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-R2: Programming new PCBs with the Golden Image

Part Number: BQ34Z100-R2
Other Parts Discussed in Thread: BQ34Z100-G1

Tool/software:

Hello,

I am using the BQ34Z100-R2 fuel gauge on a custom PCB with two 12V, 22Ah lead-acid batteries in series. I have successfully calibrated the fuel gauge with my battery pack, completed a full learning cycle, and generated a golden image, which I have verified to provide accurate State of Charge (SoC) readings.

I am now preparing to integrate new batches of PCBs with the golden image and would appreciate clarification on the following process:

  1. For each new PCB, I understand that I need to pair it with its respective unique set of batteries, flash the golden image, and recalibrate voltage, current, and temperature if necessary. Once IT is enabled, will the gauge be ready to provide accurate SoC, or does the fuel gauge need additional learning from the new batteries? Specifically, how does the gauge determine the initial SoC immediately after flashing the golden image on a blank fuel gauge with a a set of batteries?? 

  2. I have noticed that when I flash the golden image onto a new PCB with new batteries at different SoC's, the SoC often jumps to 100%? Is this expected behavior? Should I perform a full charge and discharge cycle for the gauge to accurately learn the newly mounted battery’s SoC, or is there a more efficient approach?

  3. If the batteries connected to each PCB are at varying SoC levels when first installed, would this cause any issues with gauging accuracy? Would it be best practice to ensure all batteries are fully charged before flashing the golden image to avoid initial SoC discrepancies?

I would appreciate any guidance or recommendations on how best to manage this process for easy programming of the golden image gauging across multiple PCB batches.

Thank you for your support. 

Br, 

Mathias

  • Hi Mathias,

    1 You should create your golden image with new battery cells, so once the golden image is created it should be ready to gauge new cells. After the second ocv reading the gauge should report accurate soc. No additional learning cycle are required.

    2 After the second ocv reading does this get corrected?

    3 I recommend using fully charge cells but after the second ocv reading the gauge should correct soc.

    Regards,

    Diego

  • Hi Diego,

    Thank you for the fast responsive. 

    I've tested and I am facing still discrepancies in the Full Charge Capacity (FCC) readings. Despite expecting an FCC close to 22,000mAh, I am only seeing 4,700mAh for FCC, True RC, and True FCC when flashing the golden image. However, the Qmax for Cell 0 seems accurate, reading 20,303mAh, which is close to the expected 22Ah under ideal conditions.

    I am working with a BQ34Z100-R2 IC for my battery management system, which is connected to two REC22-12I Yuasa 12V 22Ah lead-acid batteries in series, giving me a total of 24V and 22Ah.

    I have some concerns about my understanding of the system and would greatly appreciate your time in helping me clarify the ambiguities and conflicting information I’ve encountered. Specifically, I’ve noticed discrepancies between responses on forums, the technical reference datasheets, and application notes. Additionally, there seems to be inconsistency in the definitions and procedures between the BQ34Z100-R2 and BQ34Z100-G1.

    In some instances, one source advocates for a particular procedure, while another suggests a different approach, and to complicate matters further, some videos claim that certain parameters are not even critical. I’m feeling quite lost and could really use some final and conclusive guidance in order to resolve these contradictions.

    Specific Concerns:

    1. Learning Cycle Validity:
      I am unsure if something went wrong during the learning cycle and am currently repeating it. I would like validation if the learning cycle might be contributing to these first-mentioned discrepancies.

    2. Parameter Clarification:
      I need clarification on whether the parameters I have set for the BQ34Z100-R2 IC are correct for my application. I’ve gone through the Technical Reference Manuals, various support resources from Texas Instruments, videos, application notes, technical support, but there seems to be conflicting information on how to interpret pack vs. cell definitions and corresponding values.

    Application Details:

    • IC Used: BQ34Z100-R2 on custom PCB
    • Battery Configuration:
      • Two 12V, 22Ah lead-acid batteries (Yuasa 1REC22-12I) connected in series.
      • Each battery contains 6 individual cells, making a total of 12 cells in the system.

