• State
  • Locked Locked
  • Replies 16 replies
  • Subscribers 64 subscribers
  • Views 1669 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.

BQ34Z100-G1: State Of Charge discontinuity

Mirari Elortegi
Prodigy 150 points
Part Number: BQ34Z100-G1
Other Parts Discussed in Thread: BQSTUDIO, , GPCRB

Hi,

We have some golden images for this gas gauge depending on the used battery (we have 3Ah and 6Ah batteries). We are now using them on some protoptypes. We have noticed a strange behaviour on the state of charge measurement. There are some discontinuities. For example, after using the battery for some operation our prototype had 95% of SOC. Suddendly, it got down to 79%. And then, with no input current, it went up very slowly (it took 6 hours) to 95% again. Other times, the discontinuity is also in when it goes up again. Suddendly we see a 65% value, and then a 90% value. We see that the Full Capacity is also modified in these discontiniuties. But we don't see the reason. We started to think that it could be a temeprature issue, but after registering temperature we saw that sometimes it looked like the full capacity varies with temperature, but sometimes not. So we don't know what could be the reasons to these variations on full capacity, and therefore in state of charge.

We don't do any "Reset" operation to the gas gauge. There could be a "unseal" operation to modify a parameter and then a "Seal" operation when finished. Could this affect?

Could it be possible to have a "bad" golden image?

  • 0
    TI__Guru 73830 points

    Mirari,

    The gauge does run capacity simulations during the rest period and it can adjust the RM and FCC as the OCV changes. Temperature changes can also trigger capacity simulations. You did not mention which type of cells are used in you pack, but the changes that you describe are common with PbA cells. If you can send a log fie and an srec from your pack, then we may be able to offer other suggestions.
    Tom

  • 0 in reply to ThomasCosby
    Prodigy 150 points

    Hi!

    They are LiFePo4 batteries. Here you have a log file we got from a 3Ah battery pack, and our golden image files for the 3Ah model.S16003#01_R4 (GoldenImage).rarLog3Ah.txt

  • 0 in reply to Mirari Elortegi
    Prodigy 150 points

    And are these capacity simulations necessary? Is there any way to disable them? Or at least is there any way to know that it is running a simulation?

    Our prototype takes some decisions based on the state of charge, and these virtual values are conditioning the functionality...

  • 0 in reply to ThomasCosby
    Prodigy 150 points
    Hi Thomas!

    Did you manage to have a look at the files I attached? We are now thinking of a way to disable that posible capacity simulation...

    Thank you!!!
  • 0 in reply to Mirari Elortegi
    TI__Guru 73830 points
    Mirari,
    The log file does not provide an answer as to why the SOC is dropping during the discharges. Is the SOC_FLAGS register the FLAGS register in the bq34z100-G1? Could you run the profile while logging with the bqStudio program to log additional parameters that could help with the analysis?
    Tom
  • 0 in reply to ThomasCosby
    Prodigy 150 points

    Hi,

    Yes, it is the FLAGS register in the bq34z100-g1. For the moment we can't run the profile while logging with the bqStudio... Do you think it could be a capacity simulation as you suggested before?

  • 0 in reply to Mirari Elortegi
    TI__Guru 73830 points

    The SOC drops is due to a simulation. The gauge also runs capacity simulations at the Ra grid points. These simulations cannot be turned off in Impedance Track gauges. You could use a CEDV gauge, since these gauges do not have capacity simulations at grid points or during rest. The bq 34110 would be the equivalent CEDV device, but it has a different pin out.

  • 0 in reply to ThomasCosby
    Prodigy 150 points

    Hi Thomas,

    Thank you for your suggestion. Unfortunately we can't modify the gas gauge... We think that maybe we are doing something wrong. Not only the SOC value changes, the FCC changes too. Suddendly it drops to the half capacity as the usual. It only happens after some working cycles. In our application we are trying to use the batteries in the 40-60% range, so that's a cycle for us. We have been thinking various possibilities:

    1. We think that we have to let the batteries charge to 100% periodically, but for our laboratory test we are not doing it. Could this be the reason for the FCC variations?

    2. We also see that sometimes the FCC variation is related to negative battery temperature measurements, but not always. Could it be related to temperature?

