• State Resolved
  • Locked Locked
  • Replies 12 replies
  • Answers 3 answers
  • Subscribers 63 subscribers
  • Views 875 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: Microcontroller set up of BQ34Z100, device not starting

Sibe Jan Koster
Prodigy 150 points
Part Number: BQ34Z100
Other Parts Discussed in Thread: , BQSTUDIO, EV2400,

Issues: 1) Data not updated, even after FC,  2) FC remains set.

Use case: I am setting the BQ34Z100EVM not using Battery Management Studio, but the I2C port of a microcontroller. The chemistry is LiION , which is already pogrammed in the EVM by TI. (18650, one cell, 2600 mAh, 3.7(4.2V). In the '82' table, the capacity is set as 2600 mAh, the charge complete voltage at 4100 mV. Current parameter and voltage divider are adjusted to give the correct value.

The data and some settings are printed out through the microcontroller. The columns are: seconds , mA, mVbatt, temp, SOC %, remaining charge, flagsA. The last column is the calculated charge

650, -165, 3968, 251, 100, 2350, 0000001000010001, 2581
660, -165, 3968, 252, 100, 2350, 0000001000010001, 2581
670, -165, 3967, 251, 100, 2350, 0000001000010001, 2581
680, -165, 3967, 251, 100, 2350, 0000001000010001, 2581
690, -165, 3967, 251, 100, 2350, 0000001000010001, 2580
700, -165, 3967, 251, 100, 2350, 0000001000010001, 2580
710, -165, 3967, 251, 100, 2350, 0000001000010001, 2580
720, -165, 3967, 251, 100, 2350, 0000001000010001, 2579
730, -165, 3966, 251, 100, 2350, 0000001000010001, 2579

Removing the load and connecting a charger:

1350, -156, 3959, 253, 100, 2350, 0000001000010001, 2562
1360, -78, 4025, 253, 100, 2350, 0000001000010001, 2562
1370, -39, 4025, 253, 100, 2350, 0000001000010001, 2562
1380, 15, 4216, 254, 100, 2350, 0000001000010001, 2562
1390, 43, 4217, 254, 100, 2350, 0000001000010001, 2562
1400, 57, 4216, 254, 100, 2350, 0000001000010001, 2562
1410, 63, 4215, 254, 100, 2350, 0000001000010000, 2562
1420, 66, 4215, 254, 100, 2350, 0000001000010000, 2562
1430, 67, 4215, 254, 100, 2350, 0000001000010000, 2563
1440, 68, 4215, 254, 100, 2350, 0000001000010000, 2563
1450, 68, 4214, 254, 100, 2350, 0000001000010000, 2563
1460, 67, 4214, 254, 100, 2350, 0000001000010000, 2563

The device is live and communicating, as can be seen from the measurements and from the Discharge indicator.

problem 1: As can be seen from the flags, FC (full charge, bit 10) responds, but does not turn off

problem 2: SOC and remaining charge are never updated

The coulomb counter does not seem to start, or fails to update the data. The counter should start at full charge. The device is not sleeping. What else can be wrong ?

  • 0
    TI__Guru 52115 points

    Could you share your gg file? Also, have you finished the learning cycle? Thanks.

    Andy

  • 0 in reply to Andy Liu (BMS)
    Prodigy 150 points

    Unfortunately, I do not have one.  This is a consequence of not using EV2300. 

    The device is preprogrammed for LiION by TI (BQ34Z100EVM kit). 
    I suppose that means everything is set up right for LiION battery, at least in approximation.
    I just set the capacity and the terminating voltage in table '82' i.e. Data Flash/Subclass= 82 Offset=0 resp Offset =5. This is done at full charge, as indicated by FC flag.
    This is reported back by the module, but capacity is set at 2350 (should be 2600, maybe the difference is a result of an algorithm in the module kicking in)
    When charging after a complete discharge nothing changes either. This involves a restart of the microcontroller. Reported capacity is now 0 and not changing.
    Volts and amps are reported and check out allright.
    Flags are now  3200, 208, 4159, 278, 0, 0, 0000000100010110, 2704. I.e. SOC1&2, OCV relaxed, charging allowed
  • 0 in reply to Sibe Jan Koster
    TI__Guru 52115 points

    I assume that you were talking about the bq34z100-g1 device on an EVM.  BQ34z100-G1 is an IT-based gauge. Some operations (e.g. programming chemistry id) MUST be done by bqStudio.  I would suggest you get an EV2400 and use bqStudio to have all the necessary settings configured.  Without proper settings, the gauge device will not work as expected.

