BQ35100: EOS Alert is not generated even after Battery is drained fully.

Hello Sudarshan,

Can you outline what the current test procedure is doing?

Regards,

Adrian

Hello Adrian,

I request you to refer our same thread. This thread is locked by TI but I appended the above post to the same thread. Kindly go through that to recall about our problem.

We are trying to monitor the EOS for our LiSoCl2 battery by discharging it by 2% and then 5 hours rest and then GAUGE_START then 1 sec pulsed load and then GAUGE_STOP command. Earlier we were using a current sense resistor of 2 mOhm but you had suggested us to change that to 100mOhm and repeat the test. So, now we are repeating the test. Although, we didn't  see any change in Measured Z and Scaled R values and we are getting a large value of (-32768 Ohm  of Measured Z and -32767 Ohm for Scaled R)  for a Fresh battery itself.

I request you to also refer our same thread for recalling our previous discussion about the same.

With regard,

Sudarshan Chaudhary

Hello Sudarshan,

Sorry for taking a while to get back to you. I checked back at the screenshot of the bqstudio and I see the EOS_BAD_OCV bit is set. This is indicating to me that before the 1 sec pulsed load the battery has not relaxed enough before the pulse load. After doing the 2% discharge, are you relaxing for at least 5 hours? This bit will usually set if the battery was not fully relaxed before the pulsed load.

Regards,

Adrian

Hello Adrian,

Yes, the battery was relaxed for 5:30 hours. And the entire procedure was followed as per your suggestion but this bit was never set earlier. But this time when we changed that current sense resistor to 100mOhm then we observed this bit as set. I repeated the test, but again observed the same i.e. EOS_BAD_OCV bit is set.

With regard,

Sudarshan Chaudhary

Hello Sudarshan,

After changing the sense resistor, was the gauge recalibrated to ensure that current and voltage readings were accurate? I will probably need to see a waveform of the battery voltage before and after the pulse load to see how the battery voltage is behaving.

Regards,

Adrian

Hello Adrian,

Yes, I had re-calibrated the gauge. Even after that same problem was observed. Do you need to see these waveforms on Oscilloscope or in data graph of BQ studio? In Oscilloscope, it might always show a DC line of 3.6V. but not sure about Voltage data graph of BQ Studio.

With regard,

Sudarshan Chaudhary

Hello Sudarshan,

Oscilloscope will be best. Ideally I would want to see the oscope plot right before, during, and right after the learning pulse happens.

Regards,

Adrian

Hello Adrian,

Once again re-calibrated the gauge. The calibration process was:

1. Connected a load to draw 1A current from a DC voltage source.

2. Entered this 1A current value i.e. 1000mA in "Applied current" section under Calibration tab on BQ Studio.

3. Selected "Calibrate current" and then "Calibrate Gas Gauge".

Now, repeated the test by 2 % discharge of battery and followed by a rest time (Rest time has been 2 days due to holidays).

Observations after 1sec pulsed load:

1. SOH_Merit bit is set high.

2. Current is reported as 3294mA even though my discharge current is 260mA (2% discharge).

3. First time Measured Z value changed from -32768 mOhm to some different value i.e. -60 mOhm.

Attached herewith are the Waveforms before, during and after 1 sec pulsed load. The waveforms are smooth and pure DC of 3.6V. Only a small obvious change in voltage is observed while turning pulsed load on/off. 

Please analyse these test results and suggest on how to proceed further. Is this value of -60mOhm promising ?

With regard,

Sudarshan Chaudhary

Hello Adrian,

Hope you have analyzed the above test results. Please suggest on to proceed further.

With regard,

Sudarshan Chaudhary

Hello Adrian,

Looking forward to your response regarding the above issue. Please help us to troubleshoot this error.

With regard,

Sudarshan Chaudhary

Hello Sudarshan,

Sorry for taking a while to get back to you. I think the -60 is promising. Since you didn't see the EOS_BAD_OCV bit being set, I think this should be going in the right direction. Continue the testing and keep gathering data.

Regards,

Adrian