BQ40Z80: BQ20Z45 Learning failed, VOK never clear

user5767698

Part Number: BQ40Z80

Is there any one can help for me?
I have done two failed learning cycle with BQ40z80,when I finshed the learning cycle and relax for a at least 12 hours i found that VOK never clear and LStatus() is still 0x05.
1. My cell type is TOSHIBA 20Ah SCiB, I use the chemical ID 0700, is that correct?
2. I follow below step to do the learning cycle, and stop learning when failed at step9. Please see my log data attached.
Step1: Program with correct chemical ID, correct data flash.
Step2: Calibrate the battery pack.
Step3: Discharge the battery pack to empty with C/10 (voltage down to term).
Step4: Wait 5 hours.
Step5: The [VOK] and [RDIS] bits in the IT status() register clear, GaugingStatus[REST] flag is set .
Step6: Charge pack to full with C/4 until the taper current is reached and the FC bit is set.
Step7: Let it relax for 5 hours.( Update status 05)
Step8: Discharge the pack down to term voltage with C/10.
Step9: Let it relax for 12 hours.( Update status 05,vok not clear)1256.learning cycle.zip

over 6 years ago

0 Steve Schnier over 6 years ago

TI__Expert 4645 points

I just wanted to let you know I am looking at this and expect to get back to you tomorrow on this.

0 user5767698 over 6 years ago in reply to Steve Schnier

Intellectual 350 points

Thanks for your reply.

0 Steve Schnier over 6 years ago in reply to user5767698

TI__Expert 4645 points

Reviewing your data, the discharge current is below C/10, which will disqualify updates. For the learning cycle, C/5 is recommended. There is a caution about using lower than C/10 in the app note in section 4.2.5 Discharge Battery to Empty:
• A typical C/5 rate is recommend, but the rate can be as low as C/10. If using a C/10 load, make sure the gauge sees that the current is at least C/10, if the current is any lower, resistance updates can not occur.

0 user5767698 over 6 years ago in reply to Steve Schnier

Intellectual 350 points

I have done two failed learning cycle，the discharge current used in first time was C/5，but it also failed at step9.The log data as the attachm0.2c learning cycle.zipent.

0 user5767698 over 6 years ago in reply to Steve Schnier

Intellectual 350 points

I have done two failed learning cycle，the discharge current used in first time was C/5，but it also failed at step9.The log data as the attachment.

0 Steve Schnier over 6 years ago in reply to user5767698

TI__Expert 4645 points

Can you also send the .gg file that you used for the learning cycle?

0 user5767698 over 6 years ago in reply to Steve Schnier

Intellectual 350 points

4S SCiB.gg.zipthanks for your help,The .gg file as the attachment.

0 Steve Schnier over 6 years ago in reply to user5767698

TI__Expert 4645 points

Hello,

After looking through the log file and the .gg file, the Chem ID is 700 for a 20Ah Li-titanate cell, and your design capacity is set to 20Ah. However, your log file only charges 4278Ah, and then discharges 4047Ah. You need to have a full charge / discharge of at least 90% of the design capacity. You currently are only charging / discharging 20% of the design capacity.

In your log file, the discharge current is between 1200mA and 1000mA during discharge, which is well below the 2A needed for C/10 or 4A needed for C/5 (recommended). The log file is clearly not showing a 0.2C discharge rate as you expect.

I'd be happy to look at the datasheet of the cell you are using so I can verify that the settings match the cell.

Please re-run with the correct charge voltage and current, and then the correct discharge current and let me know if you have issues.

0 user5767698 over 6 years ago in reply to Steve Schnier

Intellectual 350 points

hi,steve:

The charge and discharge current is in the recommend range,beacuse of the Scale Factor is set to 5x, so the actual discharge current is between 5A--6A,and the charging / discharging capacity meets the requirements.The datasheet of the cell as the attachment. I have another question,if the Scale Factor was set to 5,the design capacity should be set to 20AH or 4AH?The cell actual capacity is 20AH.

0 Steve Schnier over 6 years ago in reply to user5767698

TI__Expert 4645 points

I think we now know the issue. Yes, all current, capacity, and protection values need to be scaled in the gauge. This is in the bq40z80 Technical Reference Manual in section 6.7 Current Scaling. It states: "The DesignCapacity mAh and Design Capacity cWh must be scaled in data flash, and all current and capacity parameters in the data flash are divided by Scale Factor." as well as "The MAC command CurrentLong() (0x0082) is provided to report current greater than 32767 mA in mA resolution. All other currents and capacities are reported as scaled values, and the host must multiply the value by the Scale Factor. All protections must also be divided by the Scale Factor."

If you re-run the learning cycle with the appropriately scaled values, I believe it should work for you.

0 user5767698 over 6 years ago in reply to Steve Schnier

Intellectual 350 points

Thanks for your help,i will re-run the learning cycle and tell you the result!

0 user5767698 over 6 years ago in reply to Steve Schnier

Intellectual 350 points

I have re-un the learning cycle and failed again.I have modified the design with correct values, we were not reading an update status of 06 after completing the 5+ hour rest , the update status still stays at 05. I didn't know what to do.Another strange question is the chemical ID recommend by GPC TOOLS is is not correct，the cell i used is SCiB，but the ID is for LiCoO2. The scale factotr == 10.

FW.zip

learning cycle_10.29.zip

4682.GPCPackaged-report.zip

0 Steve Schnier over 6 years ago in reply to user5767698

TI__Expert 4645 points

The CHEM ID not matching the chemistry may also cause the disqualification of the Qmax updates. I will follow up with you separately for next steps.

0 Steve Schnier over 6 years ago in reply to Steve Schnier

TI__Expert 4645 points

The GPC report has this conclusion: Warning: All IDs disqualified because of too high resistance deviation. No suitable ID selection available.

This is likely due to formatting or gaps in the data submitted to the tool.

We recommend CHEM ID 700 for your battery chemistry and capacity. Using this CHEM ID should not disqualify the Qmax updates based on the delta Ra being too great.

Power management

Power management forum

BQ40Z80: BQ20Z45 Learning failed, VOK never clear