BQ34Z100-R2: Full Charge Capacity not stable after learning cycle

Hi,

Did you use the GPCCHEM tool to get a chemID match with DOD% error of 3% or less?

You said you completed a learning cycle, can you verify learned status is 0x06.

Regards,

Evan

Yes, the Learned Status register is 0x06. It went from 0x04 -> 0x05 -> 0x06 as described in https://www.ti.com/lit/ug/sluuco5a/sluuco5a.pdf

I will reconfirm the chemID.

Hi,

Please share the GPCCHEM output report if possible.

Regards,

Evan

It looks like we were indeed using the wrong chem id and got this far with it. I've attached the report from the tool. The tool recommends a different one with 12.6% error. I'll run the chem id check again to confirm the results and test how this performs.

What sort of performance issues do you foresee if the error is not under 3% as you recommended, or is that classified information regarding your algorithm and gauge?

Chemistry ID selection tool, rev=2.54		
		
Configuration used in present fit:		
ProcessingType = 2		
NumCellSeries = 4		
ElapsedTimeColumn=0		
VoltageColumn = 1		
CurrentColumn = 2		
TemperatureColumn = 3		
		
Best chemical ID : 475	Best chemical ID max. deviation, % : 12.6	
		
		
		
Summary of all IDs with max. DOD deviation below 15%		
		
Chem ID	max DOD error, %	Max R deviation, ratio
475	12.6	6.17
400	14.48	0.45
		
Max. deviations for best ID is within recommended range. Chosen best chemical ID is suitable for programming the gauge.		
		
		
Selection of best generic ID for ROM based devices like bq274xx		
		
		
Device / Family #1		
Generic Chem ID	Device/ Voltage/ Chemistry	max DOD error, %
128	bq27421-G1A: 4.2V LiCoO2	28.66
3142	bq27421-G1D: 4.4V LiCoO2	29.15
354	bq27411-G1C: 4.35V LiCoO2	29.34
312	bq27421-G1B: 4.3V LiCoO2	97.56
Best generic ID 128		
Warning: Generic ID Deviation is so high that it is most likely due to anomaly in the data. Please check that data files have recomended format, units and test schedule		
		
		
Device / Family #2		
Generic Chem ID	Device/ Voltage/ Chemistry	max DOD error, %
354	bq27621:  (ALT_CHEM2) 4.35V LiCoO2	29.34
1202	bq27621: (default) 4.2V LiCoO2	29.96
1210	bq27621:  (ALT_CHEM1) 4.3V LiCoO2	34.17
Best generic ID 354		
Warning: Generic ID Deviation is so high that it is most likely due to anomaly in the data. Please check that data files have recomended format, units and test schedule		
		
		
Device / Family #3		
Generic Chem ID	Device/ Voltage/ Chemistry	max DOD error, %
3230	bq27426: (default) 4.35V LiCoO2	26.62
3142	bq27426: (ALT-CHEM2) 4.4V LiCoO2	29.15
1202	bq27426: (ALT_CHEM1) 4.2V LiCoO2	29.96
Best generic ID 3230		
Warning: Generic ID Deviation is so high that it is most likely due to anomaly in the data. Please check that data files have recomended format, units and test schedule		
		
		

Hi,

What is the chemistry of your battery?

Can you share the data you used to create this report, or if you are going to redo the discharge cycle, please share the new data.

This document may be useful - Simple Guide to Chemical ID Selection Tool (GPC) (Rev. A)

Regards,

Evan

If you use a ChemID with this much error, it could lead to SOC errors and SOC jumps.

Regards,

Evan

The battery consists of unidentified LiFePO4 3.2V 4.2Ah cells.

Here's the first data set:

chem_id_02_15.zip

And the second one:

chem_id_02_20.zip

And here's the report for the second data:

Chemistry ID selection tool, rev=2.54		
		
Configuration used in present fit:		
ProcessingType = 2		
NumCellSeries = 4		
ElapsedTimeColumn=0		
VoltageColumn = 1		
CurrentColumn = 2		
TemperatureColumn = 3		
		
Best chemical ID : 400	Best chemical ID max. deviation, % : 13.94	
		
		
		
Summary of all IDs with max. DOD deviation below 15%		
		
Chem ID	max DOD error, %	Max R deviation, ratio
400	13.94	4.56
		
Max. deviations for best ID is within recommended range. Chosen best chemical ID is suitable for programming the gauge.		
		
