bq34z100-G1 - State of Health if the battery is not used often

The gauge would provide SOH reporting, but learning will have to occur for SOH to update. Updating Qmax and the Ra table on a one month basis should be okay.

Dear Tom,

thanks a lot for your support.

I sent your answer to the customer.

Thanks again.

Best regards,
Massimiliano.

A reply from the costumer :) Let me explain.

We use the battery in standby mode. The battery only becomes active when the mains fails.
Every month the battery is tested by discharging it with about (1A) for an hour (not totally empty, let say half empty). During this test the battery is also discharged with a larger current 3A periodically (lets say every 10 seconds for 100ms). This for two reasons.
- We want to know if the battery is capable to deliver the current and to know the battery has enough capacity.
- in my opinion the fuel gauge needs this to calculate the internal resistance to determine the SOH.

The questions are:
- is testing ones a month enough, so the fuel gauge can reliable estimate the SOH
- what do we have to do during the test to allow the fuel gauge to do the measurement reliable? Is the test I described ok? Or do I have do do something else? Please advice!
- Can I pick ant fuel gauge or can you recommend one what does the job best giving the circumstances.

The Ra table should update if the current is greater than C/10 and the pack discharges through a few grid points. A 50% discharge over a one hour period should be adequate. Qmax should update, if the change in DOD is at least 37% and the DOD points get set in the rest periods before and after the discharge. The cell voltage also have to be outside of the disqualified voltage range for your ChemID and the temperature has to be between 10C and 40C. Let me know which ChemID you have selected and I will let you know the disqualified range.

We only start the test when we know the battery is full.

After the discharge test we have to charge the battery again as quick as possible (charge current 0.5C..1C), so there is no rest periode after the discharge test. Is this really needed to determine the Qmax? If so, how long is the minimum period?

The battery we us is a 2S1P pack build with two 18650 1500mA LiFePo4 cells (max discharge is 3C).

By the way, we need a fuel gauge for the 2S1P pack with balancing and protection. A key/ID is a nice feature.

There are two issues with this configuration. The rest period after discharge is required and it could take up to 48 hours for LiFePO4 cells. The second is that the disqualified range for LiFePO4 cells covers almost the full voltage range for the battery, so a full discharge will be required for the Qmax update. We can contact you off-line to discuss other options to support this use case.

Thomas. I did send a message off-line, I think. To be sure I put my reply here also:


The rest period after discharge is needed before charging again? If this not an option for us. We must have charged battery when mains fails (battery backup).

Off course it is best to full discharge the battery, the capacity can then be calculated, this also says something about the aging.

Measuring OCV before and after discharge is not reliable I think, because it is temperature depended and the battery must rest for at least 4 hours after discharging. It is better than nothing... but if the SOH can't determined reliable It is better to do nothing. To prevent false notification "battery fail". But, if we do not use the fuel gauge information to judge the battery but we use the data only logging it might say something if the target is returned with complained. We then have some statistics about the battery, together with our test data.

What if we use 18650 Li-Ion cell's or Li-Po??? You write: for LiFePo4 7% SOC to update FCC for learning, is this for Li-ion or Li-Po better?

Rody,
Required rest periods are an inconvenience, but they are required for Impedance Track devices to update Qmax as the cells age. The rest period does not necessarily take 4 hours. It just takes enough time for the cell voltages to stabilize for an acceptable OCV measurement. Li-Ion and Li-Po cells do require less rest time than LiFePO4 cells. I will contact you off-line to discuss your requirements to see if we can find a better solution.

Tom