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.

BQ34Z100: FC bit is not getting set when charging to full while running learning cycle

Part Number: BQ34Z100
Other Parts Discussed in Thread: GPCCHEM, , BQSTUDIO, , EV2400

Hi,

I got new battery with different capacity (1020mAh) for which I'm trying to generate golden file using BQ34z100 fuel gauge. FC bit never set when charging to full.

Used default chemistry in fuel gauge for my battery (Nokia BL-5C). Calibration values used for this learning cycle are generated long back with old battery (2150mAh). I have used bq Gas Gauge Evaluation Software (which I'm comfortable with used for my old battery learning cycle) for this learning cycle. Attached are the latest logs from this GUI.

Some of the Important settings which I changed while running learning cycle are mentioned below:

Taper Current: 150mA (130mA also set in another run)

Minimum Taper Capacity: 25mAh default value (2mAh and 0mAh also tried)

FC Set%: 100% (-1% also tried)

After doing all the above changes also, FC bit never set. Because of this, I was not able to complete my learning cycle after trying for a week.

Pls help what causing the issue.

Thanks

Vidyasagar

Charging_Logs.zip

  • Hi Vidyasagar,

    Is there a reason why you're not using the latest firmware?

    Your taper window and taper currents look good.

    This issue could be caused by a poor chemID match. How did you characterize your cells?

    Since you are using a new battery with different capacity, please utilize the gpcchem tool to generate a new chemID for your cells.

    If you run into any issues with finding a better match, please let me know.

    Sincerely,
    Bryan Kahler
  • Thanks for a quick response.

    As I was comfortable with old GUI, used the same now. No other specific reason.

    Provide me link for latest GUI & let me know if anything need to be updated in fuel gauge flash for using the latest GUI.

    Are you sure that mismatch in chemID is the reason for not setting FC bit ?

    I even programmed BL-4C chemistry into fuel gauge & restarted the leaning cyle. But again failed at FC bit.

    I'll provide you required data for chemID analysis on monday.

    Thanks

    Vidyasagar 

  • Hi Bryan,

    I uploaded required data for gpchem tool, but the report given by the tool did not mention exact chem ID. Attached my input & tool report for your reference.

    I just did charging, relaxation, discharging & relaxation. Tool report says the difference between initial and final DOD is less than 90% dDOD_%=  65 & mentioned to decrease the discharge rate. For a 1000mAh battery, I discharged with 120mA (which is >C/10 but <C/5). Do I need to discharge with further less current for proper ChemID detection ? please mention the required discharge current.

    Report mentioned ChemID match for bq274xx devices only. Nothing mentioned for bq34z100, which I'm using. 

    When tried with bqstudio, bq34z100EVM is not auto detected. Found on Ti forum that I need to update flash of my EVM. If I update, can I able to use this EVM with my old GUI (bq Gas Gauge EVSW). The reason for asking is that I never used bqstudio. If I'm unable to proceed further with bqstudio, atleast I can go back & use my old GUI. Provide me your response.

    Why you people won't make a single document for creating golden image file from beginning using bqstudio. This will help many TI customers & we no need refer many documents & we can easily work with bqstudio. Is there any such document ?

    Longback when I worked on bq27520 fuel gauge, there is bqEasy tab inside GUI which guides step by step preparation of learning cyle. Why there is no such thing in the later versions of the tools/for other fuel gauges.

    Based on the approx ChemID match provided by tool report, I programmed Chem ID 128 into my fuel gauge on the EVM & started learning cycle with my old GUI.

    In the mean time, plese help me on the above queries so that I can complete learning cycle successfully.

    Thanks 

    Vidyasagar3000.GPC_Tool_Input-report2.zip

  • Hi Vidyasagar,

    Thank you for sending me the GPC_report.

    Could you please also send the data log you used with the tool? A 100% error is indicative of an improper learning cycle. I understand that you're experienced with these gauges, but looking over the log will help me pinpoint the root cause.

    The GPCCHEM tool returned an error warning that the total discharge time was less than 5 hours. If this is the case, the tool either didn't detect the discharge point correctly, or the discharge was less than 5 hours (from full), which would either be an incomplete discharge cycle, or a discharge cycle at a rate greater than C/5. 120 mA would be roughly a C/8 rate of discharge assuming the battery was fully charged and fully discharged (and would require a corresponding 8 hours to discharge). Unfortunately, this is all speculation without being able to see the logs.

