• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 63 subscribers
  • Views 661 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.

BQ20Z45-R1: RSoC Error

Paul Antunes
Intellectual 2590 points
Part Number: BQ20Z45-R1

Hi,

pls find below customer issue using bq20z45-r1 in production:

BQ20z45-R1 fuel gauge implemented in 2s4p batteries (just after production) send RSoC value with +5% to +8% error after they are partially charged (from 30% to 50%).

Partial charging occurs just after production, batteries have never been fully charged but they are storaged several weeks.

RSoC error have been estimated comparing RSoC on SBS data sent by the fuel gauge and voltage characteristic (Voc = f(SoC)) of the cell used.

RM and FCC updates after Impedance Track launch (exit of relaxation mode or 5h after a short discharge) give no improvement.

A full charge or full discharge removes RSoC error.

 

Is this RSoC error is due to the partial charge of batteries that have never been fully charged ?

Parameters in dataflash called Cell 0..3 Chg DOD at EOC are at 0 on batteries that have never been fully charged, can we simulate a fully charge by increasing this value >0 ?

thanks in advance for your support.

Paul

  • 0
    TI__Genius 14985 points
    Hi Paul,
    Are they using voltage measurements taken after several hours of no current flow to estimate the SOC? Even if the answer is yes, RSOC cannot be compared to voltage data directly. RSOC meand "relative" state of charge and can depend on many factors including load and temperature.
    I would also like to confirm they did the basic configuration properly:
    How did they determine the correct chemID to use for this cell?
    Did they perform a learning cycle to optimize Qmax and the Ra tables?
    Can you please share .gg files from before they charge and after they charge to full?
    DOD at EOC is learned after a full charge and will affect RM, FCC, and SOC slightly. After they charge to full and this value is updated, it is a good idea to use the non-zero value in the golden file.
  • 0 in reply to dMax
    TI__Intellectual 2590 points

    Hi David,

    Hereafter customer answers:

    1- Are they using voltage measurements taken after several hours of no current flow to estimate the SOC?

    Yes, voltage measurement is taken 5h minimum after entering in relaxation mode.

    2- Even if the answer is yes, RSOC cannot be compared to voltage data directly. RSOC meand "relative" state of charge and can depend on many factors including load and temperature.

    Yes, we know. But we do not check a precision <5%. This control is performed in storage mode only. Temperature is always at +20°C and without load. We use a Voc=f(SoC) model measured with a standard battery. This solution works with most of batteries and we have error ≤2%.

    3- I would also like to confirm they did the basic configuration properly: How did they determine the correct chemID to use for this cell?

    We have determined ChemID from your table called “chemistry_selection_table.xls”. We use NCR18650 (2900mA.h) cells from Panasonic and the ChemID is 0x0238

    4- Did they perform a learning cycle to optimize Qmax and the Ra tables?

    Yes, a learning cycle has been performed with standard battery to get a golden image. Then this golden image is saved to all battery packs manufactured.

    5- Can you please share .gg files from before they charge and after they charge to full?

    I did not generate these files yet. I will try to give you these files asap.

    6- DOD at EOC is learned after a full charge and will affect RM, FCC, and SOC slightly. After they charge to full and this value is updated, it is a good idea to use the non-zero value in the golden file.

    OK, thank you. That is a good information. In our golden image these value are at 0. We can try to use fully charging values.

    regards,

    Paul