BQ27426: wrong full capacity estimation

Yaron Apel

Part Number: BQ27426
Other Parts Discussed in Thread: GPCCHEM, BQSTUDIO

Hi,

I am using BQ27426 with chem ID 0x1202 (4.2V), Initializatizing 'Design Capacity' to 2600 mAh.

I have encountered a situation where the fuel gauge estimates the full charge capacity as 2983 (and still battery level is at 80%).

the scenario in which this issue happend was:

1. Charge the device to 50%

2. Power off the device for 2 days (FG still connected to battery)

3. Charge the device till charging stops - at this statge the full capacity goes up from 2500 to 2893).

I attach an xls file with fuel gauge readings during this process.

How can this be explained?

Are there any parameters that can cuase this kind of behavior?

Thanks,

Yaron

over 4 years ago

0 Wyatt Keller over 4 years ago

TI__Guru** 105770 points

Hello Yaron,

I don't think the excel sheet was attached. With ROM based gauges the accuracy may be diminished because you cannot select a specific chem ID. To verify this I would run the GPCCHEM tool to see if the default ID's are a good match. Otherwise it may be better to use a flash gauge that can be programmed with a better chem ID.

It's best to check the chemID match with the GPCCHEM in the ROM based gauges to make sure it's close match.

Sincerely,

Wyatt Keller

0 Yaron Apel over 4 years ago in reply to Wyatt Keller

Intellectual 290 points

Hi Wyatt,

Thanks for the fast response.

unfortunately, replacing the fuel guage is not an option for us right now.

I have attached the file (this time...).

Please take a look at the values and tell me if you can find any possible reason for the fuel gauge to predict full capacity of 2983,

while design capacity is still at 2600.

Thanks,

Yaron

device_4_reaching_2900.xlsx

0 Wyatt Keller over 4 years ago in reply to Yaron Apel

TI__Guru** 105770 points

Hello Yaron,

Is the temperature consistent? Your log doesn't contain all the normal logged data from bqStudio.

FCC can change depending on temperature of the batteries and based on the chemID.

Sincerely,

Wyatt Keller

0 Yaron Apel over 4 years ago in reply to Wyatt Keller

Intellectual 290 points

Hi Wyatt,

The temperature starts at 306K and goes down to 297K.

Can this explain what I see?

Thanks,

Yaron

0 Wyatt Keller over 4 years ago in reply to Yaron Apel

TI__Guru** 105770 points

Hello Yaron,

I don't believe those temperatures are extreme enough to cause anything to happen.

What is the Qmax calculated by the gauge?

Sincerely,

Wyatt Keller

0 Yaron Apel over 4 years ago in reply to Wyatt Keller

Intellectual 290 points

Hi Wyatt,

The Qmax value is 16384 (default) through the whole process.

Thanks,

Yaron

0 Wyatt Keller over 4 years ago in reply to Yaron Apel

TI__Guru** 105770 points

Hello Yaron,

Can you update the Qmax to match the design capacity for your tests. Can you share your batteries datasheet?

Sincerely,

Wyatt Keller

0 Yaron Apel over 4 years ago in reply to Wyatt Keller

Intellectual 290 points

Hi Wyatt,

I have attched the battery datasheet.

I cannot update the Qmax, as most of the time the fuel gauge works fine, detecting max full capacity of 2400 - 2600.

The wrong full capacity estimation happens once in a while, and I don't know what is the cause.

Thanks,

Yaron

34262 datasheet A1.pdf

0 Wyatt Keller over 4 years ago in reply to Yaron Apel

TI__Guru** 105770 points

Hello Yaron,

It looks like the battery has a PCM, do the protections ever get triggered on the cell? I've seen this cause a lot of issues with gauges in the past.

So it's working the majority of the time but happens randomly on all your packs?

Sincerely,

Wyatt Keller

0 Yaron Apel over 4 years ago in reply to Wyatt Keller

Intellectual 290 points

Hi Wyatt,

Our PCM cut-off voltage is 4.250 +/- 25mv, and we didn’t see any cases when the Vcharge was above 4.225V.

You are correct - it's working the majority of the time but happens randomly .

Thanks,

Yaron

0 Wyatt Keller over 4 years ago in reply to Yaron Apel

TI__Guru** 105770 points

Hello Yaron,

Was the Qmax also changed during this relax period? (the 2 day relax) The Qmax may have updated during the rest to a different value.

Can you verify the default chemID will match using the GPCCHEM tool?

Sincerely,

Wyatt Keller

0 Yaron Apel over 4 years ago in reply to Wyatt Keller

Intellectual 290 points

Hi Wyatt,

The Qmax did not change during the relax period.

The GPCCHEM tool is irrelevant, as we use a ROM gauge, which has only 3 possible ChemID's, and only one for 4.2V batt.

Thanks,

Yaron

0 Wyatt Keller over 4 years ago in reply to Yaron Apel

TI__Guru** 105770 points

Hello Yaron,

The simulation that runs after the 2 day relax must of had some update of parameter, either Qmax, Ra, different temperature for it to change FCC.

The GPCCHEM tool will tell you if the default chemID will work well for your battery, this should be done before implementing any ROM gauge to verify.

Sincerely,

Wyatt Keller

Power management

Power management forum

BQ27426: wrong full capacity estimation