BQ20Z95: please check the dfi file, and how to make gg file.

Hello Wyatt,

finally we got gg file.

could you check this gg file whether it is suitable or not?

i also attach the specification of battery.

Best regards,

Chase

EGEN.pdf

C3 210210.zip

Hello Wyatt,

1. I completed learning cycle on this pack. and actually it is not a new pack. it assembled completely once, and be written previous gg file(have some problems with SOC jump). and we start again learning cycle in bqSW program with default dfi file.

2. i searched my previous post with you, i could not find about Ra tables have large resistances. could you explain more about this?

3. protections/battery specifics what you said means like undervoltage/overvoltage protection and capacity, charging current and so on?

Best regards,

Chase

Hello Wyatt,

thank you for your reply.

in fact i just run a learning cycle after the pack had been in the field.

and that pack had previous golden image(that have a problem with SOC calculation).

anyway if it looks like not all the Ra values were updated, should we do learning cycle from the beginning? or do i have another way to fix it?

if yes, can you guide me about the expected range of Ra values?

Thank you in advance.

Best regards,

Chase

Hello Wyatt,

i am sorry that i have too many questions, cause this is the first time of my learning cycle, both me and my customer.

1. what do you mean "original golden image"? is it the golden image which i first uploaded to my battery?

if so, i think there is also have a problem with Ra tables. please check the file below.

C3_150325.zip

2. if i modify some parameters in gg file such as CUV threshold, CUV recovery, Cell Shutdown Voltage, then should I do learning cycle again from the beginning? or just modify and upload it directly to the battery?

3. is there any default gg file that have only initial data of the pack, and default Ra table? how can I get it?

please check this issue. Thanks.

Best regards,

Chase

Hello Wyatt,

Thank you for your reply.

1. could you explain more about Ra tables? what is the function of Ra table and what would go wrong if all Ra tables are not updated?

if you have a document about it, please attach it.

2. can I make a golden pack with only basic parameter (voltage, current, cell structure, etc) and default DFI file, without learning cycles?

Best regards,

Chase

Hello Wyatt,

long time no see.

my customer did another learning cycles again, and made new gg file.

could you check this gg file is good or not?

C3_210303.zip

and i check the Ra tables you said, it definitely changed(actually all values are reduced) after 2 learning cycles from default values.

but i still don't know, how can I verify that all Ra values are not updated? can you give me some advice for this?

and I have one question, can I rest more time after the value is changed enough for nest step during leaning cycle?

cause as you know, learning cycles need a lot of time, especially in the end of each step we have to wait to see the changes of some parameters, and need more time if that time is middle of the night. we can start next step in the morning.

please check these issues. Thanks.

Best regards,

Chase

Hello Chase,

It looks like cell3 resistance values have updated but the others are the same as before. Do you have any logs of the learning cycle/schematic of the connections?

It's strange that only one cell has updated.

It is okay to let it relax longer than specified, that's just the minimum.

Sincerely,

Wyatt Keller

Hello Wyatt,

Here is my log during learning cycle. could you Please check it?

EGEN_210304.zip

Best regards,

Chase

Hello Wyatt,

we are still in testing with latest gg file, but have some troubles.

1. we tried several times to get chemistry id, but we had each different id we have tried. what do you think is the reason? and can I use anything of them? (we got 1169, 1229, 3589, 2032. they have different capacity.)

2. after I write latest gg file, each cell capacity is like the picture below;

after 2 hours later, the cell capacity changed. not similar, each have so different values.

why is it happen? did I something wrong while I do my learning cycle?

3. last time you said about using Ra table  for single cell to another. is it mean that I can copy the Ra table to another cell with same value?

4. how did you calculate MaxError set to 1%? is it from learning cycle data or gg file?

5. this is silly question, but i am asking you. we have just 3 cells in serial. but I can see the Cell0, 1, 2, 3 in gg file. is it normal?

I am sorry for too many questions. please understand that this is our first learning cycle.

Best regards,

Chase

Hello Chase,

You should use the chem ID recommended by the GPCCHEM tool. It should only recommend 1 chem ID with the best match, this should be used.

Your cells are very imbalanced, they should not become that imbalanced from 2 hours of rest, do you have a load on the pack?

Sincerely,

Wyatt Keller

Hello Wyatt,

1. of course I got one chem ID if I upload same data, but I did several learning cycles to make gg file in bq evaluation software (old version), they have little different results each. and every learning cycle got different chem ID. i am little confused what should I use,

2. I did not connect the load on the pack. just upload new gg file, and rest 2 hours. what should I doubt? schematic or the process of learning cycle? or something else? 

3. could you answer for rest of my questions?

Hello Wyatt,

I made new gg file with my customer. could you check this gg file is okay?

EGEN_210406.zip

I also attach chemID report. please check it too.

EGEN_210325-report.zip

Best regards,

Chase

Hello Wyatt,

Thanks to your support.

you mean that I can use R_a3 (Ra Table) for all the Ra Tables?

[R_a3(Ra Table)]
Cell3 R_a flag = FF55
Cell3 R_a 0 = 88
Cell3 R_a 1 = 91
Cell3 R_a 2 = 102
Cell3 R_a 3 = 99
Cell3 R_a 4 = 97
Cell3 R_a 5 = 103
Cell3 R_a 6 = 107
Cell3 R_a 7 = 124
Cell3 R_a 8 = 131
Cell3 R_a 9 = 147
Cell3 R_a 10 = 155
Cell3 R_a 11 = 201
Cell3 R_a 12 = 260
Cell3 R_a 13 = 413
Cell3 R_a 14 = 1068

Can I apply this values to Cell0~2 and xCell0~2? all the same values?

how about each R_a flag? should I leave this value as its original value? or should I change it all the same as FF55 which mentioned the table above?

please check this issue. Thanks

Best regards,

Chase

Hello Wyatt,

I am very sorry, but I cannot understand clearly what you said.

you mean that I can use R_a3 (Ra Table) for all the Ra Tables? or should I do learning cycle from the beginning?

please check this issue. Thanks.

Best regards,

Chase

Hello Wyatt,

I am sorry, I have one more question.

how about the R_a flag in each field? should I enter the same value as cell 3 value(FF55)? or should it remain as each original value?

Best regards,

Chase

Hello Wyatt,

we tested the modified gg file as you suggested, the capacity seems about half compared to normal. (about 1318mAh)

left picture is writing modified gg file(capacity is 1318mAh), right picture is writing gg file which just set the default, not did learning cycle. (stopped right before learning cycle, and made gg file - capacity is 2459mAh)

could you advice something for this matter?

Best regards,

Chase

Hello Chase,

When you changed the Ra values the FCC dropped? The one with more capacity is with the higher Ra values?

Reducing the Ra values should increase FCC because there is less resistive losses during discharge. FCC is based on a simulation using application current and Ra values.

Sincerely,

Wyatt Keller

Hello Wyatt,

When you changed the Ra values the FCC dropped? The one with more capacity is with the higher Ra values?

=> yes, I changed Ra values as you mentioned, and wrote it to new battery, FCC dropped as the screenshot.

Can you guess what is the reason?

Best regards,
Chase