    With respect to bqStudio, what problems are you experiencing? Please make sure you're using the latest version 1.3.80. The bq34z100-G1 bqz file is included with the latest version of the software. The EVM user guide provides a walkthrough of using the EVM with bqStudio, found here: www.ti.com/.../sluu904

    Are you using an EV3200 to connect to the device? If you EV2300 is version 3.1C, please update the firmware of the EV2300. Update instructions and files can be found in this thread: e2e.ti.com/.../317313

    Thank you for your feedback about the lack of documentation that you find helpful for creating a golden image. Are you looking for an updated document with bqStudio that is similar to the following document? www.ti.com/.../slua544.pdf

    Sincerely,
    Bryan Kahler
  • Hi Byyan,

    Attached input data given to the tool (GPC_Tool_Input_old) for your reference.

    Today with the bqstudio GUI & with default battery chemistry, FC bit was set when battery fully charged. But even after relaxation of more than 2hours, OCVTAKEN is not Set, VOK is not clear. Qmax Cell 0 & Update Status are not updated. Can you suggest what could be the issue & how should I proceed.

    Attached today's learning cycle logs of battery (GPC_Tool_Input) also for checking at your end.

    Please confirm if issue could be due battery & suggest how can I confirm this.

    Thanks

    Vidyasagar

    GPC_Tool_Input_old.zip

    GPC_Tool_Input.zip

  • Hi Bryan,

    Today I ran the learning cycle again with my old battery (2000mAh) using default chemistry. I successfully reached last step (discharging) without any issues. By tomorrow morning relaxation will complete & Update status will be updated to 06.

    My target is to load the above generated file as golden image into other devices & convert the communication into HDQ mode so that 4 LED's controlled by fuel gauge can show the battery SOC

    I have few queries which require your help in this process

    1. Can I verify the accuracy of fuel gauge by running one more charge & discharge cycle after update status goes to 06 ? Do I need to fire any commands to gauge before doing this ?

    2. Can you help with the commands which I need to fire to the gauge on the EVM before exporting the recently generated file as golden image. Also help me with commands which I need to fire once this image is loaded into other devices.

    3. I want to verify the accuracy of fuel gauge once it is converted into HDQ mode. Can I cross verify LED's status with GUI SOC% in this mode ?

    4. Will I able to reprogram in HDQ mode ?

    Thanks

    Vidyasagar 

  • Hi Vidyasagar,

    To answer your questions:
    1. Can I verify the accuracy of fuel gauge by running one more charge & discharge cycle after update status goes to 06 ? Do I need to fire any commands to gauge before doing this ?

    Assuming the device is now fully configured, once IT ENABLE is set (this is the second to last command to be sent to the device in production, with sealing the device being the last command), the gauge will begin updating as you run the charge and discharge cycle. This is a great opportunity to watch the registers and watch the OCV values set, perform an update, and reset the max error %.

    2. Can you help with the commands which I need to fire to the gauge on the EVM before exporting the recently generated file as golden image. Also help me with commands which I need to fire once this image is loaded into other devices.

    When creating the golden image for master programming, please optimize across 10-30 devices and average the values (assuming variance between the devices is low).

    Please make sure to set the update status to 0x02 and set cycle count to 0 on the golden image.

    3. I want to verify the accuracy of fuel gauge once it is converted into HDQ mode. Can I cross verify LED's status with GUI SOC% in this mode ?

    The gauge will still communicate with bqStudio in HDQ mode with the EV2300/EV2400. After issuing the command to convert to HDQ mode, please reconnect the lines from I2C to HDQ.

    4. Will I able to reprogram in HDQ mode ?

    Yes, but it will be much slower than I2C and is not advised.

    Sincerely,
    Bryan Kahler

  • Hi Bryan,

    Thanks for your response.

    I read steps to be followed from multiple documents (as there is no single such document available) while exporting golden image & loading the same on production line. Please confirm steps I mentioned below are okay or not. Modify if required.

    1. Export .gg file of golden image. Change update status to 02. Import it. Now generate golden image (.srec) file with update status as 02.

    2. Load this golden image (.srec) file into new boards

    3. Send IT Enable (In this case what should be update status ?)

    4. Send Seal Command.

    5. Convert to HDQ

    Thanks

    Vidya Sagar

  • Hi Vidya,

    These steps are close. You will need to convert to HDQ, switch your wires from the I2C to HDQ interface, and then seal the device with HDQ.

    Sincerely,
    Bryan Kahler