• State TI Thinks Resolved
  • Locked Locked
  • Replies 11 replies
  • Answers 2 answers
  • Subscribers 63 subscribers
  • Views 332 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

BQ35100EVM-795: how to test EOS mode using BQ35100EVM-795,BQstudio,EV2400 for battery pack of 2 primary shaft battery(LS14500) in parallel configuration.

karan patel1
Expert 1085 points
Part Number: BQ35100EVM-795
Other Parts Discussed in Thread: BQ35100, BQSTUDIO, , EV2400

my original application behaviour is like:

our device working is like it wake up 5 times a day, wake up period is nearly 60 sec and during this time take measurement from sensor and send data on zigbee and go to sleep,active mode current is nearly 80 mA and sleep mode is 1.6 mA. we are using two new Li-socl2 non rechargeable battery in parallel configuration(reason is that continuous current of one cell is 50ma so we use 2 cell in parallel to counter this situation).cell name is LS 14500 shaft battery.we need notification for battery change so for that we planning to use EOS mode of BQ35100.

Things i have:

BQ35100EVM-795 , BQstudio,EV2400 , battery pack of 2 primary shaft battery(LS14500) in parallel configuration.

Planning is first i test behaviour using BQ35100EVM-795 and than integrate it in original product.

Activity done till now is summarised here:


-install latest version of bqstudio and driver for ev2400.

-cross check all jumper as mentioned in doccument(on J2 jumper is on 1-s side and J3 on 2s).

-use load of 49ohm between BAT+ and -PACK to draw 70 mA current.

-done voltage,temperature and current calibration.

-change mode to EOS and other setting as per my understanding in data memory.

now to understand behaviour of measuredZ and scaled R i follow procedure(send start gauge->apply load for some second->remove load->send stop gauge command->wait for data to be updated->wait for 60 sec or more before repeat cycle)

i tested this with 2 type of battery for at least 3 above cycle,and log both measuredZ and scaled R in log

1- fully new battery(here i send new battery command also)

2-partially used battery

i find that after first cycle in both case both resister value is same i think it should be different. here is log file and other files(srec and data memory)

0100_1_02-bq35100_with_cal_EOS_mode.srec.txt

bq35100_data_memory_with_cal_EOS.gg.csv.csv

5037.log.txt

So what is standard procedure to test and validate my requirement?

what should be standard way to test it?

i also don't get why in both case values are same is i am doing it wrong?

  • 0
    TI__Genius 10370 points

    Karan,

    Everything about your procedure look correct except it seems you are missing turning the gauge off. The gauge expects the device to be powered down between pulses to clear out the RAM is has accumulated. 

    This can be accomplished by setting the GE pin low or issuing a reset command. 

    Thanks,

    Eric Vos

  • 0 in reply to Vos
    Expert 1085 points

    Thank you for quick reply,

    but i have some more question,

    why in both case measuredZ and scaled R is showing same after first cycle? is it the due to same reason that i am not pulling low GE pin?

     one other thing i noticed is that every time i execute this cycle soh is decremented by fix value for ex 5%.I know that in eos mode soh reading will not be accurate but in this case if i execute cycle 20 times than soh becomes 0% which is not practical.i dont want precise reading for soh but even if i get appropriate reading than that data can be utilised also.

    Is there any document which provide good explanation for various terms under data memory tab (ex short trend and long trend pulses ) because there are many terms there?

  • 0 in reply to karan patel1
    TI__Genius 10370 points

    Karan,

    The value you get after you start the test and after sending the "new battery" command are most likely accurate since the RAM is cleared out. After that since RAM is not cleared your measuredZ and scaledR are calculating out to be 32K. This value would result in your SOH wanting to report 0%. However, we do not allow SOH to increase for user experience and we do not want to allow the value to go to 0 in one read. Therefore we limit the amount SOH can change for any single Gauge Stop event. By default this is 2, but it is open DF which might have been changed to 5 for your board. 

    Short/Long Trends are TI internal Filters to calculate when you have reached the End of Life. When these two values deviate by more than the threshold EOS is detected. There is no mapping for these filter since the filter duration "could" be changed (not recommended). 

    Thanks,

    Eric Vos

  • 0 in reply to Vos
    Expert 1085 points

    Thank you Eric for prompt reply,

    now i am sending restart command after every cycle and both values is working fine(mean its not 32k) in EOS mode.