    Clarification Needed on the Following:

    1. Definition of Pack and Cell (in the context of the BQ34Z100-R2 and not the G1 variant):

      • Pack: Should I consider the two batteries in series as the "pack" (24V, 22Ah)?
      • Cell: Should I be treating the characteristics of a single cell as 2V, 1,833mAh (22,000mAh/12) or 2V3,666mAh (22,000mAh/6)?
    2. Cell-Based Parameters: Below are parameters I have identified to be based on an individual cell. Is this correct?

      • Cell Charge Voltage:
        • T1-T2: 2308mV (Cell-based)
        • T2-T3: 2308mV (Cell-based)
        • T3-T4: 2308mV (Cell-based)
      • SOC Thresholds:
        • SOC1 Set: 1905mAh (Cell-based)
        • SOC1 Clear: 2000mAh (Cell-based)
        • SOCF Set: 1850mAh (Cell-based)
        • SOCF Clear: 1900mAh (Cell-based)
      • Voltage Thresholds:
        • Cell BL Set: 1950mV (Cell-based)
        • Cell BL Clear: 2000mV (Cell-based)
        • Cell ELL BH Set: 2420mV (Cell-based)
        • Cell BH Clear: 2300mV (Cell-based)
      • Cell Voltage at Charge Termination: 2315mV (Cell-based)
      • Number of series cells: 12
    3. Pack-Based Parameters: Below are parameters I have identified to be based on the entire pack (24V, 22Ah) Is this correct?

      • Design Energy: Is this parameter based on the entire pack or a single cell, and how is it calculated? (Pack-based)
      • Design Voltage: I have set this to 24V, corresponding to the two lead-acid batteries in series. (Pack-based)
      • Design Capacity: Set to 22Ah for the series configuration of the two batteries. (Pack-based)
    4. Voltage over 5V: As my Design Voltage is beyond the 5000mV max. value, what would have to be done and what above mentioned parameters will be affected ?

    I've attached the gg.csv file as well. I apologize for the long follow-up question. 

    Br, 

    Mathias lead_acid_config.gg.csv

  • Hi Mathias,

    1)

    Yes, the two batteries in series are your pack, 12 cells in series.

    Please divide the total capacity of your pack in mAh by the the 12 cells in series contained within your pack.

    2)

    Cell Charge Voltage

    these most likely don't need to be adjusted, these are based on a singe cell.

    SOC Thresholds

    these are based on the pack, these are used to set flags

    Voltage Thresholds

    These are also thresholds for flags, these are based on a single cell

    Cell Voltage at Charge Termination

    Are you referring to cell taper voltage?

    Number of series cells

    12

    3)

    Design Energy is based on the entire pack.

    Design voltage and capacity are correct.

    4)

    Please use an external voltage divider and configure VOLTSEL.

    Configure Voltage divider parameter, the ratio you are using.

    If scaling is needed, please scale parameters.

    (+) [FAQ] BQ34Z100-G1: BQ34Z100-G1 FAQs - Power management forum - Power management - TI E2E support forums

    Regards,

    Diego

  • Hi Diego,

    Thank you for your previous response, it helped clarify some points.

    I’m using an external voltage divider with the VOLTSEL properly configured, and everything is functioning correctly in terms of the schematic.

    However, I still have some concerns. 

    1.  Design Energy and Design Voltage settings.

    2. Currently, I’m seeing a mismatch between the Remaining Capacity, Full Charge Capacity, True RC, and True FCC, all showing 4415 mAh, which is significantly lower than the expected 22000 mAh.

    Interestingly, the Qmax Cell 0 value seems accurate at around 20582 mAh. This mismatch persists even after completing a learning cycle. Despite the discrepancy, the SoC appears accurate, and the state of charge changes as expected during charging and discharging.

    Could you help identify what might be causing this capacity mismatch?

    I've attached the screenshots for the parameters mismatch: 

     (Qmax Cell 0)

     (RC, FCC, True RC and TRUE FCC). 

    Thank you for your assistance.

    Best regards,

    Mathias 

  • Hi Mathias,

    1

    With R2 FW, there is separate scaling for voltage, current, and energy.

    You are correct, design voltage and energy should be for a single cell, this is my mistake 

    2

    Are you doing any current scaling?

    If you have a log file, please share that as well.

    Regards,

    Diego

  • Hi Diego, 

    Thank you for the clarification. Should the Design Capacity be considered for a single cell as well or does that remain as pack-based? 

