• State TI Thinks Resolved
  • Locked Locked
  • Replies 15 replies
  • Answers 3 answers
  • Subscribers 65 subscribers
  • Views 365 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

Original question:

BQ27411-G1: Learning Cycle Error for BQ27411-G1

BQ27411-G1: BQ27411-G1

Mridula B
Prodigy 20 points
Part Number: BQ27411-G1
Other Parts Discussed in Thread: BQSTUDIO, , GPCCHEM

Hi,

I have been trying to complete the learning cycle for BQ27411-G1 evk with a 950mAh LiPo battery. The learning cycle is performed using bqStudio in Automatic mode, with a Relay control setup for switching between load and charger. But after the charging cycle, an error comes:

"

Error: [QMAX_UP: Not Set] 

Learning cycle cancelled

"

I have been trying to debug this issue and thus I logged the voltage, current and control register values.

I have attached the register log, image of the battery used ,screenshots of values given to RA table and values given to the data memory.

3240.log3.log

    

Kindly help me out to resolve this issue and complete the learning cycle successfully to generate the Golden Image.

  • 0
    TI__Guru 53616 points

    This means that the gauge couldn't update Qmax for one of several reasons (that can't be clearly identified with the provided data):

    You screenshots show that most parameters aren't configured (0 by default). Please go through each parameter and set them according to the meaning as described in the TRM.

    For example, the Dsg, Chg and Quit Current (Thresholds) aren't set at, hence the gauge won't know which state the cell is in (charge, discharge or relax). These and all other parameters that are 0 by default must be set to reasonable values before running a learning cycle.

  • 0 in reply to Dominik Hartl11
    Prodigy 20 points

    Hi,

          As suggested we did run a learning cycle by setting the parameters in the data memory table. But even after setting those values we were unable to complete a successful learning cycle, the cycle was canceled because of the same problem as before. Below is the log of the learning cycle :
    Thu Jun 29 18:10:09 IST 2023: Learning cycle started...
    Thu Jun 29 18:10:09 IST 2023: Start task: Discharge ( TerminateVoltage = 2900 mV )
    Fri Jun 30 01:36:44 IST 2023: End task: Discharge completed in 07:26:34
    Fri Jun 30 01:36:44 IST 2023: Start task: Relax ( Duration = up to 360 minutes )
    Fri Jun 30 01:36:44 IST 2023: End task: Relax completed in 00:00:00
    Fri Jun 30 01:36:44 IST 2023: Start task: Charge ( TaperCurrent = 50 mA )
    Fri Jun 30 08:44:38 IST 2023: End task: Charge completed in 07:07:53
    Fri Jun 30 08:44:38 IST 2023: Start task: Relax ( Duration = up to 120 minutes )
    Fri Jun 30 08:44:40 IST 2023: Error: [QMAX_UP: Not Set]  
    Fri Jun 30 08:44:40 IST 2023: Learning cycle canceled

    From the log, we noticed that the relaxation task was not happening after the discharge task even though we had set the Dsg.Relaxation time for 360 minutes.
    Can you please help us solve this issue and help us to find what we are missing in completing a successful learning cycle? 
    Another thing we found is that the bat_det bit becomes low sometime after the start of the learning cycle could this be the cause of the learning cycle getting failed? If so can you suggest us some solutions to stop this from happening 
    I am attaching the register logs and the screenshot of the values we have set in the data memory configurations.
  • 0 in reply to Mridula B
    TI__Guru 53616 points

    The BAT_DET bit must not change. The gauge will not learn anything if BAT_DET is 0. You can configure the gauge to not monitor the battery detection pin but instead force BAT_DET = 1:

    2.5.4 Battery Detection (BIN), TRM: www.ti.com/.../sluuas7b.pdf

  • 0 in reply to Dominik Hartl11
    Prodigy 20 points

    Hi ,

    Sorry we could not find a pin called BIN as suggested in the document mentioned . I am attaching the picture of our board and the data sheet . Please help us to find the pins 

  • 0 in reply to Mridula B
    TI__Guru 53616 points

    BIN is not available on this EVM. If the problem is with BAT_DET, then please make sure you set OpConfig[BIE] to 0, followed by sending the BAT_INSERT subcommand. This should "stick" BAT_DET to 1. If it doesn't, then the gauge may experience an unexpected reset (=check battery voltage and make sure it won't drop too low temporarily).

  • 0 in reply to Dominik Hartl11
    Prodigy 20 points

    Hi,

    We looked up the OpConfig[BIE] after sending the BAT_INSERT  command and we found that it was set as zero already. So hope it is not the problem.

