BQ27Z561: bq27z561 learning cycle question

Thanks Nick! 

Any issue using standard charge current with power supply during learning process-Or must the solar charger (or power supply setting simulating the solar charger) be used?

Thanks, Nick.

Here I was asking a similar question here:

(1) BQ27Z561: BQ25185DLHR LOW CURRENT CHARGING - Power management forum - Power management - TI E2E support forums

I guess my question is the charge term taper current, will not really indicate end of charge in our setup. Our charge will always be supply less than battery cutoff so it's not a good marker for FULL charge. Is the taper current used to indicated FULL charge?

We'll have to manually verify end of charge by other means (shutdown the charge process in software monitoring voltage, or setting a timer....TBD).

Is the gauge smart of enough to track charge for the little current that is injected to it OR will it see a low current and say FULL charge when its actually just in the middle of of charge? - THAT IS MY CONCERN.

This is the IT algorithm doc:
Evaluation 3: Theory/Implementation Impedance Track Battery Fuel-Gauging Algo (R ev. B

From there:

DODcharge is assigned equal to DOD0 aft first DOD0 update after CHARGE TERMINATION BY TAPER CURRENT

where 

Qstart = Qmax (DOD0 - DODcharge)

Thus Qstart would equal ZERO and thus be equivalent to a full charge? 


I'm concerned that the taper current is being used to determine FULL CHARGE


Additionally. Our application has an active state and sleep state with current ranges from 1.8A during message transmission, averaging  maybe 500mA and hopefully sub-milliamp (microamp) current levels during sleep. The sub=milliamp is hopeful/optimized levels we have NOT reached yet. Are realistic current discharge is probably a couple milliamps.

Will the gauge work accurately given the low current draw?

What DSG Current Threshold should be used given our application-Should it be based off the wake state?

I'm not an expert in the alogrithm. Most of my time spent on the driver was implementing the calibration procedures on a microcontroller and working through errors in the documentation. 

From a battery perspective (not an expert), the end of charge is determined by 4.2V being reached (Constant voltage mode entered) and the current dwindling decreasing from a current higher than cutoff current (say 1700mA), down to the cutoff current. 

We can't track this decrease as our solar charge current is at best 50ma (always below BATTERY cutoff) and thus cannot use this for determining charge. To me, the IT alogrithm indicates taper current (perecursor to cutoff current) is utilized in indicating full charge

IF the gauge is using current level as the determining factor to indicate FULL CHARGE then i fear it could indicate full charge when it's actually not. OR the gauge doesn't work for our setup.

I feel the gauge would work fine with a standard charger (High charge current in CC phase that dwindle to cutoff current in CV phase)...but our current setup is suspect to me.

So three 'cutoff currents at play'

CHARGER cutoff current -    30mA

GAUGE Assigned taper current (related to the cutoff current of charger), set by us but deduced from charger...say 35 mA

BATTERY cutoff current  at standard charge 1700mA CC mode, 4.2V CV mode, 68mA for a cell--->272mA for the 4p pack

I'm not certain these are aligned-hence I'm concerned about hw (charger->gauge->battery) compatibility

Thanks!

Hello,

Today's a national holiday. We will get back to you tomorrow.

Regards,

TI Apps Team

Thanks, Nick.

So here's my settings. current pack/device does is not optmized so we'll like never settle below the 1mA range at this time...Is this going to be a problem.

I'm leaning towards the ideal values...but I don't know if they're compatible with with our current unoptimized setup. Note we are not using the device on a pack at this time, the cells soldered/welded in place on the device. Pack construction is saved for a later date.

• Design Capacity mAh = 3400mAh x 4 = 13600 mAh

• Design Capacity cWh = (3.6v)x(3400mAh)x4 = 4896 cWh

• Design Voltage = 3600mV

• Term Voltage = 2650mV (NOTE: This is the pack cutoff for learning mode, and must be updated to system cutoff AFTER LEARNING.

• Charge Term Taper Current = USUALLY C/20 = 13600mAh/20 = 680 mAh BUT…Current CU hardware cutoff is 30mA, this is less than pack cutoff for standard charge (272mA for pack-->68mA per cell).
  https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/1382777/bq27z561-bq27z561-learning-cycle-question
  Use 35 mA … IDEAL 68+5 =73mA

• Dsg (discharge) Current Threshold = This is the current threshold above which the gauge detects that it is in discharge mode. A positive number should be entered and the firmware automatically interpret it as a negative (discharge) current threshold. The default value is suitable for many applications, but if customized it should be set to a value lower than the minimum system draw when in active mode. When discharge current exceeds this threshold the algorithm enters the Discharge state. A value below C/10 is generally reasonable. This setting is found in the Gas Gauging, Current Thresholds section of Data Memory. USUALLY default (60mA), (NOTE: default usually works, recommend below C/10, comply with NOTE 1 below)…BUT use 10mA … IDEAL 2mA

• Chg (charge) Current Threshold = default (75mA), (NOTE: comply with NOTE 1 below)… BUT use 25 mA …IDEAL charger cutoff current or a little less (25mA)

• Quit Current Threshold = default (40mA), (NOTE: default usually works, recommend below C/20, comply with NOTE 1 below)…BUT use 10mA… IDEAL 1mA

• BQSTUDIO CHEMISTRY: Chem_ID 2162, (NOTE: read back with CHEM_ID COMMAND, will equal 0x2162-That is HEX), ALSO a checksum that can be verified…), corresponds to INR18650-35E (3400mAh)

Hello Craig,

If the current draw from the gauge never goes below 1mA then the gauge will never enter it's "rest" state, which may cause inaccurate gauging. I think that testing will have to be done with both values of settings to see if the gauge reports the correct behavior in practice. However, depending on the load profile, it may not be necessary to use the ideal values. The setting values looks fine, but I would change Quit Current to 8mA since Quit current threshold needs to be below Chg and Dsg current threshold.

Regards,

Nick Richards

Thanks Nick, I'll probably run with the wider (non-ideal) margins. 

One thing though with a charger connected. The charger will bear the load as well as charge the battery. We'll be monitoring open-circuit battery voltage to end the charge.

The charger will support the load AND charge the battery full charge voltage (or timeout/ as defined by us) is reached. At this point the charger will STOP charging the battery but continue powering the load.

While the charger powers the load without charging the battery the current 'for the battery' should fall to below 1mA.

Current charger being utilized is one of yours (bq25185):

www.ti.com/.../bq25185.pdf

Hey Craig,

If the load falls below 1mA then that is the ideal scenario for the gauge.

Regards,

Nick Richards

Sounds lke currently our filed charger will never charge. Any issue with factory charging and then just using the learned image with the IDEAL values?

Hey Craig,

There should not be any major issues with using the factory charging and using the ideal values.

Regards,

Nick Richards