BQ27520-G4: Learning cycle was fail

I'm sorry that it is very late to reply.

I've confirmed SLUA903 carefully.

But I couldn't understand how to do Learning Cycle.

So could I confirm the procedure step-by-step?

At first, about 1st Discharge Battery to Empty.

1.Do I need to send any command to IC?

 For example, 0x21. In Figure 3. it seems that sending this command is after Relax for 5 Hours.

2.Discharge condition is that current is 274mA, discharge it until the voltage reaches 3.0V. My Terminal voltage setting is 3.0V.

 Is it correct?

Best Regards,

Yuichi Tsutsui

Dear Wyatt

I understand sending it is after relaxation.
After sending command(0x21), would update status set 04 automatically?
And which update status is it?
I think there are 4 update status as below.
 -Update Status0: Gas Gauging->State
 -Update Staus1:Gas Gauging->State
 -Update Status:OCV Tables->OCVa0 Table
 -Update Status:OCV Tables->OCVa1 Table

I have a questions about the design capacity too.
SLUA903 says that this can be copied from the battery pack datasheet.
Our battery's datasheet says this is 2700mAh.
So is it correct that design capacity is 2700mAh?
And first discharge current should be more than 270mAh?

Best Regards,

Yuichi Tsutsui

Dear Wyatt

I will try to do learning cycle with Design capacity 2700mAh, and discharge current 270mA.

I will share the result once I get it.

-About update status
 There are some update status. I'm sorry that those are making me confused.
 Which one in EVM is same with SLUA902 and SLUA903.
 Please refer to the attached material.

-About Design capacity
I have experience that when I change the design capacity to the datasheet capacity in other applications, the operating time gets shorter.
Specifically, when operated under the following conditions, the operation time after the second time became shorter.
    Nominal Capacity:2300mAh (8.4V charge and 0.46A discharge) <- from battery pack specification
    Nominal Capacity:2050mAh(8.2V charge and 0.46A cischarge)  <- from battery pack specification
    Set Design capacity:2050mAh
This is because the RM value was fixed at 2050mA, I think.

After that, when the design capacity was set to 2300 mAh, such a thing did not occur.
And RM value became between 2050 to 2300mAh.

So I set the design capacity to a value larger than the datasheet capacity, but what do you think about these?

About BQ27520_G4.pptx

Thanks

Best Regards,

Yuichi Tsutsui

Dear Wyatt

Thank you for your reply.

From today, I try to do learning cycle again with design capacity 2700mAh, and discharge current 270mA.

After relaxing in five hours, I'll check these update status and share them.

Please wait once I've confirmed them.

Thanks

Best Regards,

Yuichi Tsutsui

Dear Wyatt

I've finished 1st discharge and relax in 5 hours.

I'll share the following logs (attached zip file). Please confirm them whether there are some mistakes.

   after_1st_relax_in_5H_200630.gg.csv: after 1st discharge and relax in 5hours

   after_1st_relax_in_5H(after_send_0x21)_200630.gg.csv: After sending command 0x21

   after_1st_relax_in_5H(after_send_0x21)_200630_EVM.png: At starting to charge battery EVM capture

   learning cycle_200630.log: log file from EVM from 1st discharge and relax.

01_1st_Discharge_Relax.zip

Thanks

Best Regards,

Yuichi Tsutsui

Hello Yuichi,

The bqStudio dashboard looks correct to start the learning cycle, you have a valid OCV reading and Ra table updates are enabled. Were you able to complete the learning process successfully?

Sincerely,

Wyatt Keller

Dear Wyatt

I've finished the 1st charge and relax in 2 hours.

Just now I'm starting the 2nd discharge.

But the Ra value and update state don't seems update.

The attached files are the following log.

Could you confirm whether the leaning cycle is success?

    after_1st_charging_relax.gg.csv: after 1st charge and relax in 2 hours

    after_1st_charging_relax.png:after 1st charge and relax bqStudio dashboard

    Just_after_discharge.gg.csv: get it from bqStudio after starting to discharge 

    Just_after_dicharge.png: bqStudio dashboard after starting to discharge 

    learning cycle_200630.log: log file from EVM from 1st discharge to 1st charge and relax

02_1st_charge_Relax.zip

I'm not sure the learning cycle is success or not.
Could you tell me which part I should confirm to check it?

Dear Andy

Thank you for your confirmation.

I try to do the learning cycle with changing Qmax Cell 0/1/  to 2700mAh.

If you have the concern about my gg.csv, please let me know.

The attached file is updated gg.csv.

