• State TI Thinks Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 3 answers
  • Subscribers 63 subscribers
  • Views 497 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

BQ40Z60: Cell Grid Tracking

Asfan Fiirizky
Prodigy 20 points
Part Number: BQ40Z60
Other Parts Discussed in Thread: GPCCHEM

Tool/software:

Hi everyone, 

So, we have a lifetime test for our battery pack, which cycling (charge and discharge) the battery using a randomize RSOC as a target.

At the first time cycle after learning cycle procedure, the upper target RSOC was 95% and lower target was 15%.
The cell grid was updated along with the RSOC decrement, which indicates that the gauge tracks the DOD and report us through the cell grid (1 - 13) after it passed 11.1% (before knee region).
This can be describe by figure 1.

Figure 1. RSOC vs Cell Grid at cycle count 1.

However, after several cycle further, with large RSOC difference from 90 to 40%, the cell grid was not updating.
The cell grid updated after transition from discharge to charging state.
This can be described on figure 2.

Figure 2. RSOC vs Cell Grid at cycle count 4.

Is there any explanation, how this case happened? 

I just attached 3 files, which consist of learning cycle data, data at cycle count of 1, and data at cycle count of 4.
Including the data memory files. 

2025-04-17_15.07.02-FirstCycleDataLog-DischargeUntilTermV.csv2025-04-21_05.47.19-QMAX_Docking_1_Batt_0x1c.csv2025-04-21_19.17.30-QMAX_Docking_14_Batt_0x1c.csv2025-04-21_05.47.19-Cycle-1-DataMem.csv2025-04-21_19.17.30-Cycle-14-DataMem.csv

  • 0
    TI__Mastermind 26595 points

    Hello Asfan,

    Could you attached your GPCCHEM tool report for your chemID, it should've been a zip file that you received in email from the tool that helped you picked your chemID.

    GPCCHEM Tool:
    https://www.ti.com/tool/GPCCHEM

    Thank you,
    Alan

  • 0 in reply to Alan Frias
    Prodigy 20 points

    Hi Alan,

    Apologize, I've not uploading my battery data to GPPCHEM yet, soon we will upload it.

    However, based on your answer, there's possibility unmatch chemistry between my battery cell with an OCV that were used as a reference for the gauge.
    I can answer it, yes it's different. 

    A long time ago, we're using a random battery cell from marketplace, they assuming that their product was authentic which is Sony US18650VTC6 3000 mAh (Chem id 2141). However, we found an unbalance issue during cycling procedure. Hence, we decide not using that battery again. 

    Now we're moving on, to bought a battery cell from authorized distributor from Molicell, and the battery is INR-18650-P28A (Chem id 5271).
    However, our gauge still using the old chem id 2141. Based on our understanding of your impedance track algorithm, as long as the battery chemical was the same (NMC), It's safe to use the old chem id (2141), and also the battery that we're used has lower QMAX capacity (2700 mAh) than the OCV (3000 mAh). 

    Therefore, the gauge should be able to track where the gauge position (DOD) along discharge period. 
    Is my understanding correct?

  • 0 in reply to Asfan Fiirizky
    TI__Mastermind 26595 points

    Hello Asfan,

    Asfan Fiirizky said:
    long as the battery chemical was the same (NMC), It's safe to use the old chem id (2141)

    This statement is wrong and the battery gauge would act incorrectly with all measurements. Since you are updating the chemID you would also need to update your parameters within data memory tab and redo a calibration in the calibration tab to be safe. You would also need to rerun the learning cycle to get an updated qmax. Once you complete these steps you should be getting accurate measurements. I've attached the EVM guide that talks about upbrining the gauge, document for learning cycle and the GPCCHEM calculator to pick your chemID.

    EVM:
    https://www.ti.com/lit/ug/sluub71/sluub71.pdf?ts=1745422355635&ref_url=https%253A%252F%252Fwww.ti.com%252Ftool%252FBQ40Z60EVM-578

    Learning Cycle (Same concept but different gauge):
    https://www.ti.com/lit/an/slua848/slua848.pdf?ts=1745422267940&ref_url=https%253A%252F%252Fwww.google.com%252F

    GPCCHEM calculator:
    https://www.ti.com/tool/GPCCHEM

    Thank you,
    Alan 

  • 0 in reply to Alan Frias
    Prodigy 20 points

    Hi Alan,

    First of all, I do appreciate for your answer. 

    Apologize, if my understanding wrong. However, could you help me understand it properly about the gauge and in parallel I'll follow your suggestion to use a suitable Chem id for my battery cell.

    As you can see on first learning cycle file, the grid point updated from 1 until 10 (knee region), which indicating that the gauge was able to track the DOD and also updating the resistance, even the OCV reference is 2141/3000mAh (where my battery is 5271/2700mAh). 

    Would that be make sense if we refer to one of OCV curve at certain temperature, which all I knew that the OCV is function of T (Temperature). If you said that my statement is wrong, I can accept that, if the OCV curve has a lot different curve at variant temperature. Is that right? 

    Thus makes the gauge has different resistance measurement with the last previous cycle that using another T (Temperature). But, the updated resistance values are always on positive sign since the OCV has higher voltage than the underload voltage. Is that correct? 

    The last one, the 100% of my battery DOD (5271) is way lower than 100% of 2141 DOD. Because, the present battery capacity way lower than my old battery. So, the gauge should be able track its DOD, and it's proven by the first learning cycle data. Is that correct? 

    Thanks for your help. 
    Asfan.

  • 0 in reply to Asfan Fiirizky
    TI__Mastermind 26595 points

    Hello Asfan,

    Asfan Fiirizky said:
    OCV curve has a lot different curve at variant temperature. Is that right? 

    Yes this is correct the OCV curve can vary based on your temperature of your cell and these OCV curves are often given to you by the cell manufacture.

    Asfan Fiirizky said:
    So, the gauge should be able track its DOD, and it's proven by the first learning cycle data. Is that correct? 

    This is not necessarily correct due it being able to track and update the DOD after a charge, discharge rest cycle however having the wrong chemID would never give you the accurate results or never update enough after multiple cycles to show true state of curve.

    Thank you,
    Alan

  • 0 in reply to Alan Frias
    Prodigy 20 points

    Hi Alan,

    Alright then, let me follow your suggestion. 

    If I choose a correct chem id based on GPCCHEM tool, it should be the cell grid are always updating during dicharge period right?
    Because the only way for us as a user to monitor whether the gauge is able to track our battery is using the cell grid.

    Thank you,
    Asfan.

  • 0 in reply to Asfan Fiirizky
    TI__Mastermind 26595 points

    Hello Asfan,

    SOC updates periodically depending on charge, discharge or rest events, I am not sure what the cell grid are meant to represent low level since the axis aren't label. However they should update and SOC should be reflected correctly when your chemID is updated, without this being updated it is really difficuilt for the battery gauge to update correctly. Also if you are interested in the each cell groups voltages there is a register for that you can view in the .log aswell. 

    Thank you,
    Alan