• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 3 answers
  • Subscribers 62 subscribers
  • Views 1604 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.

BQ40Z50-R2: AOLD, SOH, Recovery-Reset time configuration questions

Ivan Tokarev
Prodigy 30 points
Part Number: BQ40Z50-R2
Other Parts Discussed in Thread: BQSTUDIO, BQ40Z50, EV2400

Hello. My company has battery pack product which uses bq40z50 with R2 2.08 firmware inside. I am using EV2400 and BqStudio to analyze and program.

  1. Some battery packs have only AOLDL bit HIGH. We have Latch Limit = 0 and NR = 0 settings. AOLD current protection threshold is 100 mV / 3 mOhm = 33 Amps.  I actually don't understand why this hardware based protection triggered if OCD1,2 didn't. I tried draw 20 Amps from battery with my custom setup and OCD2 triggers and AOLD doesnt. Could you give me an idea why AOLD could trigger? 
  2. Here I tried R3 firmware. Why the DSG FET off time isn't stable when Recovery time = 0, Reset time = 0, NR bit = 1?  I measured time difference with oscilloscope on DSG FET gate and it was floating from 0.3 sec to 1.5 sec. Can you give me explain about that situation? What can I do with power disabling in Trip stage?
  3. I know that SoH calculated with simulated FCC at 25 Celsius and fixed DC/SoH Load Rate current. But I dont understand when this simulated FCC will degrade? We have batteries with FCC at 70% but SoH 99-100%. I understand that FCC can decrease with temperature and dsg current but SoH is very high. SoH updates at the same time with RSOC and ASOC updates. How it actually calculates these SoH FCC and SoH Energy?

This is my bq40z50 settings R2_settings_forum.gg.csv

2241 Chemistry ID

  • 0
    TI__Guru** 105870 points

    Hello Ivan,

    The hardware and firmware based protections operate much differently.

    When you set the hardware protections (ALODL) it is essentially setting a comparator value for the 100mV along with the timer to trigger the protection it is all hardware. This means it can protect the battery very quickly.

    The firmware based protections operate on the ADC readings, if you set the value to 0 in firmware, we still have to take measurements the conversion period to determine when to trigger the protection, which is why you can see high variation. Since we use delta sigma ADC, the first conversions may not be over the firmware trip point yet, so we don't trigger the first 500ms for example (your worst case 1.5s) then on the next 1s conversion and firmware check we are, so we trigger the protection.

    SOH is influneced by the the Ra table, which increases over time, and the Qmax value which decreases over time, that's what we use to help determine the SOH. If you set your design capacity incorrectly it can also lead to errors in the SOH estimation, the R4 SOH calculation has a feature to help with SOH estimation.

    Sincerely,

    Wyatt Keller

  • 0 in reply to Wyatt Keller
    Prodigy 30 points

    Thanks for help. I would like to have a bit more explanation of your answer.

    1) Does current spike(more than my 33 Amps threshold on AOLD) trigger only AOLD because it's hardware and it works faster? I ask you because I have a bunch of battery packs that have only SafetyStatus()[AOLDL] triggered and others have only SafetyStatus()[OCD1,2]. Lifetime data sometimes updates about such high current, sometimes not. Does constant high current below 33 Amps will trigger OCD1,2? 

    2)That is, I cannot rely on fix time recover power capacability? Recover and Reset = 0 don't help me to save my load from power disabling. What can I do to prevent damaging my load due to immediately power disabling? If I have alert before such event(I understand that impossible) or have some power disable delay with alert, I will save my load.

    3) That SOH FCC and SOH Energy calculation algorithm pays attention to Ra table and Qmax. If I use my battery always between 20% and 80-100%(not full charge-discharge cycle), when will it update Ra table and Qmax value? 

    4)Question isn't about above your answer. Can you give explaination how is Max error(in %) actually calculated ?

  • 0 in reply to Ivan Tokarev
    TI__Guru** 105870 points

    Hello Ivan,

    Yes the hardware protections work much faster, all the protection settings are in the TRM. You can see the hardware protections are in the micro-second range and the firmware protections operate on the firmware refresh rate, minimum of 1s.

    I'm not sure I understand your question, the protection is designed to prevent damaged to the load is there is a short or other abnormally high current. The protections should only be triggered if something goes wrong to protect the battery and load. For the firmware based protections you will have an alert, for hardware you will not.

    Qmax needs a 37% change in DOD to update, you don't need 90% like the learning cycle. Ra values will update during normal discharges. There are other features like the fast Qmax that can also help you update Qmax with less relaxation periods.

    Max error is just incremented a specific amount per cycle, I believe the default is 0.05%. It's just used to help indicate when a Qmax or Ra update is needed. Some customers perform a conditioning cycle to reduce the max error.

    Sincerely,

    Wyatt Keller

  • 0 in reply to Wyatt Keller
    Prodigy 30 points

    Thanks again for your help. I appreciate your efforts to help me investigate bq40z50 theory.

    1)If I add Latch Limit > 0 to my AOLD, ASCD, ASCC settings and NR = 1(because PRES pin always ground and I cannot make low-high-low transition on it), it will help me recover battery packs from protection. For example if current spike hits, it will not trigger XDSG and block permanently my battery pack. For the first 0.5-1.5 seconds it disables power and when failure condition ends it will enable power. I am right?

    2)I was incorrect a bit. My load can be damaged because it can fall down Slight smile due to hardware protection power disable feature. That's why I want to know the power disable duration at Recovery and Reset = 0.

    3)If I discharge battery more than 37% Qmax will be updated? Does battery need a lot of time to RELAX after hard work(high current ~8-15 Amps) to learn something special or it can be easily placed in charger and I dont need care about such learning problems?

    4)Does high(5-10%) max error indicate that battery didn't update Qmax or Ra or I need to recalibrate battery? Can I use Max error to indicate that battery has degraded enough?

    5)Can you explain how cycles are calculated internally? I only read that cycle increases when battery capacity reaches FCC or DC(depends on CCT bit).

    6)Last question. Does sensor resistor's resistance(from 3mOhm to the lowest working resistance - 1mOhm) changing really help to prevent trigger hardware protection less often and not to afraid that battery can be damaged?

  • 0 in reply to Ivan Tokarev
    TI__Guru** 105870 points

    Hello Ivan,

    That's correct, the gauge will reset the latch counter after the latch reset time when NR = 1. The recovery time can be used to recover each individual trip condition, if they keep happening then we can disable the pack.

    If both are set to 0, there will be short periods of reset and recovery, if you want to disable the feature all together (which is essentially what setting both times to 0 would do) you can set the current very high in the hardware settings so it won't trip.

    The change must be 37% of the absolute scale on the chem ID DOD. Yes the gauge must see a relax period long enough to take on OCV. You can use Fast Qmax also, there is more information in the TRM about the update conditions.

    Yes high MaxError means there hasn't been a Ra or Qmax update for many cycles.

    That's correct, please see the TRM: 15.13.3 Cycle

    Since the hardware protections are based on the voltage across the sense resistor, decreasing the sense resistor will require higher currents to set the hardware protections.

    Sincerely,

    Wyatt Keller