMS parts work.

  • Hello James, 

    Thanks for the update here. The SoC error you were seeing was likely caused by the inadequate ChemID match. You can try to reference the ChemIDs for the BQ27426/BQ27427 to see if the DoD error is less. Another option here is to use a flash gauge. With the flash gauge, you can choose from our list of chemIDs and have a higher likelihood of matching to a cell. If you still do not have an adequate match, another option could be to have a custom ChemID created for your cell. These options are only available with the flash gauges. 

    Regards,

    Jonny. 

  • Thanks. I will keep these options in mind