		
Selection of best generic ID for ROM based devices like bq274xx		
		
		
Device / Family #1		
Generic Chem ID	Device/ Voltage/ Chemistry	max DOD error, %
354	bq27411-G1C: 4.35V LiCoO2	30.31
3142	bq27421-G1D: 4.4V LiCoO2	31.17
128	bq27421-G1A: 4.2V LiCoO2	37.24
312	bq27421-G1B: 4.3V LiCoO2	97.56
Best generic ID 354		
Warning: Generic ID Deviation is so high that it is most likely due to anomaly in the data. Please check that data files have recomended format, units and test schedule		
		
		
Device / Family #2		
Generic Chem ID	Device/ Voltage/ Chemistry	max DOD error, %
354	bq27621:  (ALT_CHEM2) 4.35V LiCoO2	30.31
1202	bq27621: (default) 4.2V LiCoO2	30.86
1210	bq27621:  (ALT_CHEM1) 4.3V LiCoO2	33.08
Best generic ID 354		
Warning: Generic ID Deviation is so high that it is most likely due to anomaly in the data. Please check that data files have recomended format, units and test schedule		
		
		
Device / Family #3		
Generic Chem ID	Device/ Voltage/ Chemistry	max DOD error, %
3230	bq27426: (default) 4.35V LiCoO2	27.91
1202	bq27426: (ALT_CHEM1) 4.2V LiCoO2	30.86
3142	bq27426: (ALT-CHEM2) 4.4V LiCoO2	31.17
Best generic ID 3230		
Warning: Generic ID Deviation is so high that it is most likely due to anomaly in the data. Please check that data files have recomended format, units and test schedule		
		
		

In the second data set, I tried to discharge the battery closer to the 10V/2.5V per cell limit. This apparently dropped the previous id suggestion 475 from the recommendation or it was dropped due to other variations. We assume that the lower voltage limit for these cells is 2.5V but we can't reach it exactly since at that point the battery's own protection will cut the power and disrupt the data.

Hi,

I had previously overlooked that you said that you are using an LiFePO4 battery. These batteries have a long and flat ocv curve and are difficult to match perfectly.

I recommend proceeding with the 12.6% error ChemID and enabling LFPRelax bit.

(+) BQ34Z100-G1: Pack Configuration B Register Parmeter LFPRelax - Power management forum - Power management - TI E2E support forums

Are you using EV2400 + BQstudio?

Regards,

Evan

Alright. I'll test how it behaves. And yes, we're using ev2400 and bqstudio.

Hi,

Please share a bqstudio log file moving forward. :)

Regards,

Evan

I did include the log from bqstudio with the capacity drop in the initial post. Did it not get added properly? It should be the last thing in it. Or do you mean something else?

Hi,

Yes, the original log file shared was correct, the log files shared after that appear to be missing data.

Regards,

Evan

The 0475 chemistry with 12.6% error didn't end up working. During the discharge cycle, the RUP_DIS flag was set when it was updating the ra table. Should I try the 0400 chem id that had 14.5% and 13.9% errors? Or does the error mean that it will also automatically fail?

I've attached the full logs from bqstudio. They ended up being in two parts since the tool was needed briefly elsewhere.

02-26 failed discharge cycle bqstudio logs.zip

Hi,

Your learned status is 0x05?

how long did the battery rest for after discharging?

Is this the chemID you had previously completed a successful learning cycle with?

Regards,

Evan

-Your learned status is 0x05?

I did not reset that register. It's still 0x06 from before. Before the calibration attempt, I ran the RESET command from bqstudio which cleared at least the ra table values.

-how long did the battery rest for after discharging?

The battery was:

• Discharged
• Left to rest for 5 hours
• Charged to full in 17 hours
• Left to rest for 98 hours
• Started to discharge and while discharging the RUP_DIS flag was set indicating an incorrect chem id as per the Technical Reference Manual:
  https://www.ti.com/lit/ug/sluuco5a/sluuco5a.pdf?ts=1703775513458&ref_url=https%253A%252F%252Fwww.ti.com%252Fproduct%252FBQ34Z100-R2#%5B%7B%22num%22%3A253%2C%22gen%22%3A0%7D%2C%7B%22name%22%3A%22XYZ%22%7D%2Cnull%2C717%2Cnull%5D
  6. During discharge, the resistance table may update for a while and then stop. When this happens, RUP_DIS
  is set. This indicates that the Chemistry ID choice is incorrect. This means that the gauge has measured a
  resistance value that just does not make sense (that is, is negative). A chemistry cycling is needed to identify
  the correct chemistry profile.

-Is this the chemID you had previously completed a successful learning cycle with?

No. I was trying to run the calibration for the new chem id 0475 that you recommended to try based on the GPCCHEM tool's report:

Diego said:

I recommend proceeding with the 12.6% error ChemID and enabling LFPRelax bit.

And yes, the LFPRelax bit in Pack Configuration B is enabled.

So your guess is that the gauge didn't actually achieve a proper 0x05 learned status and that's why the ra update indicated an incorrect chem id? Does achieving the 0x05 learned status depend on the selected chem id and how does the LFPRelax bit change that part of the calibration?