BQ27426: Questions about Golden Learning

#1: This ensures high quality initial Qmax. The gauge disables several Qmax checks during a learning cycle hence the gauge requires a large change in DOD to qualify the measured passed charge for Qmax updates during a learning cycle.

#2: This is outside the scope of a true learning cycle (which is intended for a new cell, not a severely aged cell). You'd have to change the 90% limit but I don't recommend doing this because it can adversely affect Qmax accuracy.

HI Dominik，

How can I change the 90% parameter?

Dominik Hartl11 said:

#2: This is outside the scope of a true learning cycle (which is intended for a new cell, not a severely aged cell). You'd have to change the 90% limit but I don't recommend doing this because it can adversely affect Qmax accuracy.

如何修改90%这个参数？

Hello Dominik，

I'm very sorry, still want to know the detailed reason for needing passedcharge>Design Capacity*90%.

Please help to elaborate:.

Why does passedcharge need to be more than Design Capacity*90% for Golden Learning?
What is the reason?

Dominik Hartl11 said:

#1: This ensures high quality initial Qmax. The gauge disables several Qmax checks during a learning cycle hence the gauge requires a large change in DOD to qualify the measured passed charge for Qmax updates during a learning cycle.

非常抱歉，还是想知道需要passedcharge>Design Capacity*90%的详细原因。

请帮忙详细说明下:

为什么Golden Learning时，passedcharge需要超过design Capacity*90%？

Hello Dominik，

1, Bq27426 does not force 90%:
The default requires passedcharge>Design Capacity*90%, but the 90% limit can be modified via registers.
Is this understood correctly?

2、How to modify the limit?

1、Bq27426不强制90%，是否可以理解为：默认需要passedcharge>Design Capacity*90%，但可以通过寄存器修改90%的限制？

2、如何修改该限制？

The 90% is what TI requires for a dedicated learning cycle. The bq27426 only enforces the 37%. There is nothing to modify.

Hello Dominik，

Back to the problem at the beginning:

Inserting a battery with SOH=73%, loading the Fs file (reserve capacity 350mAh), and again failing to learn successfully via Golden learning.

Is it possible to make Fuelgauge learn the FCC of a battery with SOH=73% by cycling through multiple rounds of full charge (SOC100%) full discharge (SOC0%)?

回到一开始的问题：

插入SOH=73%的电池，加载Fs文件(储备容量350mAh)，再次通过Golden learning无法学习成功。

是否可以通过多轮循环的满充(SOC100%)满放(SOC0%)来使Fuelgauge学习到SOH=73%电池的FCC？

Wang Dong said:

2、Using the aging battery SOH=73% for Golden Learning, the passedcharge<design capacity*90%, what can be done to solve this problem and make Golden Learning successful?