Thanks

Best Regards,

Yuichi Tsutsui

Flight5_SP_FS_Ver08_200630.gg.csv

Could you tell me the charging voltage you see on the battery specification?   Thanks. 

Andy

Dear Andy

It is 8.2V.

Thanks

Best Regards,

Yuichi Tsutsui

Hello Yuichi,

Let us know if the updated Qmax values let you get a successful learning cycle. Your charging voltage looks correct for the battery spec, what chemID are you using? did you use the GPCCHEM tool to find it?

Sincerely,

Wyatt Keller

Dear Wyatt

Yes, I used the GPCCHEM tool.

I attached GPC report.

I apologize that I confirm the battery datasheet again and Nominal Capacity is 2740mAh.

So I changed the design capacity and Qmax value.

Then I started to discharge the battery from yesterday.

Now I completed the discharge and relax in 5 hours. And I'm starting the 1st charging after sending command 0x21.

The attached files are the followings.

If you have any concern about them, please let me know.

-Attached material

  GPC_Tool_200531-report.zip:The report from GPCCHEM tool

  1st_discharge_relax.gg.csv:gg file after 1st discharge and relax in 5 hours

  1st_discharge_relax.png:bqStudio dashboard after 1st discharge and relax in 5 hours

  1st_discharge_relax_after sending 0x21.gg.csv:gg file after sending command 0x21

  1st_discharge_relax_after sending 0x21.png:bqStudio dashboard after sending command 0x21

  At the beginning of 1st charging.png: bqStudio dashboard at the beginning of 1st charging

  learning_cycle_200701.log: the log file getting from bqStudio between the beginning of 1st discharge and the relaxing in 5 hours.

01_1st_discharge_and_relax.zip

Thanks

Best Regards,

Yuichi Tsutsui

Hello Yuichi,

The dashboard looks correct to start the learning cycle, just continue to make sure you follow the procedure in the quickstart guide.

Sincerely,

Wyatt Keller

Dear Wyatt

Thank you for your confirmation.

I've completed the 1st charging and relax in 5 hours.

I'm starting the 2nd discharging.

Before starting the discharging, I confirmed gg file.

But it doesn't seems to change from the beginning one.

Could you confirm it is correct?

The attached files are the followings.

  -after charging and relax.png: bqStudio dashboard after charging and relax in 5 hours.

  - after charging and relax.gg.csv: gg file after charging and relax in 5 hours

  -at the bigining of discharge.png: bqStudio dashboard at the beginning of discharge after charging and relax

  -at the bigining of discharge after chargind.gg.csv: gg file at the beginning of discharge after charging and relax

  -learning_cycle_200701.log: getting from bqStudio from now.

02_1st_charging_and relax.zip

Dear Wyatt

I've completed the 2nd discharge and relax.

The attached materials are the following ones.

Please confirm them.

  -after_2nd_discharge_and_relax.png: bqStudio dashboard after 2nd discharge and relax in more than 2 hours.

  - after_2nd_discharge_and_relax.gg.csv: gg file after 2nd discharge and relax in more than 2 hours.

  -learning_cycle_200701.log: getting from bqStudio from now.

03_2nd_discharge.zip

Ra value and update state didn't change during this learning cycle.

Is this success?

Could you confirm whether the learning cycle is success or not?

Hello Yuichi,

It looks like it was not successful, you can check this in the update status to make sure you update from 0x00 to 0x01 after the first charge cycle, if it does not do this you don't need to continue with the learning cycles.

Do you have some other load on the battery when at rest? It looks like your voltage is dropping while there is no load, this could cause the gauge to never relax. If you didn't have enough time between charge and discharge the timer may not have reached the forced time either.

Sincerely,

Wyatt Keller

Dear Wyatt

Thank you for your confirmation.

I see.

Some devices are connecting to the battery which is driving at rest.

So I will conduct the learning cycle with these device removed.

But some IC and resistance couldn't be removed because these are needed for charging IC and battery management IC(BQ27520).

Detail is the attached material and is it acceptable?

Connecting to the battery_200707.pptx

Dear Wyatt

Thank you for your reply.

I did 1st discharge without removing some material which are the ppt. sending by me before.

Just now the battery is relaxing in 4 hours after discharge, so I'm waiting until the relaxing time is more than 5 hours.

I confirmed the log file and I think the battery voltage doesn't seem to drop down. Is it OK?

Could you confirm the following materials?

   -after_1st_discharge_200708.gg.csv: gg file after discharge and relax in about 4 hours

   -learning cycle_200707.log: log file getting from bqStudio from the starting to learning cycle and now.

