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.

bq40z50 high current application

Other Parts Discussed in Thread: BQ40Z50, BQSTUDIO

I'm attempting to use a bq40z50 in a high current application (2-3C discharge). I did the IT calibration as described, with a C/5 discharge, and the gauge seems to be tracking capacity well at low rates of charge/discharge. However, when I try to discharge at 3C, the 'true' capacity measurement drops mid way through discharge (from say 70% to 0%) even though the battery still has approx 70% charge remaining. The 'float' capacity measurement moves more slowly.

Is there a configuration I should set to make the gauge perform better? Should I do the IT calibration process differently?

Thanks,

Ben

  • Ben, 

    There may be some benefit to updating the Ra table at the higher discharge rate. Will you always discharge at the 2 to 3C rate or will you discharge at lower rates as well. We need to review you Load Select setting as IT uses this for FCC simulations during discharge. Do you have log files and gg files that you can share? I can help to analyze these an offer a better recommendation. Also, how did you verify your ChemID? Did we characterize cells, was is a Mathcad match or just a selection table match? 

    Tom

  • Hi Tom,

    This is my (pre calibration) gg file:

    6443.hamilton2.gg.csv.zip

    Yes, we only care about discharges at the 2-3C rate. It's possible some discharges will be much slower (C/10) but we don't care about accuracy in those cases. It's going to be a constant power load.

    The batteries I'm using are Samsung INR18650-25R (http://congrevape.com/wp-content/uploads/2014/04/131101_Introduction-of-INR18650-25R.pdf). They aren't listed in the selection table. I'm using chem id 206 (SDI ICR18650-24F) which is also NCA chemistry and therefore I believe it to be similar... although I haven't actually done a Mathcad matching.

    One other thing- because the current may at times exceed 30A, I'm calibrating the current shunt at 0.5x the actual current (bq40z50 thinks the current is 2A, it's actually 4A). Correspondingly I am halving the capacity and current limits in the registers (every time there is a unit of mA)- so I set max capacity to 7500mAh/2=3750mAh etc. I know this is unorthodox but I don't believe it should impact any of the IT calculations.

    Thanks,

    Ben

  • What are the triggers to set RSOC to zero? Do you think it might be that batteries are sagging and some voltage trigger is happening? I think all my limits are 2.9V or below and it definitely isn't sagging to that level, but perhaps I'm missing something.

  • Ben, 

    Your Termination voltage is set to 11600mV, so reaching this threshold would force the SOC to 0. The wrong ChemID or Ra table could force it as well. Your Load Select is set to 3, so this will use current and voltage values from the present discharge in the FCC simulations. You may need to use User Rate and set the User Rate value to your high discharge rate to improve the accuracy. I would verify the ChemID first. If you can provide discharge data, then I can check it with Mathcad. (You can use the attached document for reference in collecting the data.) You can also provide a log file for a max C rate discharge and I can calculate that the Ra table values need to be for the best performance. 

    Tom

    Chemistry selection.pdf
  • Hi Thomas,

    I was able to calibrate at a high C discharge rate and make a golden image. I used the image on an almost new pack and took the attached discharge log. The capacity remaining was much more accurate in this discharge than it was when I had calibrated at C/5, although there are small jumps. (with the c/5 calibration the jumps were at approximately at the same times, but went to 0% remaining capacity and then back up later)

    0714.discharge.zip

    Notes on the discharge log:

    - We seem to have some kind of thermal problem that is affecting current measurement accuracy. The true current for the entire discharge was 17A, or 8500mA after the division by two that I discussed earlier. In the recording it starts out measuring around 8500mA but increases as it gets hotter.

    - I didn't do the charge to C/100 as described in the PDF you sent because the pack I had available was already charged- let me know if you need me to repeat with the charge step

    - The computer went to suspend so there's some data missing in the middle of the rest state (again, let me know if I should repeat)

    I also included data flash values from before and after the discharge.

    Would it be useful for me to do a C/5 discharge? It seems to me that the Ra table values must be quite different, so a high C discharge is more relevant.

    If the battery is discharged at a low C rate, will it update with new Ra table values that would break accuracy during high C rates? What about if I set the [Ra Filter] register to something low, like 5%? Can I set it to 0% / what impact would that have?

    Can you describe how User Rate works in a bit more detail?

    Thanks so much for your help! Please let me know what discharges I can do that would be most useful for you.

    Ben

  • Ben

    I am sorry about the delay in getting back to you. I checked the data and Mathcad predicts ChemID 246 to be a better match than ChemID 206. If you have continued testing with ChemID 206 and it is performing well, then you can stick with it. I attached the Ra table and Qmax updates to match ChemID 246. You can update the DF with bqStudio.

    The gauge will used a current for FCC simulations and the Load Select setting will allow you several choices to choose from. The User Rate will allow you to fix the current to a specific value for all of the simulations. This could be a good selection, if the current never changes. Otherwise, it would be better to choose a selection that updates as your load current updates.

    Tom

    ChemID 246 Ra Qmax.xlsx