    When we did run again by confirming this we did end up having another error saying : 

    Tue Jul 04 17:05:09 IST 2023: Learning cycle started...
    Tue Jul 04 17:05:09 IST 2023: Start task: Discharge ( TerminateVoltage = 2900 mV )
    Tue Jul 04 17:12:11 IST 2023: End task: Discharge completed in 00:07:02
    Tue Jul 04 17:12:11 IST 2023: Start task: Relax ( Duration = up to 360 minutes )
    Tue Jul 04 23:12:17 IST 2023: Error: [RUP_DIS / RDIS: Not Clear]
    Tue Jul 04 23:12:17 IST 2023: Learning cycle cancelled

    Another concern is that we are using a fuel gauge with unmatching chemical id because in the data sheet it says that it is for a LiCoO2 type battery but our battery is LiPo. When we read the chemical id already programmed in the fuel gauge it was  0x0128 .

    Does this information help you to figure out our problem.

    Kindly help us since we are running out of our time .

    Thank you

  • 0 in reply to Mridula B
    TI__Guru 57697 points

    Hello Mridula,

    The first step in designing with this gauge is to use the GPCCHEM tool to find out if this gauge supports the battery chemistry. If the chemistry is not a good match, then the gauge will not be accurate and learning cycle will usually fail.

  • 0 in reply to Shirish
    Prodigy 20 points

    Hi,

    We were not able to get a complete log using bq-27411 evm ,so :

    1)We used another gauge (bq-27542G1) in which chemical id can be programmed for taking the logs to upload to the GPC tool is this ok?

    2)We are planning to use bq-27441 G1B gauge for our battery in our custom board ,Is this compatible for our battery (battery image given above)?

  • 0 in reply to Mridula B
    TI__Guru 53616 points

    #1: this is ok.

    #2: The bq27411 doesn't support Lithium Iron Phosphate chemistries (the built in ChemID doesn't cover the voltage range nor the curve for Lithium Iron Phosphate cells).

  • 0 in reply to Dominik Hartl11
    Prodigy 20 points

    Hi,

    Since bq-27411 doesn't support our battery can we use bq-27441 G1B gauge ? We use a LiPo (Lithiyum Polymer) battery of 950mAh capacity.

  • 0 in reply to Mridula B
    TI__Guru 53616 points

    Oh, sorry, I misread that you use Lithium Iron Phosphate, not LiPo.

    The bq27411 and bq27441 both support LiPo but it depends on the actual cell chemistry. The GPC tool that Shirish mentioned will tell you which gauge is compatible with your specific cell. There is no other way to tell. Please follow the instructions from GPC and check the result returned by GPC for the bq27411 and bq27441.

  • 0 in reply to Dominik Hartl11
    Prodigy 10 points

    Hi,

    We have been attempting to create a golden image for the bq27441-G1B fuel gauge however, we failed to create the same using bq27441 G1B fuel guage too..  The cycle was canceled after completing the relaxation time following discharge.

    We were unable to obtain the chemical ID of the battery used as you suggested. we could not generate a log as required by the GPC tool since the learning cycle as demanded by the tool was not successfully completed.

    We are seeking your assistance in resolving this matter and ensuring the successful creation of the golden image. 

  • 0 in reply to Saranya R
    TI__Guru 57697 points

    Hello Saranya,

    The logs for GPCCHEM are independent of learning cycle.

    You should log data using the Registers window while applying charger/rest/load as specified in the instructions.

    The learning cycle plugin is useful after the correct chemID has been determined.

  • 0 in reply to Shirish
    Prodigy 10 points

    Hi

    Based on your suggestion, we tried to get the chemical ID using the GPC tool. When we first tried, the GPC tool gave us a warning message. After looking at the log data more closely, we found some unexpected negative values at the end of the charging phase and just before the relaxing phase. We fixed this by changing those negative values to zeros and tried again. This time, we successfully got the chemical ID. Please take a look at the attached log files and GPC report.

    According to the report, the best chemical ID is 2603, and we couldn't find our fuel gauge (bq27441-G1B) on the recommended devices list.

    Could you kindly verify whether it is not possible to utilize the bq27441-G1B fuel gauge for our battery?

    /cfs-file/__key/communityserver-discussions-components-files/196/bq_5F00_chemid.zip

    /cfs-file/__key/communityserver-discussions-components-files/196/bq_5F00_chemid_2D00_report.zip

    /cfs-file/__key/communityserver-discussions-components-files/196/log_5F00_4_5F00_8_5F00_23.log

  • 0 in reply to Sabik T K
    TI__Guru 57697 points

    Hello Sabik,

    You cell is not the generic kind as it does not match any generic ID.

    bq27441 will not give good results with this cell.