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.

BQ35100: Scaled R negative value

Part Number: BQ35100
Other Parts Discussed in Thread: EV2400

Dear Sir/Madam,

I tried to read Scaled R value and saw that it was almost negative value. And it seems not stable such -193, -2211, -65536.... while not using battery

What does negative value means? is it right? 

Using battery EVE model ER34615 19Ah, CHEM ID= 0x0615

Best regards,

KPK

  • Your question has been assigned to an engineer and we will get back to you.

    Andy

  • KPK,

    If you do not have a battery attached then the gauge has no correct voltage reference to use which will result in bad values being fed into the filter. You will result in an impedance based on the current you measure with no voltage drop. could result in a negative value

    When you do connect a battery you need to follow this procedure

    1) GE high

    2) Gauge Start Command (wait got [GA]=1)

    3) Apply load (This step should cause at least a 100mV battery voltage drop)

    4) Remove load

    5) Issue gauge stop command (Wait for GDONE, 15Sec by default)

    6) Read parameters. 

    If your current is not high enough or you do not get enough of a voltage drop you could have issues with the math inside the IC. 

    Thanks,

    Eric Vos

  • Thank Erik,

    The lastest picture is I plugged battery already. What does ScaleR negative value mean?

    I had 2 more questions:

    1. Can we use SOH (%) to estimate remaining capacity of battery in EOS mode? Or we need base on ScaleR value? As I see ScaleR seem not stable

    Your 6 steps:

    1) GE high

    2) Gauge Start Command (wait got [GA]=1)

    3) Apply load (This step should cause at least a 100mV battery voltage drop)

    4) Remove load

    5) Issue gauge stop command (Wait for GDONE, 15Sec by default)

    6) Read parameters.

    2. Can I pull GE low after Read parameters done to save power consumption? I think Bq35100 still collect information from battery when sleep by GE low, correctly?

    thanks and regards,

    Tuyen

  • Tuyen,

    1) SOH is relative. This gauge does not report remaining capacity out. I would not trust SOH since it could hold at a specific value for a while. 

    2) Yes after you read the values it is intended GE would go low for this device. 

    Thanks,

    Eric Vos

  • Thanks Eric Vos,

    I also had more questions:

    1. When we use BQ Studio and EV2400 to load right chemical ID to bq35100. Do we need change any parameters to suitable with our hardware design? I calibrated voltage and current already.

    2. As I know we need to discharge battery from Full to Empty to get ScaleR at this point. But my battery capacity 19Ah while my device is consumed 5-100mA , so it took too long to get the result. How can we speed up this process? Can we use a heavy load?

    3. In the discharge process, It seem scaleR not stable such as: 1567, 2056, 3010, 56, 0, 0, 0, 0, 1650.... it was be 0 frequently. What is reason in this case? How we decide warning empty battery when it seem not stable?

    4. After load chemical ID to BQ35100. I saw Ra Table as below

    It mean when read Scale-R > Ra 14 (5072) then battery is flat? Or we need get this value from step in question 1?

    regards,

    Tuyen

  • Tuyen,

    1) You will need to configure the gauge to be in EOS mode after you finish calibration. If you are in a multicell configuration you will also need to setup the board to use the VIN. There are lots of document at the bq35100 web page. 

    2) You should follow a procedure where you gauge your battery 5 or so times with rests between, discharge at a higher rate for a longer time,  then let your device rest for an extended time, gauge 5 times with rests between. The major issue most people have is they do not allow sufficient rest between pulses to allow the battery to return to OCV

    3) This is most likely due to one of three reasons. 1) the battery not being is a well rested OCV state when you start the "Gauge pulse" outlines above in an earlier post. 2) The you are following the needed procedure. 3) You are not applying a large enough load to cause the correct voltage drop on the battery. 

    4) The image did not show up so i am not sure. In general Ra0 is battery full, Ra14 is empty. But they are not evenly distributed. They get closer as you get more discharged

    Thanks,

    Eric Vos

  • hi Eric,

    Here is the snapshot that Tuyen is missing out:

  • Tuyen/Poh Kok,

    This particular ID does need to have a very flat profile. I am not sure SOH will be the best judge. This battery seems like the main benefit from it will be the EOS (End of Service) alert when the battery is near empty.

    Thanks,

    Eric Vos

  • Thanks Eric,

    I have some  question

    1. I read 8.2.2.5 Current Sense Resistor Selection in this guide

    www.ti.com/.../bq35100.pdf

    I had concern about unit of is mohm or ohm? Let example my device consumption maximum is 200 mA then I need R = 0.625 mohm, correct? If don't have 0.625 then can i change to 0.5 mohm (appropriate 250mA)?

    2. I saw in 11.8 BatteryAlert() 0x0B of this guide

    www.ti.com/.../sluubh1c.pdf

    It had EOS(Bit 3): ALERT is  triggered because of EOS.

    Can we check thit bit to decide a empty battery in EOS mode?

    regards,

    Tuyen

  • Tuyen,

    1) I would expect the size of the senseR to be around 100mOhms. The ADC has a range of ~100mV. 100mV/200mA = 500mOhm for full range. so i would leave some margin and use around a 100mOhm.

    2) The alert bit will trigger for lots of reasons. there is a register in the DF to configure what to trip for. Please check what is enabled.

    Thanks,

    Eric Vos