And if I need to remove other parts, please let me know

01_1st_discharge.zip

Thanks

Best Regards,

Yuichi Tsutsui

Hello Yuichi,

This looks better. Continue with the rest period and then charge to check if the update status goes from 0x00 to 0x01, if it does then continue to the final discharge, if not there may be a different problem.

Sincerely,

Wyatt Keller

Dear Wyatt

Thank you for your reply.

I'm happy to hear that.01

The battery is relaxing in all day yesterday. This is because the voltage of it is getting up slightly.

So I'm starting to charge the battery after checking your message.

The attached material is the following ones.

I will check if the update status goes from 0x00 to 0x01 after completing the charging.

But I have a question about it.

Which is this update status you mean?

In my understanding, there are four of them in gg file (Update Status0, Update Status1, Update Status(OCVa0), Update Status(OCVa1)).

   -after_1st_discharge_and_relax_200709.gg.csv:gg file after discharge and relax in very long time
   -Just starting to 1st charge_200709.png:bqStudio dashboard at the starting 1st charging
   -learning cycle_200707.log:log file getting from bqStudio from the starting to learning cycle and now.

6052.01_1st_discharge.zip

Thanks

Best Regards,

Yuichi Tsutsui

Dear Wyatt

I completed the 1st charge and relax.

I confirmed gg file after relaxing.

But all of update status is 0x00. So learning cycle may be fail.

Is it correct? Please confirm the following materials.

And what do you think is the reason for fail?

 02_1st_charge.zip

Thanks

Best Regards,

Yuichi Tsutsui

Hello Yuichi,

I will look through your logs and try to find any problems next week.

Quick question: how much longer after the charge did you pull the .gg file? Was it long enough for the update status to trigger?

Sincerely,

Wyatt Keller

Dear Watty

Thank you for your reply.

I did that about 16 hours after charging.

I'm keeping to logging date in this weekend. So I attached the log file and gg.file which I got this morning.

Please confirm them.

200713.zip

Thanks

Best Regards,

Yuichi Tsutsui 

Dear Wyatt

I apologize I'm late to reply.

My final voltage is higher than terminate voltage(3.0V). This is because I discharge the battery with electrical load which the UVP was set to 6.0V.
So my final voltage is slightly higher than 3.0V.
And I started to discharge it with UVP 5.8V, then I think the final voltage will become lower than 3.0V.

I confirm the quick-start guide, and I change the value, taper current, Quit current, Charge current and discharge current.
The latest my gg file is the following. Please confirm it if there are any mistake.
Flight5_SP_FS_Ver10_200727.gg.csv

Thanks
Bes Regards,

Yuichi Tsutsui

Dear Wyatt

Thank you for your confirmation.

I'll continue to do learning cycle and I'll get back to you once I finish the 1st discharge.

By the way, I would like to confirm the way to set the following value.
Now I'm following the way which another member did before, but I don't know the way is correct and this follows your manual.
So could you give me the material which mention about the way to set the following value?
 -CC Gain
 -CC Delta
 -CC Offset
 -Board Offset
 -Pack V Offset

Thanks
Best Regards,

Yuichi Tsutsui    

Dear Wyatt

Thank you for sharing the document.
I'll confirm it.

I've already finished 1st discharge, relax in more than 5 hours, 1st charge and relax in more than 2 hours.
The following material are gg.file and log.file at that time.
Could you confirm them there are any problem?
I think the update status didn't change so the learning cycle was fail.
Ver10.zip

Thanks
Best Regards,

Yuichi Tsutsui

Hello Yuichi,

I have been going through your logs but I can't find anything preventing the Qmax from updating yet. I will look through them more tomorrow and let you know if I find something.

Sincerely,

Wyatt Keller

Dear Wyatt

Thank you for reply.

From now I started to 1st discharge with setting OpConfig[IDSELEN] to 0.
I'll get back to you once the 1st discharge and relax is finished.

The attached files are the followings.
 -Dashboard at the starting of 1st discharge
 -gg. file with setting OpConfig[IDSELEN] to 0 
6014.Ver10.zip

Hello Yuichi,

Let me know if you are able to get an updated status.

Sincerely,

Wyatt Keller

0842.Ver10.zipDear Wyatt

I've finished 1st discharge, relax, 1st charge and relax.
The followings are log.file and dashboard at the each steps.
Is there any issues?
All Update status seem not to set...

And I'm taking mistakes after 1st charge and relax.
So I restart learning cycle from 1st discharge...

Thanks
Bes Regards,
Yuichi Tsutsui

Dear Wyatt

I'm starting the 1st charge after discharge and relax.
The attached files are the latest log file and gg.file and dashboard at the starting to charge and after discharge and relax.

