BQ27421-G1: Operation at low current consumption

This looks like the chemistry isn't a good match for the built in ChemID.

If the absolute current drops below the quit current threshold, the gauge will use voltage measurements to adjust depth of discharge and that relies on a good match between cell and build in chemistry. 

You can check this with https://www.ti.com/tool/GPCCHEM. This web based application will tell which bq27421 version is the best match for your specific cell chemistry.

Dominik
thank you.

GPCCHEM takes time, so let's start right away.

"There is no problem with discharge current of 80mA and 40mA, but there is a problem with 5mA."
The situation is that the ChemID is not correct.

Is it correct to understand that "80mA/40mA/5mA all give similar results if ChemID is correct"?

Please tell me how to change ChemID of bq27421-G1A.

The ChemID OCV tables will have a direct effect on results if the gauge enters relax mode (because the gauge will eventually take voltage measurements and correlate them to the ChemID OCV tables). So if the gauge is in relax (e.g. if your configuration treats 5mA as a relaxed cell) and the ChemID isn't compatible, you'll notice this more immediately than if the gauge is in discharge state (where it will affect cell resistance measurements mostly).

Changing the ChemID is not directly supported by TI tools for the bq27421. The bq27421 is sold with different ChemIDs (G1A/B/D):

Dominik
Thank you for your reply.

I got the log of GPCCHEM.

Never used Mathcad, but tried.
I installed "PTC Mathcad Prime 9"
The extension mcd was non-compliant.
Is there another way?

I will also attach the voltage, current and temperature files.

GPCCHEM(ChemID).csv

You do not need mathcad. You just need the roomtemperature rel/dis/rel files and follow the instructions from GPCCHEM to obtain a ChemID match. The report shows the max. DOD error for each G1A/B/D version of the gauge.

Dominik Hartl11
thank you.

As a result of GPCCHEM, Chem_ID was "1976".
Unfortunately, based on reports, bq274xxFamily is not worth using.
Do you know GaugeIC that can use this battery?

The bq27z561 can use this battery.

Dear Dominik Hartl11

Thanks for your reply.

Unfortunately, looking at the datasheet, I decided that the bq27z561 cannot be used.

However, I will purchase the bq27z561EVM-011 and challenge myself.

The batteries for the products we are developing have already been decided.

I would like to know more FuelgaugeICs that ChemID:1976 can use.

How can I find out how to get there?

Is there no way to select FuelGauge from ChemID?

From what I've researched, bq27220 might work.

Please tell me how to put the parameter of ChemID: 1976 here.

TI has several gauging algorithms with Impedance Tracking and CEDV being the main ones.

The bq27220 runs the CEDV algorithm. CEDV doesn't use a ChemID. It fundamentally relies on previously measured coulomb count from full to empty, which it uses for FullChargeCapacity. The main drawback is that it can't adapt to changing conditions during a discharge (e.g. temperature or load) which can lead to SOC inaccuracies.

You only need to configure EDV2/1/0 voltages (which are voltage thresholds for your cell and application that correspond to 7%,3% and 0% SOC), a 10 point voltage vs. SOC table and initial Full Charge Capacity. See the TRM for more information.

The bq274xx gauges use Impedance Tracking. Impedance Tracking uses ChemID, which describes how a specific cell behaves. The ones without flash memory cannot be programmed with a specific ChemID but they have pre-programmed ChemID, which may work for many cells (this can be checked with GPCCHEM).

TI gauges with a "z" in the name use Impedance Tracking and can be programmed with a ChemID. For example bq27z746. bq28z610. bq40z50. bq27z561. TI has a selection tool on www.ti.com/.../products.html