BQ34Z100-G1: Ra Table and Qmax not updating during battery use.

Hello Douglas,

Qmax is updated only if certain conditions are met.

TRM section 3.2.8 Quit Current, Dsg Relax Time, Chg Relax Time, and Quit Relax Time has more information

Also make sure that correct charge termination is achieved during charge cycle

I used the same settings when the golden image was successfully created. Quit current, 45mA, was at least 10 times more than idle. Dsg and Chg current thresholds were 30% more than quit. The dsg and chg relax times were at their default of 60s. Charge termination is set to 250mA while charger is set to 150mA.

A key difference is that during the golden image creation there was low charge and discharge rates; C/2 charge, C/5 discharge.

Our application of flying quad drones has discharge rates of C3-C8 depending on the maneuver.

My key question right now: Can the bq34z100-G1 track batteries that have high discharge rates? 

If it gauge will work for high discharge current applications I assume we can figure out the setting needed for it to track. 

Is there an app note for applications with high discharge rates?

From the RM 

After about 6 minutes in RELAX mode, the device attempts to take accurate OCV readings. An additional
requirement of dV/dt < 1μV/s is required for the device to perform Qmax updates.

I assumed that if after 6 minutes the dV/dt was > 1uV/s it would hold off taking a OCV reading. 

Does the OCV reading try and fail instead?

If this is the case then maybe I would need Dsg Relax Time of 1-2 hours since dV/dt is much more than 1uV/s 6 minutes after a 4.5C discharge.

Hello Douglas,

I believe that the gauge will wait for dv/dt condition to be satisfied before taking a reading. Do you have a log of the registers to check for approximate dv/dt?

I realized I had a mistake according to TRM. I left the golden image update status at 0x06 so it was loaded with 0x06 already set. On the test unit I changed update status to 0x02 and then sent the Enable IT command to set update status back to 0x06 before running another cycle. 

This latest cycle updated Ra Table and Qmax. Unfortunately I didn't have access to the Arbin so it was a C/4 charge C/2 discharge cycle so I don't know if is fixed or if it only works for slow discharge cycles.

I won't have access to the Arbin again until next week. At that point I'll do some C1 charge and C5 discharges and see whether it will update or if there is still a problem.

BTW, I didn't make a log of registers, I'll do that if it fails to update after high discharge cycles.

Hello Douglas,

The discharge rate may be too high. The gauge has to perform a mini learning cycle to update Ra tables and Qmax.

I ran it on the Arbin the last 2 days and it updated both qmax and Ra Tables. I'll be trying it with a high current variable load next week.

Hopefully it will still update but if not could any of Ra Filter, Res Current, Max Res Factor and Min Res Factor be used to tune for my load?

I didn't find descriptions for these values but their names seem revelant.

Hello Douglas,

Thanks. will try to get you more information.

Ra Filter controls the low-pass filter for each cell resistance update. This is about how much the gauge allows Ra to change between cycles and not about your load.

Res Current is the threshold that the load current must exceed to allow Ra measurements. This should not be less than C/10 because the voltage drop over the internal cell resistance won't be high enough for the gauge to calculate Ra with reasonable accuracy.

Min/Max Res Factor control how much Ra is allowed to change based on a grid point update. The gauge calculates a scaling factor = round(100 * Ra_new/Ra_old) and if this is outside the Min/Max Res Factor, it will disable updates.