• State
  • Locked Locked
  • Replies 27 replies
  • Subscribers 64 subscribers
  • Views 6440 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.

BQ40z60 not balancing

Svenn S
Prodigy 170 points
Other Parts Discussed in Thread: BQ40Z60, BQSTUDIO

Hi All,

I'm trying to charge a 4 cell 1P 18650 batterypack with the BQ40z60EVM kit and BQStudio.

The BQ40z60 does not seem to go into balancing mode. Bit 2 (CBR) and bit 0 (CB) in the balancing configuration are set.

The bit registers show me that the balancer is enabled during charging and rest.

The GAUGE_EN bit is set in the Mfg Status init register.

After a reset, the balancer starts balancing for around 30 seconds, then the Cell balance times are zero again and the balancer stops. The Enable bit in the update status register is also cleared when the balancer stops.

The delta voltage of the cells is 17mV. The Min STart Balance Delta is set to 3mV.

How do I get the balancer to keep balancing? The firmware is updated to V0.15 build 21.

  • 0
    TI__Guru 73830 points

    Cell balancing does not work properly on the original v0.13 firmware, b ut you said that you have updated the EVM to v0.15. Does Update Status = 0E? Cell balancing is inhibited until the optimization and field Qmax updates have occurred and the Update Status = 0E.

  • 0
    Prodigy 170 points
    The update status register = 0E but Qmax has not updated yet.
    The balancer is balancing in rest, but not during charging. While charging, the cell voltages are not balanced and charging stops when one cell reaches 4200mV. The charger is not charging in constant voltage modus.
    How can I get the cells balanced during charging and enable constant voltage charging?
  • 0 in reply to Svenn S
    TI__Guru 73830 points
    Svenn,
    Please send a log file for the rest and charging profile and I will try to determine why the cell balancing is not working.
    Tom
  • 0
    Prodigy 170 points
    Hi Thomas,

    I just logged a charge cycle which took 3,5 hour:
    s23.postimg.org/.../Charge_Cycle.png

    The charger didn't balance during this charge cycle, and doesn't charge with concstant voltage.
  • 0 in reply to Svenn S
    TI__Guru 73830 points
    Svenn,
    Do you have the log file that you can share? I can review some of the register data.

    Tom
  • 0 in reply to ThomasCosby
    Prodigy 170 points
    Do you mean the data memory export or a register log during charging/rest?
  • 0 in reply to Svenn S
    TI__Guru 73830 points
    I was referring to the log data showing current, voltage, capacity, etc.
  • 0 in reply to ThomasCosby
    Prodigy 170 points

    I discharged a fully charged battery to about 40%. Then I let it rest during a night.

    Today I fully charged the battery.

    charging log: www.dropbox.com/.../charging log.log
    rest after discharge log: www.dropbox.com/.../rest after discharge log.log
    data memory: www.dropbox.com/.../data memory.gg.csv

    I hope you can help me balance and charge the battery pack in the right way.

  • 0 in reply to Svenn S
    TI__Guru 73830 points
    Svenn,
    I reviewed the log data and it looks like cell balancing is working as we would expect. I think that the problem is that the Relax Balance Interval is set too short and it is causing the balancing times to be recalculated too often. It should be set to 18000s. The balancing times are calculated properly based on the DOD data. Can you restore the Relax Balance Interval and try again?
    Tom
  • 0 in reply to ThomasCosby
    Prodigy 170 points
    Hi Tom,

    I changed the Relax Balance Interval to 18000s.
    The balance times seem te be still incorrect which causes the wrong cells to be balanced (see picture and datalog).

    Log: www.dropbox.com/.../langdurige rest.log
    Picture of cell voltages in log during balancing (rest): www.dropbox.com/.../Balancing 18h.PNG
    settings: www.dropbox.com/.../04-02-2016.gg.csv
  • 0 in reply to Svenn S
    TI__Guru 73830 points

    Svenn,

    The gauge is setting the cell balancing times correctly based on the DOD that it sets during rest. The problem is the the balancing is too aggressive. Bit 1 in the Balancing Configuration register should be set to 1. (RSVD_1). This will disable simultaneous cell balancing and use sequential cell balancing. We found that simultaneous cell balancing will cause multiple current loops when adjacent cells are balanced. You can also adjust the Bal Time/mAh Cell 1 and Bal Time/mAh Cell 2–4 registers to make the balancing less aggressive. You can try dividing both numbers by 2x. If you find that it does not balance quickly enough, then you can increase the numbers later. 

    Tom

  • 0 in reply to ThomasCosby
    Prodigy 170 points

    Hi Tom thanks for your response!

    I decreased the cell balancing times and cell balance interval timer. Balancing now seems to work better.

    The charger is still not going into constant voltage mode. When the cell voltage reaches 4200mV, the COV bit is set and charging stops.

    How can I get the charger charging in constant voltage modus after constant current charging like the Technical Reference Manual is showing in figure 4-2?