I have a question about the relationship between sending RESET command and FAS, SS, and RUP_DIS bits.
Before starting the discharge, gg.file is acquired and the RESET command is sent.
When I send the RESET command, the FAS, SS, and RUP_DIS bits become High which makes it impossible to access gg.file.
After discharge is completed, when UNSEAL FULL ACCESS command is sent for getting gg.file, FAS, SS and RUP_DIS bits become Low.
When I send the RESET command after getting the gg.file for next charging, the gauge displayed on the left edge of the dashboard is changed from 0% to 1%.
Is it correct to changing gauge?

Thanks
Best Regards,
Yuichi Tsutsui

0257.Ver10.zip

Hello Yuichi,

I've been reviewing your files with a colleague, I'm hoping to have some helpful information soon.

Looking at your logs you're hitting 0% SOC much before your current is stopped, this may be because the chem ID is not properly matched. 

Before the initial discharge you use the RESET to stop resistance updates, they should clear once you rest the batteries. You shouldn't need to send the RESET command after the one sent before the discharge.

Sincerely,

Wyatt Keller

Dear Wyatt

I appreciate that you confirmed this with your colleague.
Thank you very much.

From your reply I seemed to misunderstand the steps of learning cycle.
Is the following procedures correct for learning cycle?

1.Sending reset command
2.Discharge the battery to less than terminate voltage with C/5-C/10
3.Relax in more than 5 hours
4.Charge the battery (not sending reset command)
5.Relax in more than 2 hours
6.Discharge the battery to less than terminate voltage with C/5-C/10 (not sending reset command)
7.Relax in more than 5 hours <- completed learning cycle!

I have sent rest command at step1 and step4. I think this is the cause for failing learning cycle based on your reply.
But I have some concern about sending this command and how to get gg.file.
After sending rest command, FAS and SS bits become high. I know under this conditions I can't get gg.file.
So I need to send UNSEAL_FULL_ACCESS command to get it during the learning cycle.
If I get gg.file during the learning cycle, is there anything specially I need to do before restarting the learning cycle?
For example sending RESET command or IT_ENABLE command, etc.

So I don't get gg.file and send any command except sending RESET command before discharge during the now ongoing learning cycle.
Now I've completed step1-step3, and step4 now is going.
The attached files are dashboard captures at each steps.
If you have any questions, please let me know.

Thanks
Best Regards,
Yuichi Tsutsui  

Ver10_4th.zip

Dear Wyatt

I've completed the learning cycle as the prior steps.
The attached materials are the dashboard at the each steps and gg.file after learning cycle.
Ra and update status look not changed. So I think the learning cycle was fail.
Could you confirm them and please give me any advice.

Thanks
Best Regards,
Yuichi Tsutsui

1425.Ver10_4th.zip

Hello Yuichi,

Are you sure you have programmed the correct parameters for your battery? Even after SOC reaches 0% it continues to discharge for some time after.

Your steps above are correct, only reset at the beginning like outlined in SLUA903. Send the GAUGE_EN before the RESET command as well as outlined in the app note.

Can you also export the .gg file with the logs so we can double check your setting for each cycle?

Sincerely,

Wyatt Keller

Dear Wyatt

I think the parameters are correct because these are programed after sending the data to gpcchem.

As the email from Kang-san, I try to chose the Chemical ID again.

The attached file is the .gg file. Is it correct which you requested?
8228.Ver10_4th.zip

Thanks
Best Regards,
Yuichi Tsutsui

Dear Kang-san

Thank you for your reply.

The IVT file is the attached file. Is it correct? 

This is the one I get from bqstudio during the learning cycle.

learning cycle_200804.log

Thanks
Best Regards,
Yuichi Tsutsui

Dear Kang-san

Thank you for your reply.

The IVT file is the attached file. Is it correct? 

This is the one I get from bqstudio during the learning cycle.

7024.learning cycle_200804.log

Thanks
Best Regards,
Yuichi Tsutsui

Dear Wyatt

I resubmitted the data to GPC tool.
And I get the new chemical ID.
This is because when I get the prior chemical ID from GPC tool, I took mistake as your reply (Jul 6, 2020 7:06PM).

So from today I started the learning cycle with the new gg.file.
The attached materials are the new gg.file and the submitting data and the report from GPC tool.
I'll get back to you once the learning cycle is completed.

If you have any concerns about the attached materials, please let me know.

 GPC data and ggfile.zip

Thanks
Best Regards,
Yuichi Tsutsui