• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 63 subscribers
  • Views 384 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.

BQ34Z100EVM: IT learning with no charging cycle

Julian Eyre
Prodigy 240 points
Part Number: BQ34Z100EVM
Other Parts Discussed in Thread: BQ34Z100

Hi,

I have a system whereby it is not possible to charge the Lion batteries in situ.  They must be charged separately from the system in which they are used.

I.e. when I perform the initial learning cycle in order to optimise the IT parameters, in order to create a Golden Image, I cannot connect the battery pack during the charging phase.  I can only switch a fully discharged and relaxed battery with a fully charged one (I can maintain power during the switchover).  When I do this, the system never moves from a learned status of 4.  The FC flag remains at 0. It seems that the system has not recognised that the battery has moved from a discharge state to a charged state at all. In fact, when I then discharge the battery followed by relaxation, the system remains in learned status 4.  The battery has been Chemistry ID'd, with the correct ID and battery parameters loaded into the system.

Observed status was:

Discharged battery, fully relaxed: RUP_DIS 0  VOK 0  OCVTAK 1

Switch to fully charged battery :   RUP_DIS 0  VOK 0  OCVTAK 1  (i.e. unchanged)

Discharge cycle :                          RUP_DIS 0  VOK 1  OCVTAK 1

Discharged battery, fully relaxed: RUP_DIS 0  VOK 0  OCVTAK 1

It seems from the various guides that this is an integral part of optimising the system (indeed, without performing this first learning cycle the system is unusable).

My question, therefore, is to ask whether it is possible to optimise a system without performing the charging cycle in situ?  It is similar, I would imagine, to asking how one can optimise the system for a primary battery which has no charging cycle at all, which is, in fact, another task I must accomplish.  The documentation does not appear to offer any guidance on how to do this with the BQ34Z100 - G1 device.

All help gratefully received.

Julian

  • 0
    TI__Mastermind 42535 points

    Hi Julian,

    I think this is possible. For this you need to do one thing, if your battery is brand new, then set your Qmax to your battery's design capacity. Then set update status to 0x05. What I would suggest here is that you take a fully charged and relaxed battery with update status of 0x05 then check that VOK is cleared and OCVTAKEN is set.

    Then you can dsg it to empty and relax. You will see Ra update first during dsg at  the grid points and at the end of the cycle Qmax will update as well.

  • 0 in reply to Batt
    Prodigy 240 points

    Hi Batt,

    I did as you suggested, and the Ra did update during the discharge with RUP_DIS set to 0 and VOK set to 1 during the discharge as expected.

    However, once the pack reached the cell terminate voltage (3000 mV) and the load disconnected, even after several hours of relaxation (> 12 hours), the Update Status did not change to 6 and the QMax was not updated.  VOK was cleared as expected, so I do not know why the QMax was not updated.

    I have attached the register screen seen during the relaxation period after discharge.

    Regards,

    Julian

  • 0 in reply to Julian Eyre
    TI__Mastermind 42535 points

    It's your configuration that is causing an issue. I see that you have NiDV and NiDT set in pack configuration. This is seen in the flags. Your system is a Li-ion battery system. So, please reset them and run the cycle. Ni batteries use a very different method for chg termination and gauging. Once you set yours to a Li-ion setting, you updates should happen at the correct time.

  • 0 in reply to Batt
    Prodigy 240 points

    Hi Batt,

    The BQ34Z100 Technical reference states the following:

    "NiDT: Performs primary charge termination using the ΔT/Δt algorithm. See Charging and
    Charge Termination Indication. This bit is only acted upon when a NiXX Chem ID
    is used.

    NiDV: Performs primary charge termination using the –ΔV algorithm. See Charging and
    Charge Termination Indication. This bit is only acted upon when a NiXX Chem ID
    is used."

    I am not using a NiXX Chem ID, so these two bits should not have any effect on the learning process.

    Also, I am not using the charging part of the cycle, only the discharge part, as my pack is not charged in situ.  Those bits apply only to charge termination.

    However, I did run the discharge cycle again according to your last instructions (setting the learning status to 5 first), and by ensuring that the load was minimised during the relax period, the cycle completed correctly.  Before, I had a single low-power LED lit to indicate that the discharge load was disconnected (drawing about 3 mA), but it seems that this was enough to prevent a QMax update (the delta-V / delta-t was > 1uV/s).  Disconnecting the LED allowed the QMax update to take place.  I also disconnected the load at 200mV below the termination voltage rather than at the termination voltage (set at 3000mV), which may have helped.

    I suspect that this voltage change sensitivity may cause learning cycle failures for other customers, and it would be useful to emphasise the requirement that the load is absolutely minimised during the relaxation period after discharge, when battery voltage can easily change with time, even with very small loads.

    Julian