  • 0 in reply to Svenn S
    TI__Guru 73830 points
    Svenn,
    You may need to adjust the Charging Voltage parameters to compensate for a limitation in the Voltage Resolution parameter. I have an app brief on this available for download from the bq40z60 website.
    Tom
  • 0
    Prodigy 170 points
    Hi Tomas,

    Constant voltage charging is working correctly now, but the current is not flat during constant current charging.
    The charging current is set to 1000mA at all temperature and voltage ranges, but it starts charging with 950mA, then it drops below 800mA and rises back to 1000mA.

    Here is picture of a charge cycle.
    www.dropbox.com/.../Laden 11-2-2016 Graph.png
  • 0 in reply to Svenn S
    TI__Guru 73830 points

    Svenn

    I have not seen the current react like that. Can you send the log data for this run?

    Tom

  • 0 in reply to Svenn S
    TI__Guru 73830 points
    Svenn,
    The log data does not yield any hints as to why the current varies. Is there a possibility that there could be an external load drawing current from the charger? This would cause the current through the sense resistor to be reduced. The current also looks noisy and this could be caused by noise in the current sensing system. Can you try changing the current settings to a lighter value to see if this changes the behavior?
    Tom
  • 0 in reply to ThomasCosby
    Prodigy 170 points
    Hi Tom,

    There's no load connected to the EVM board during charging. During charging, an 18V (capable of devivering 2,5A) lab power supply, the batterypack and the EV2300.
    Today I tried with 500mA instead of 1000mA. This doesn't solve the problem.

    Here's the log with 500mA charging current:
    www.dropbox.com/.../17-2-2016 laden.log
  • 0 in reply to Svenn S
    TI__Guru 73830 points
    Svenn
    Can you export an srec from the pack and I will try charging using our EVM?

    Tom
  • 0 in reply to Svenn S
    TI__Guru 73830 points
    Svenn,
    I downloaded your srec file and will try it. I did notice a couple of concerns in the parameter set. The Term Voltage is still setup for a 3S pack. You should probably set it to 12000mV. You are using the default ChemID (1210). Did you verify that it is a good match to your cells?

    Tom
  • 0 in reply to Svenn S
    TI__Guru 73830 points

    Svenn,

    I ran a charge profile with your setup and I did not find the current to drop. The only things that I changed were the Term V (to 12000mV) and i increased the charging current to charge faster. The device is very sensitive to layout, so can you send me your schematic and layout files. I will try to find some potential issues.

    Tom

  • 0 in reply to ThomasCosby
    Prodigy 60 points

    Hi, Thomas!

    Could you please help me with a similar problem: the cell balancing does not affect cell voltage level difference.

    I took three mid-charged cells  (#1, #2, #3, approx.3.8V), and the #4 cell approx. 3.7V, all connected to bq40z60EVM and using the bqStudio for controlling. I have set the UpateStatus to 0xe and made the gas guaging enabled. Anyway, the cell balancing had no effect in either rest or in charge states: the difference of approx. 100 mV between the 'high" cells and the "low" one did not decreased.

    I`m also a bit dizzi about the OperationStatus[CB] flag: it is set to ONE signaling that some balance fet(s) is(are) active, i.e. the board shows the balancing process is in progress. It happens in rest, but I tryed to leave it for many hours - the voltage difference did not decrease. The same happens in charge state (but till the CUV flag is set, after what the charger turns off whithout keeping the constant voltage mode). I tried to vry balancing parametes but also had no result. 

    Could you give some idea where to look in the bqStudio for getting some detaild status information about active balancing fets? Or about the balancing process as a whole.

    Thank you!

    Pavel Rudenko.

  • 0 in reply to Pavel Rudenko
    TI__Guru 73830 points
    Pavel,
    It not valid to just set Update Status to 0E to achieve good balancing. The Qmax and Ra tables need to be updated as well. You also need to have an accurate DOD point set before it will balance. Finaly, make sure that the Balancing Configuration register is set to 07. If you a log file, then I can check it to see if you will provide any hints.
    Tom
  • 0 in reply to ThomasCosby
    Prodigy 170 points

    Hi Tom,

    the chemistry seems to have reset when I programmed the latest firmware. I've fixed that and I have connected a different battery pack, so we can rule out any issues related to the battery. The battery is a 4S1P, 6,8Ah LI-ION pack (SAFT MP176065N). The wires to the battery pack have also been shortened, they are now about 10cm long. The Term V has also been set to 12000mV as you mentioned. The charge profile now looks like this:

    The voltage drop at the right of the graph was caused by me, because I restarted the logging sequence.This seems to be a little bit better, but the constant current is still not what  I set it to (1000mA). Using a lower current (IE 500mA) doesn't change the shape of the charge profile by much.

    I'm using the BQ40Z60 EVM, not a custom layout.

  • 0 in reply to Svenn S
    Prodigy 170 points
    We're still waiting for a response, 2,5 weeks later. Can anyone help me with the above issue?
  • 0 in reply to Svenn S
    TI__Guru 73830 points
    Svenn,
    Are you working two problems here? I have made comments on both the cell balancing issue and the irregular charge current profile. I tested your srec on my pack and did not see the irregular charge current profile. There must be some other issue, since we are using the same type of hardware and firmware. You mentioned that you have long wire running from the battery to the EVM. Could you take a picture of your complete setup and I can try replicating that. Maybe we can find why we are getting different results.
    Tom