R_a values of zero

Sebastian Manz

Other Parts Discussed in Thread: BQ20Z80

Hello everybody,

I created a golden Image for a bq20z80 batterypack. After the learning cycles completed, I noticed that some R_a values in the .gg-file were zero: e.g. "Cell1 R_a 0 = 0".

I wonder if such values could be correct. From what I know of lithium-ion batteries, they have impedances > 0 mΩ in the usually examined frequency regions. So are these zero values only a matter of the underlying algorithm and perfectly fine? Or could they lead to problems in the gauging calculations later on?

At the moment, the SOC values I receive from the battery seem reasonable that is.

Thanks,

Sebastian

over 9 years ago

0 ThomasCosby over 9 years ago

TI__Guru 73830 points

Ra table values should not be equal to zero. Can you attach the log file from your optimization cycle? I can check it with a Mathcad program.

0 Sebastian Manz over 9 years ago in reply to ThomasCosby

Intellectual 470 points

Hello Tom,

unfortunately I did not log the optimization cycle, but I can repeat the procedure, because I have an image of the flash before the cycle. What logging interval should I choose? Should I start logging when charging the pack to 100% or should I start when discharging to 0%?

Thanks,

Sebastian

0 ThomasCosby over 9 years ago in reply to Sebastian Manz

TI__Guru 73830 points

Yes, log during the charge cycle as well. Is this problem with the bq20z80 as well?

0 Sebastian Manz over 9 years ago in reply to ThomasCosby

Intellectual 470 points

Yes, this problem appears together with bq20z80.

0 ThomasCosby over 9 years ago in reply to Sebastian Manz

TI__Guru 73830 points

I checked the Ra table values in the senc file that you sent and they look normal in one table, but low in the other table. Do you have a discharge profile logged from one of your runs? How did you verify that the ChemID was a good choice?

0 Sebastian Manz over 9 years ago in reply to ThomasCosby

Intellectual 470 points

I'll run and log the discharge profile, but it will take some time.

I chose a ChemID which is exclusively for the Samsung cells in this design (chem id 2019) according to the description text. I flashed it with "bqEasy" over the default senc file with chem id 0100. The SOC values I receive seem plausible for different pack/cell voltages.

0 Sebastian Manz over 9 years ago in reply to ThomasCosby

Intellectual 470 points

Hello Tom,

I have made three logging runs with two different packs and not a single Ra value was equal to zero. Maybe I messed up the calibration in the cycle with the zero values... I'm probably using the new values for the golden image.

What could happen if the Ra values were wrong? Would this only affect the SOC calculation?

Thanks,
Sebastian

0 ThomasCosby over 9 years ago in reply to Sebastian Manz

TI__Guru 73830 points

Yes, incorrect Ra values should affect capacity estimates and SOC values.

Power management

Power management forum

R_a values of zero