     

    And.

  • 0 in reply to Andy Liu (BMS)
    Prodigy 150 points

    Thank you. I come the same conclusion. Maybe it can be used in production as well.

    Reading some other posts, I conclude that the bq34Z100-G1 needs to have a switchable I2C port for programming (position 1) and readout in service (position 2). Can this be resolved by using 2k7 pullups on the I2C line ?

  • 0 in reply to Sibe Jan Koster
    TI__Guru 52115 points

    What are the Position 1 and Position 2 are you talking about? The bq34z100-g1 has only once i2c port. 

  • 0 in reply to Andy Liu (BMS)
    Prodigy 150 points

    The I2C communication with a microcontroller has been discussed before on this forum.

    One conclusion was that it is not advisable to keep the microcontroller connected when the EV2400 is used. This will lead to communication errors.

    Then I suppose the only way to program a bq34Z100 - on a production board, using the golden file - is to disconnect the I2C of the bq34Z100 from the microcontroller and connect it to the EV2400. This could be done by a 4-pole switch or a 4-pole jumper. 

    Sorry for the confusion. 

    My EV2400 will arrive next week. 

  • 0 in reply to Sibe Jan Koster
    TI__Guru 52115 points

    You can also use a df.fs file for bq.fs. file for mass production.  Bascially, any host with an i2c port can program a bq34z100-g1 device. 

    See the FlashStream Parser in the app note below.

    http://www.ti.com/lit/an/slua801/slua801.pdf

    Andy

  • 0 in reply to Andy Liu (BMS)
    Prodigy 150 points

    Hello Andy,

    I now have a EV2400. Reading out the device is indeed a lot easier. 

    Unfortunately, the result is still the same.

    It seems the available energy has gone negative (-25921 mWh). I have connected the battery between pack+ and batt- on the bq34Z100EVM. The load and the charger are connected to pack- and pack+. A discharge current is indicated with a minus, as it should be.

    The remaining capacity is 2552, buts jumps to 0 in the log, when a charger is connected. Also when connecting a load, the remaining capacity junmps to 0. Also the remaining time to empty jumps to 0.

    There is no change in the remaining capacity, when discharging. The available energy grows somewhat more negative.

    It seems charging is never registering and discharging is. That would account for the negative remaining energy. 

    I have attached a golden file0100_0_16-bq34z100G1_LiION.zip.

     

  • 0 in reply to Sibe Jan Koster
    TI__Guru 52115 points

    Could you provide a bqStudio log since you already have an EV2400? 

    Andy

  • 0 in reply to Andy Liu (BMS)
    Prodigy 150 points

    Hallo Andy,

    Please find attached my log from a charge to full and discharge to breakdown (the LiION battery has an electronic protection). Charging i done be a lab supply that tops off at 4300 mV and with a current limit set to C/10. Discharging is done by a resistor, I halved the value near the end.

    The device responds to a change from charge to to discharge. Number of cycles is still zero. State Of Charge is 100 when charging and turns to 0 on discharging. At the same time the Remaining Capacity turns from 2555 to 0. The Coulomb counter is still not running. 

    I have set the taper current to 100mA, this had a negative value (-156mA), which seemed to be wrong.

    Considering the picky nature of the device, do you think it can be made to work with a solar MPPT charger ? The charging current will vary constantly, and full charge may not be reached at all.

    Kind regards,

    Sibe Jan Koster

    log20200604.zip

  • 0 in reply to Sibe Jan Koster
    TI__Guru 52115 points

    I took a look at your log file.  The Update Status is 4 in your log file, which means the learning cycle is not done yet. 

    Have you finished the learning cycle yet? If not, please go ahead and do the learning cycle.

    Andy

  • 0 in reply to Andy Liu (BMS)
    Prodigy 150 points

    Thank you. I did get the gauge to work after changing the battery chemistry to 1126.