    But in SOH mode with right chemistry id and same GE pin pulling high and low sequence still soh is decremented by fixed (state of health max delta %) value in each cycle?

    now should i changes this Short/Long Trends values or leave it as it is,because in reference manual its default value is 0 but here its showing 720452,720220?

    i modified data under gas gauging tab according to my need but  there is ra table is also there ,so is it  fetch from chemistry data automatically? or i should change and manually feed it?

    i am using 2 cell in parallel configuration so according to your suggestion in one separate thread that i need to scale down scaled resistance by 2x so,i am not getting how to do that?

    is chemistry id playing role in EOS mode or its for SOH only?

  • 0 in reply to Vos
    Expert 1085 points

    hi eric can you reply?

  • 0 in reply to karan patel1
    TI__Genius 10370 points

    Karan,

    1) I am not sure what you mean by restart (you should be sending reset or putting GE low) not sending the newbattery everytime

    2) SOH comes from impedance correlation with the ChemID (mostly From Ra table) 

    3) Ra table is updated with chemistry update. Also other internal gauging parameters

    4) You should not touch the filters. the gauge will auto update those (if you changed them i would start over with the new batt command)

    5) The gauge should learn the scale factor to adjust for the 2 parallel cell. If you wanted to cut the Ra table by 2 this would be a decent step to take

    6) EOS is Independent of chemID

  • 0 in reply to Vos
    Expert 1085 points

    1) I am not sure what you mean by restart (you should be sending reset or putting GE low) not sending the newbattery everytime

    restart  means sending soft restart command and i am not sending new battery command every time i send it once when connect new battery.

    5) The gauge should learn the scale factor to adjust for the 2 parallel cell. If you wanted to cut the Ra table by 2 this would be a decent step to take


    does this mean, i should replace the Ra table values with (current value/2) ?

    for soh i follow steps,

    set GE high -> wait 1s -> send gauge start() -> apply load and wait for 30 sec -> remove load -> send gauge stop command() -> wait for G_DONE set -> read SOH -> set GE low

    but still soh is decremented by  fix (soh delta % value) which is 2% on every cycle. chemistry id is correctly programmed.

  • 0 in reply to Vos
    Expert 1085 points

    In reply to Eric Vos24:

    1) I am not sure what you mean by restart (you should be sending reset or putting GE low) not sending the newbattery everytime

    restart  means sending soft restart command and i am not sending new battery command every time i send it once when connect new battery.

    5) The gauge should learn the scale factor to adjust for the 2 parallel cell. If you wanted to cut the Ra table by 2 this would be a decent step to take


    does this mean, i should replace the Ra table values with (current value/2) ?

    for soh i follow steps,

    set GE high -> wait 1s -> send gauge start() -> apply load and wait for 30 sec -> remove load -> send gauge stop command() -> wait for G_DONE set -> read SOH -> set GE low

    but still soh is decremented by  fix (soh delta % value) which is 2% on every cycle. chemistry id is correctly programmed

  • 0 in reply to karan patel1
    TI__Genius 10370 points

    Karan,

    You can replace the Ra table with Ra/2. Or you can allow the Scale Factor do the adjustment for you. 

    To check SOH decreasing please look at your measuredZ value (prior to scale Factor learning) and scaledR (after) against the Ra table. where that value falls is approximately where the SOH is trying to go to. Most likely you have added resitance in your setup causing the measuredZ to be higher than the ChemID Ra table resulting in a lower SOH.

    The Ra table is not evenly distributed across SOH range. Near the end of SOH are more points to better capture the end profile. Ra0-Ra6 covers approx 70% of the SOH range. 

    If my answer address your question please do not mark them rejected. If you have additional questions, please continue to ask. Marking them rejected will influence other peoples search results who might be facing the same issue. 

    Thanks,

    Eric Vos

  • 0 in reply to Vos
    Expert 1085 points

    Thank you Eric,

    actually i have in impression that if i do not cancel it than your last answer is considered as final and my further message will not be notified to you and i will not get answer for further query ,so i  have to create new thread that's why i cancelled every time.

    i replaced ra table by ra/2 but issue is same.i am not able to understand reason for SOH behaviour.There is not additional resistance other than device.

    i am collecting log for my observation in that i record measuredZ,scaledR,and also other few terms.May be this log will give some idea?

    BQ35100_log_TI.csv

  • 0 in reply to karan patel1
    TI__Mastermind 24870 points

    Hello Karan,

    Can you provide a gg.csv file here as well?

    I looked at the log file and I don't think you should replace ra table by Ra/2.

    You may want to go through the EVM user's guide here for initial setup and run the battery through cycles to understand the SOH behavior over time.