BQ40Z50-R2: Balancing issue

VASANTH MITHILACODY62

Part Number: BQ40Z50-R2
Other Parts Discussed in Thread: BQ78PL116, , BQ40Z50, BQSTUDIO, BQ4050

Hi,

We are seeing the issue of 4S8P 19.2Ah LiFePO4 battery pack is not charging to 100%.

The pack was able to charge to 100% in the beginning of production but after 4 months of usage, the pack is able to charge up to 80% only. When we read the pack through battery management studio, the cell 2 has shown up zero balancing time where as cell 1, 3 and 4 balancing time are higher ( refer to attached image ).

During charging, Cell 2 reaches 3.6V when other cells are in the range of 3.5V. During discharging cell 1 drops to <2.8V.

When we disable the cell balancing ( CB, CBR, CBS=0), pack is able to charge to the full capacity.

what are the parameters affecting the cell balance time calculation ? Is there any provision to change the settings to voltage based balancing instead of capacity based balancing as in bq78PL116 ?

over 7 years ago

0 Swaminathan Ramanathan75031 over 7 years ago

TI__Expert 6575 points

The cell balancing in bq40z50-R2 is capacity based. It uses the DOD measurements and shunts capacity out of the cells with the highest capacity.
Looking at the CBDOD_x values, it appears that the CBTimers are being calculated correctly.
The bq40z50 does not have an option to balance based on voltage.
If you can post a bqStudio log of the charging cycle as well as the gg file and the chemID you are using, we can investigate further to see what is going on.
Also are you using internal cell balancing or external cell balancing?
Internal cell balancing can only bypass 10mA, so for such a large capacity pack, it may not be able to keep up.
Regards,
Swami

0 VASANTH MITHILACODY62 over 7 years ago in reply to Swaminathan Ramanathan75031

Intellectual 475 points

GG file and charging log.rarI have attached the SREC file and charging cycle data. We are using external balancing and the balancing resistor is 33ohm.

0 Swaminathan Ramanathan75031 over 7 years ago in reply to VASANTH MITHILACODY62

TI__Expert 6575 points

The .rar file seems to be corrupted. I could not open it.
Could you please post a .zip file.
Regards,
Swami

0 VASANTH MITHILACODY62 over 7 years ago in reply to Swaminathan Ramanathan75031

Intellectual 475 points

Zip file attached

GG file and charging log.zip

0 Swaminathan Ramanathan75031 over 7 years ago in reply to VASANTH MITHILACODY62

TI__Expert 6575 points

Hi,
This is a large capacity pack, and from the log and gg it looks like the cell balancing calculations of the gauge itself is correct, but it might have a hard time balancing the cells:

The DODs here show that Cell1 has the lowest capacity, so the cell balancing time for that is 0, however Cell3 has the highest capacity and the time to balance this capacity down is ~56000 seconds or 15 hrs.

Also, looking at the gg file, the impedance values are in the single digits, so they are not likely very accurate.

I would recommend the following options:

1. Using current scaling. The application guide for using current scaling is at http://e2e.ti.com/cfs-file.ashx/__key/communityserver-discussions-components-files/180/6102.how-to-guage-high-capacity-packs.pdf

2. Changing to a bq4050 which is a CEDV gauge and uses voltage based balancing, and does not use the impedance values.

Regards,

Swami

0 VASANTH MITHILACODY62 over 7 years ago in reply to Swaminathan Ramanathan75031

Intellectual 475 points

Hi Swami,

Thanks for the reply and suggestions.
Few more questions.
1)As per datasheet, IC supports 29Ah. The pack what we are using is 19.2Ah. Any concerns here ?
2)What is the reason for single digit impedance values. Anything related to learning cycle ?
3)We are seeing that pack is able to hold only 80% of the capacity. The balance current is 100mA. With 75% duty cycle, average balance current is 75mA. So to balance the 20% of capacity difference, we require 3.84Ah/0.075A = 51.2 hours. We kept the pack for more than week in the idle mode, but still not able to get full capacity.
4)How current scale will help the balancing issue ?
5)Is bq4050 is pin to pin compatible with bq40Z50-R2. What are things we need to take care while migration to bq4050 ?

Regards,
Vasanth

0 Swaminathan Ramanathan75031 over 7 years ago in reply to VASANTH MITHILACODY62

TI__Expert 6575 points

Hi Vasanth,
1. The capacity by itself is not the concern, but the high capacity is causing the impedance to be low enough that it impacts the impedance track algorithm.
2. This is likely normal for these cells in a high parallel cell count configuration.
3. To debug this you might want to log during relaxation and see if the "CBTimex" is decrementing or something is inhibiting cell balancing. Also, check that you are above the "Min RSOC for Balancing" during the rest period. When the gauge takes its OCV readings for gauging, it will disable balancing for 30 minutes to ~5hrs so that it can get a clean rested voltage, so this will also have to be accounted for.
4. Current scaling will help with more resolution for the impedance. For example a /10 current scale will result in 10x impedance values. But, yes from a purely cell balancing perspective, it likely does not help much.
5. The bq4050 is a pin to pin compatible device. Only the gauging algorithm is different.

Regards,
Swami

0 VASANTH MITHILACODY62 over 7 years ago in reply to Swaminathan Ramanathan75031

Intellectual 475 points

Hi Swami,

Thanks again for the quick response. Appreciate your effort in sharing the detailed information.

1)Based on the above discussion, is there any possibility for us to continue with the bq40z50-R2 design by doing some changes in the balancing algorithm.

Since in the field, we cannot guarantee that there is enough resting time between each charge or discharge cycle. Sending reset command while charging will help us in recalculating the balance time and improve the balancing ? Suggest us if there is any method to solve this issue

2)We will replace bq40z50-R2 with bq4050 if we cannot achieve balancing.

Please confirm same external balancing circuit can be used for bq4050 too ( shared the PDF )0243.3022.bq40z50 external cell balancing.pdf

Regards,

Vasanth

0 Swaminathan Ramanathan75031 over 7 years ago in reply to VASANTH MITHILACODY62

TI__Expert 6575 points

Hi Vasanth,
1. I think it may be difficult with the bq40z50-R2. You could disable balancing al together to see if the performance is acceptable.
2. The external balancing circuit will be the same for bq4050. You can find more details on the working of the voltage based balancing at: www.ti.com/.../slva155.pdf
Regards,
Swami

0 joseph richards over 7 years ago in reply to Swaminathan Ramanathan75031

Prodigy 40 points

Is there support available inside the US for our engineers working this issue here?
Direct contact information to local Field Application Engineer is required ASAP as we are nearing production.

0 VASANTH MITHILACODY62 over 7 years ago in reply to Swaminathan Ramanathan75031

Intellectual 475 points

Hi Swami,

We have observed that at the end of charge and discharge, cell voltages are at different levels and when we send reset command during this time, looks like the bq is able to calculate the balance time appropriately (At the end of charge, cell with higher capacity shown higher balance time whereas the lowest cell shown zero balance time )

Do you see any issues when we send reset command at the end of each charge cycle ?

What are the paraments going to be impacted when we reset the bq IC ?

Ra table, calibration parameters will be impacted during reset ?

Regards,

Vasanth

0 VASANTH MITHILACODY62 over 7 years ago in reply to VASANTH MITHILACODY62

Intellectual 475 points

Hi Swami,

Any update on the above query ?

Regards,

Vasanth

Power management

Power management forum

BQ40Z50-R2: Balancing issue