BQ20z655 can't perform optimization cycle

Could anyone can help? Or could someone send me default GG file or SENC file for BQ20z655-R1?

I would be very grateful.

Karol
Please provide an email address and I will send to default senc file.
Tom

Tom,

My e-mail adress is kplocha@wamtechnik.pl

Thanks in advance

Hello Tom,

I flashed SENC that you send me. After this I made following steps:

- set design capacity and all QMAX

- set term voltage and taper current

- perform calibration

- apply proper chem id for cells

Then I type 0021 and start charging. After charging I wait 5-6 hours, then I start discharging process and also I wait 5-6 hours for rest. Unfortunately, the update status is still 04. The VOK and QEN are set. I have already made 6-7 cycles and still nothing.

I must say that I am out of ideas now. Could you provide any support? The case is urgent.

I attach:

- all designing files

- GG file

- some logs

8272.1163.20.01.3A.zip

1163 logs.zip

GG_file_1163.zip

hI Karol
You gg file has all zeros for the parameters. Pls ensure your device isnt sealed and then extract a new gg file so we have the correct parameters.
Secondly your log file doesn't show a continuos profile of the discharge-rest-charge-rest-discharge-rest-charge-rest-discharge profile. You need two full charge and disharge cycles with rest periods in between to archieve successful learning.

Also ensure your discharge current is between c/5 and c/10 and ensure it is constant currect. Furthermore, ensure FC flag gets set during charging indicating the battery is fully charged. pls review the primary charge termination section of the technical reference manual.

I hope this helps
thanks
Onyx

Hi Onyx,

Thank you for answer. This is really strange that GG file is filled with 0. I will check this and then I will attach correct GG.

I know that logs contains only end of charge / end of discharge data, but I think that this end data is most important. But I also will make full cycle with all data logged and then I attach this.

Please know that I have already performed many learnings for other TI ICs like: BQ34z100-R1, BQ20z95, BQ40z50-R1, BQ40z60 and I never have any problems with this. With BQ20z655-R1 I already made 6-7 cycles with proper current, proper charge termination (FC was set), but still no success.

When I collect data from cycles, I will write again.

Best,

Karol 

Hi Karol,
Good to know you have already performed learning cycle in the past on other devices. It helps to know you aren't a newbie :) Now lets go over certain things that could have resulted in failure to learn

-Make sure to have atleast 5 hours rest periods after discharge and 2 after charge.
-Set your design capacity to what is stated in the manufacturer data sheet.
-Charge to the nominal charging voltage stated in the battery data sheet and discharge to the EDV stated as well.
- Ensure the chem id you are programming are the correct chem id for your cells. Use our gpcchem tool to identify your chem id if you haven't already done so.

If you do all this and learning still fails, send the complete log file of the learning cycle and and senc and gg file.

thanks
Onyx

Hi Onyx,

I'm Karol college from work and now I'm trying to solve this problem.

We tried those first 3 steps and still no progress.

Then I have used a tool GPCCHEM and it shows a slightly ChemID than there we were using. Our cells are LG18650HG2 and it's got a ChemID 2112, but the tool showed that the ID is 2125 or 3758 with less than 1% error.  I've uploaded the new Chem with ID 2125 and it is also LG18650HG2 but with option -10C. I'm cycling ot now with changed ID and I will see what will happen and maybe it will be working.

I'm also sending You a file for gpcchem and report from this tool beside the gg file. Log is in the file for gpcchem.

1072.Files.zip
Thanks
Michal

I checked ChemID's 2112 and 2125 and the OCV tables are almost the same, so either option should work for you. I would use 2125, because the GPCCHEM tool identified it as the best choice. The -10C means that it was a special characterization and it was evaluated down to -10 degC. The other one was characterized down to 0 degC.

Hello All,

I still have problem with BQ20z655 during performing learning cycle. I have tried all tips and advice that you wrote, but it won't help. I even tried to perform learning cycle with NCR18650B, but still no success.

I send you complete log with full charge -> rest -> discharge -> rest, GG file and SENC file.

The case is urgent and I ask for solution/explanation. Could it be that BQ20z655 has some bug?

Thanks in advance,

Karol

1401.CompleteLogAndGGAndSENC_1163.zip

Karol

I checked your data and there are several discontinuities in it, so I am not sure whether they caused the cycle to fail. the Mathcad program does show that ChemID 2125 should be a good match for the cells, so I used it to calculate the Ra table and Qmax. I updated these in the attached gg file.

Program the default senc file

load the ChemID

Load the gg file

Reset the device

Calibrate the device

Issue the IT Enable.

Run a charge and discharge cycle and test gauging accuracy.

Tom

https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/196/GG_5F00_file_5F00_1163-optimized.gg

Many Thanks Tom!

I also wonder about this discontinuities and I hope that was only communication problem. The strange thing is that there was so many current spikes and I don't know where they come from.

Anyway, I will flash software, then I will make tests and in the next week, I will provide you feedback.

Best,

Karol

Hello Tom,

We checked the GG file that you send us, and now all looks OK and BMS behave well.

Thank you for your support. We really appreciate your help.

Hello All,

I have another question about BQ20z655. During our tests we have strange current readings. About each 9 minutes the current value that is showed in BQ Evaluation Software equals about -64mA. The output pads P+ and P- are floating and they aren't connected to any charger or load so that current  is really strange.

Could you tell if this issue is design mistake from our side or there are something weird with the BQ20z655 IC.

I attach you GG file and log from charging and discharging.  I also attach you our design files.

1163_full_log.zip

1163_Design_files.zip