BQ27542-G1: Device reporting Wrong SOC

Shaival Shah

Part Number: BQ27542-G1
Other Parts Discussed in Thread: BQ24190, CC3200, BQSTUDIO, EV2400

We are having an issue on the IOT devices reporting wrong SOC even though the battery does charge completely. This started happening on a handful of devices since past 2 weeks and we are using the fuel gauge BQ27542. So the SOC does get updated correctly for a certain period of time e.g. 0 to 73% and then it stays on 73% after that. When I checked the voltage on the battery it says 4.2V which is completely charged.

So I just wanted to ask why after a certain period of time the same fuel gauge did start reporting the SOC incorrectly? These devices were running same hardware for past 6 months and were reporting the SOC correctly.

I am using CC3200 as the master device to program the gauge. I feel somehow firmware update might have changed the coordination between charger IC and fuel gauge. The charger IC is BQ24190. The battery is rechargeable Li-ion. Any comments/help would be appreciated.

over 8 years ago

0 Onyx Ahiakwo over 8 years ago

TI__Guru* 76230 points

hi Shaival,

Did you identify your chem id for your battery correctly ?
Did you carry out a learning cycle after you identified your chem id?

thanks
ONyx

0 Shaival Shah over 8 years ago in reply to Onyx Ahiakwo

Prodigy 120 points

Yes I did identify it correctly and I did run the learning cycle and generated the golden image accordingly too. As I mentioned the gauge and the SOC reporting was correct for almost 6 months before this problem showed up.

0 Onyx Ahiakwo over 8 years ago in reply to Shaival Shah

TI__Guru* 76230 points

check to see that you have had field qmax updates. If the gauge has been used in the field for 6 months and you haven't had qmax updates occurring (qmax would be significantly different from your golden file value), that could be the cause of the problem. You would want to try to run a learning cycle on that pack to see if the problem goes away.
thanks
Onyx

0 Shaival Shah over 8 years ago in reply to Onyx Ahiakwo

Prodigy 120 points

Hi Onyx,

Thank you for the suggestion. I did run the learning cycle again but it didn't fix the problem. One more update when I unplug the battery from the device completely for 2 minutes and plug it back in the gauge reports correct SOC i.e.99% charged which says Qmax value should be updated fine. Actually the same problem happened on one of the device freshly shipped from factory a week ago where the charge was stuck at 76%. This was freshly calibrated and programmed in the factory using golden image and the learning cycle was performed too. So now it gets the issue more weirder.

Let me know your thoughts on this.

Best,
Shaival

0 Onyx Ahiakwo over 8 years ago in reply to Shaival Shah

TI__Guru* 76230 points

HI Shaival,
pls send your srec, your golden gg, a gg extracted from one of the units behaving awkwardly, and a log file of the issue using bqstudio for analysis.

thanks
Onyx

0 Shaival Shah over 8 years ago in reply to Onyx Ahiakwo

Prodigy 120 points

fuelgaugefiles.zip

0 Shaival Shah over 8 years ago in reply to Shaival Shah

Prodigy 120 points

Hi Onyx,

I have attached the golden image and the srec files. Since this being a IOT device I have no way to access the bad unit and get the log file. Let me know what calibration data you need and am happy to send that.

Thanks,
Shaival

0 Onyx Ahiakwo over 8 years ago in reply to Shaival Shah

TI__Guru* 76230 points

Hi Shaival,

I requested a gg.csv file which you can extract using bqstudio. if you are able to reproduce the issue, you should do so using our EVM. That way you will be able to log using bqstudio.

thanks

Onyx

0 Shaival Shah over 8 years ago in reply to Onyx Ahiakwo

Prodigy 120 points

Hi Onyx,

So this issue occurs very intermittently and is not being reproduced on the EVM. I still cannot hook up the IOT device using bqstudio so is there any other way I can obtain the gg.csv file ?

Thanks,

Shaival

0 Onyx Ahiakwo over 8 years ago in reply to Shaival Shah

TI__Guru* 76230 points

How did you extract the srec you sent me? I thought you said the srec was from a failing unit. Is that srec a golden file?

thanks
Onyx

0 Shaival Shah over 8 years ago in reply to Onyx Ahiakwo

Prodigy 120 points

Yes it is a golden file, it is not from a failing unit. Sorry if I was not clear.

0 Onyx Ahiakwo over 8 years ago in reply to Shaival Shah

TI__Guru* 76230 points

I am afraid that will not help us much. Without a log file capturing this issue or atleast a gg file from the failing units, it will be difficult to say what the cause of the problem is.

You will need to get failing units returned to you to run tests.

thanks
Onyx

0 Shaival Shah over 8 years ago in reply to Onyx Ahiakwo

Prodigy 120 points

Hi Onyx,

So ours is IOT embedded device and do you have a way you can think of how to connect that to the bqstudio to get the gg.csv file ona failing unit ??

0 Onyx Ahiakwo over 8 years ago in reply to Shaival Shah

TI__Guru* 76230 points

blue wire the com lines and connect an ev2300 or ev2400 to the computer. make sure your host is turned off to avoid bus clashes.

That way you will be able to extract the srec file from the gauge and log data as well.

thanks

Onyx

0 David C over 8 years ago

TI__Expert 3350 points

Shaival,

Based on the files you submitted

It seems that you never programmed a chem id, or did a learning cycle. The chem id on your file was the default chem id. The update status was 00 indicating a learning cycle was never done. It ought to be either 02 or 06 if a learning cycle was completed.

The gauge might seem like it’s working but since it’s hasn’t been properly configured the SOC is just a generalized calculation.

-David

Power management

Power management forum

BQ27542-G1: Device reporting Wrong SOC