    3. We saw a parameters called Qmax Change. Could we limit the FCC variation using this parameter?

    4. Could the FCC variation be related to OCV measurements that are taken before the battery is really relaxed? Could we limit it using DSG Relax Time, Chg Relax Time...

    5.  We have StateOfCharge() Smoothing activated. Is this necessary?

    I know that some ideas could not be related to our "problem" but we are just brainstorming...

    Thank you!!!

  • 0 in reply to Mirari Elortegi
    Prodigy 150 points

    Hi again Thomas!

    We have some more information about the moments that we see that the SOC jumps. In this case the Full Charge Capacity does not change. Just the SOC. For example we have seen that being at 100%, changes to 29%, after being discharging for 3 minutes at 8 Amperes. Our batteries are 6Ah. When we did the golden images we did them discharging at around C/5. Could the problem be that this 8A load condition can disturb the estimated SOC value? Do we have to do the golden images with more load?

  • 0 in reply to Mirari Elortegi
    TI__Guru 73830 points
    Mirari,
    What does the gauging accuracy look like when you charge the pack to full, rest and then discharge it to empty? This will help to determine whether the optimization cycle results were good. The gauge should be able to compensate for the variations in current as long as your cells do not have some typo of abnormal non-linearity in the series resistance.

    Tom
  • 0 in reply to ThomasCosby
    Prodigy 150 points

    Hi Thomas,

    The optimization cycle looks good.

    Can we get phone support from Texas Instruments?

    We tried to get support from your distributor Avnet Silica in Spain but received a negative answer from your side, as this thread was already open. We need to appoint a conference call between P4Q, Silica and Texas Instruments. This issue has already taken too long... and we see that this post method is not solving the issue (State Of Charge value discontinuities and jumps).

    Please confirm a date for the conference.

    Best regards,

    P4Q Team.

  • 0 in reply to Mirari Elortegi
    TI__Guru 73830 points
    Mirari,
    We will not be able to support a call at this time. It is good that gauging is good with the continuous discharge cycling. I reviewed your log file again and found that you are operating at extremely low temperatures and an RB tweak will probably be required to adjust the resistance tables to operate at these temperatures. You can use our online tool to run the Rb tweak. The tool can be found at this location.

    www.ti.com/.../GPCRB

    Tom
  • 0 in reply to ThomasCosby
    Prodigy 130 points

    Hello Thomas,

    Which of the Texas-provided documents should we follow to proceed with the suggested test? Is it the SLUUB45A (Simple guide to CEDV Data Collection for GPC) or is it SLUUBD0 (Golden GG Maker and Resistance Temperature Compensation Optimizer)? Both can be found in data collecting instruction manuals.

  • 0 in reply to IKER JOARISTI URTUSAGASTI
    TI__Guru 73830 points
    This is the tool that you would use for an Rb tweak.

    www.ti.com/.../GPCRB
  • 0 in reply to ThomasCosby
    Prodigy 130 points

    Hello Thomas,

    On THAT link there are two more links (on step 1) to downloadable documents explaining the procedure for data logging to later update a zip file to your servers (step2) for postpreccesing.

    The fact is that on the one hand, document SLUUB45A requires 7 files:

    1) Configuration file with: NumCellSeries, CellTermV, ChemType, VoltageColum, CurrentColum, TemepratureColum,ElpasedTimeColum, FitMaxSoC, FitMinSoC, LearnSoC

    2) Data logs of 6 experiments (6 files): hightemp_highrate, hightemp_lowrate, roomtemp_hihghrate, roomtemp_lowrate, lowtemp_highrate, lowtemp_lowrate

    On the other hand, document SLUUBD0 requires 4 files:

    1) Configuration file with: ProcessingType, ChemID, NumCellSeries, VoltageColum,CurrentColum,TemperatureColum,ElapsedTimeColum

    2) Data logs of 2 experiments (2 files): roomtemp, lowtemp

    3) gg.csv file

    The question is which of the documents is the right one for GPC optimization.

    Best regards

  • 0 in reply to IKER JOARISTI URTUSAGASTI
    TI__Guru 73830 points

    Iker

    The SLUUBD0 covers the data and steps to use the GPCRB tool to run the Rb tweak for improve low temperature performance.

    Tom