BQ35100: When to run gauge start

Hello,

This question has been assigned and will be reviewed when possible. In the meantime you can provide the associated .log and .gg files with the project.

Thank you,
Alan

EOS mode doesn't coulomb count so it doesn't account for any power draw. It uses a significant discharge current to calculate cell resistance, which is the reason why you have to turn on the gauge and then issue the gauge start command. This tells the gauge to start measuring cell resistance. This should be turned off with the gauge stop command when the discharge stops (the system enters a low power state). The gauge will continue to measure after gauge stop (15 seconds default) and then calculate its EOS results (resistance, SOH). It can then be powered down.

So you should not leave the gauge started all the time. Only while there's an active discharge happening.

So should I be using a different mode then? How would I best accurately track the state of health given that most of the charge will be drawn while in idle mode over a long period? I suppose the idea is that the discharge current will give it a resistance reading that it uses to determine the soh of the battery in a non-relative way. So if I don't necessarily plan for it to be doing a significant discharge that often, I should either increase the max soh delta or put a controlled discharge circuit on the board I can regularly discharge through to run the measurement?

The answer depends on the chemistry and your application. It's a choice that the customer has to make. The EOS (which also includes and SOH measurement) algorithm doesn't need the information from the cell when it is discharges slowly in an idle state. The idea is that if there's a significant discharge (e.g. the product is turned on for a few minutes), the gauge can measure cell resistance and then deduce SOH and decide if it's time to end service for this cell.

If you want to use EOS or SOH then I would add a method to apply a significant load every once in a while (like you wrote). Otherwise the gauge won't be able to tell what's happening with the cell (e.g. if it's always trickle discharged as a permanent backup supply for a RTC or similar).

And I should use EOS with a lithium thionyl chloride battery, correct?

So, if I want to do a rundown test I will just have to take periodic measurements while doing discharges that are no more than 2% of the capacity? Like I can run it for a while at .5 mA, figure out how long it will be to run down 1.8% or so, then turn on the gauge for a second and turn it off again? Can I keep the current draw applied constantly while doing that?

You should stop the current whenever you stop the gauge.

The reason for EOS for this chemistry is that accumulator mode is usually not accurate enough over the lifetime of the cell (though if you have a continuous discharge test, this does not apply) and that the OCV is very flat so it's not possible for the gauge to determine depth of discharge with a voltage measurement alone. However, it is possible to determine cell resistance if the gauge can measure the current under load so if there is a discharge load that is high enough to cause a voltage drop over the cell's internal resistance then the gauge can calculate this resistance and this resistance is the basis for the EOS determination.

This is very helpful thank you. Can you outline a timetable or series of steps for a suggested rundown test method?

I would discharge it in maybe 10 steps until 80% of rated capacity is depleted and then continue with 4 steps until 100% of rated capacity is discharged.

The absolute load depends on the capacity of the cell. I'd use a load that is well below the maximum rated discharge rate but high enough that this doesn't take too long.

so does that mean i should change the max soh delta to 20%?