    I don't have the latest log file, but will run through a new learning cycle again - just want to make sure that all the parameters are correct as a starting point as the learning cycles are very time-consuming. 

    I am not doing any current scaling.However, I've been running the learning cycle with the Design Capacity set to cell-based and a value of 3666 mAh. I assume this would be incorrect for my setup of Two 12V, 22Ah lead-acid batteries (Yuasa 1REC22-12I) connected in series

    I've attached the GPC report that I've received after doing a custom chemID identification from which I've chosen the ChemID_814. Please let me know, if this would be a mismatch or if there isa better chemID from the report.

    In addition, I've attached the config with which I ran my learning cycle. 

    I appreciate your continued support. 

    first_learning_cycle_config.gg.csv

    Best regards, 

    Mathias 

  • Hi Mathias,

    I don't see the GPC report, maybe you forgot to attach, it would be good to verify, it will hard if not impossible to complete a learning cycle without a good match.

    Design Capacity is technically for the pack but adding cells in series will not increase capacity only adding them in parallel.

    Therefore, multiple the nominal capacity of a single cell by the number of cells in parallel.

    Regards,

    Diego

  • Hi Diego, 

    You're right - my mistake. I've attached it here. Currently running with ChemID814.   

    2555.GPC_report.txt
    0	0	0
    0	0	0
    0	0	0
    0	0	0
    0	0	0
    0	0	0
    0	0	0
    0	0	0
    0	0	0
    0	0	0
    0	0	0
    0	0	0
    0		
    		
    Summary of all IDs with max. DOD deviation below 15%		
    		
    Chem ID	max DOD error, %	Max R deviation, ratio
    807	2.51	17.78
    814	2.75	-3.66
    803	3.91	0.38
    801	3.99	0.85
    813	4.47	0.46
    808	5.01	0.02
    802	5.63	0.26
    812	7.28	17.58
    805	8.94	-10.5
    800	11.91	0.89
    809	12.77	3.86
    		
    Max. deviations for best ID is within recommended range. Chosen best chemical ID is suitable for programming the gauge.		
    		
    		
    Selection of best generic ID for ROM based devices like bq274xx		
    		
    		
    Device / Family #1		
    Generic Chem ID	Device/ Voltage/ Chemistry	max DOD error, %
    128	bq27421-G1A: 4.2V LiCoO2	100
    3142	bq27421-G1D: 4.4V LiCoO2	195.35
    354	bq27411-G1C: 4.35V LiCoO2	197.79
    312	bq27421-G1B: 4.3V LiCoO2	200
    Best generic ID 128		
    Warning: Generic ID Deviation is so high that it is most likely due to anomaly in the data. Please check that data files have recomended format, units and test schedule		
    		
    		
    Device / Family #2		
    Generic Chem ID	Device/ Voltage/ Chemistry	max DOD error, %
    1210	bq27621:  (ALT_CHEM1) 4.3V LiCoO2	100
    354	bq27621:  (ALT_CHEM2) 4.35V LiCoO2	197.79
    1202	bq27621: (default) 4.2V LiCoO2	200
    Best generic ID 1210		
    Warning: Generic ID Deviation is so high that it is most likely due to anomaly in the data. Please check that data files have recomended format, units and test schedule		
    		
    		
    Device / Family #3		
    Generic Chem ID	Device/ Voltage/ Chemistry	max DOD error, %
    3230	bq27426: (default) 4.35V LiCoO2	100
    3142	bq27426: (ALT-CHEM2) 4.4V LiCoO2	195.35
    1202	bq27426: (ALT_CHEM1) 4.2V LiCoO2	200
    Best generic ID 3230		
    Warning: Generic ID Deviation is so high that it is most likely due to anomaly in the data. Please check that data files have recomended format, units and test schedule		
    		
    		
    

    That makes sense for the design capacity and I suspect that the wrong parameters I've run the first learning cycle with could be the cause of the mismatch in read values. I'll run a new learning cycle with the updated parameters and get back and hopefully close and resolve the thread. 

    Best regards, 

    Mathias

  • Hi Mathias,

    ChemID match looks good! assuming your input data is ok.

    Please double check all critical parameters, critical parameters can be found in section 2.1 here: Achieving The Successful Learning Cycle (tij.co.jp)

    Please log and share the new data when available.

    Regards,

    Diego