BQ40Z50: bq40z50R5_tooltips CycleCnt jumps to 65535

JeremyG

Part Number: BQ40Z50
Other Parts Discussed in Thread: BQSTUDIO, , BQ78350-R1A, BQ78350

Tool/software:

Hi Support,

Following on from our Golden File program, calibrate and functional test routine, we run a few condition cycles to ensure packs perform as expected.

The packs are cycled using a programmable power supply, electronic load and control PC that also logs data using bqstudio.

We are finding that during the condition cycling, the CycleCnt value on some packs unexpectedly jumps to 65535, as captured on line 1372 of the attached log file.

This appears to be an issue with related threads elsewhere in this forum, but it is unclear what the cause or resolution is.

Has there been further progress on resolving, or narrowing down the cause of this?

Although I only have a small dataset thus far, on first observation, it appears more likely to occur for batteries with higher error counts in the *.err log file.

Thank you,

Jeremy.

D00D_Cycle_1.log

6 months ago

0 Adrian 6 months ago

TI__Mastermind 37751 points

Hello Jeremy,

This is an issue within the firmware. I would highly recommend updating the FW on the gauge.

https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/196/4500_5F00_5_5F00_05_2D00_bq40z50R5.bqz

https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/196/bq40z50R5_5F00_v5_5F00_05_5F00_build_5F00_96.srec

Regards,

Adrian

0 JeremyG 6 months ago in reply to Adrian

Intellectual 631 points

Hi Adrian,

I haven't seen it on all of the packs we are testing.

Has the cause been identified and is there anything that can be done to reduce the likelihood of it occurring?

Will we see this issue listed in the Release Notes?

For reference, the full listing of FW versions for the bq40z50 is here;

https://www.ti.com/tool/download/BQ40Z50-DEVICE-FW

Regards,

Jeremy.

0 Adrian 6 months ago in reply to JeremyG

TI__Mastermind 37751 points

Hello Jeremy,

I believe the release notes should be on the download from that website. I pulled the files from that download and sent it directly to you. As far as I know from firmware team is that the newest version of firmware has a much less possibility of causing the cycle count of jumping to 65535.

Regards,

Adrian

0 JeremyG 6 months ago in reply to Adrian

Intellectual 631 points

Hi Adrian,

Thank you for clarifying - could you confirm with the firmware team please? The most recent release notes appear to be January 2023.

Thank you,

Jeremy.

0 Adrian 6 months ago in reply to JeremyG

TI__Mastermind 37751 points

Hello Jeremy,

Yes, firmware team confirmed that the build I sent has a less likelihood of this issue occurring.

Regards,

Adrian

0 JeremyG 6 months ago in reply to Adrian

Intellectual 631 points

Hi Adrian,

The website https://www.ti.com/tool/download/BQ40Z50-DEVICE-FW only provides Release Notes for major Revisions (R1, R2, R3, R4, R5), but the downloadable Firmware Bundles provide greater detail, with what appears to be a full and cumulative listing of all builds released.

As there is no mention of the CycleCnt issue in the Release Notes for the latest Firmware Bundle v5.05 (build 96), it is unclear to me how the likelihood of the issue occurring is reduced, unless it somehow relates to some other aspect.

Given the significant disruption involved in recalling and upgrading batteries, I'd like to be confident that it will address the issue, so any further advice or insight you can provide would be greatly appreciated.

Regards,

Jeremy.

0 Adrian 6 months ago in reply to JeremyG

TI__Mastermind 37751 points

Hello Jeremy,

I have confirmed with the firmware team that newest version of our firmware has a less likelihood of the issue occurring.

Regards,

Adrian

0 JeremyG 6 months ago in reply to Adrian

Intellectual 631 points

Hi Adrian,

It appears that other customers have been / are being impacted by similar issues and may also benefit from this firmware upgrade, including one example where the CycleCnt does not increment.

https://e2e.ti.com/support/tools/simulation-hardware-system-design-tools-group/sim-hw-system-design/f/simulation-hardware-system-design-tools-forum/891868/webench-tools-bq40z50-bq40z50-cycle-count-65535-changed?tisearch=e2e-sitesearch&keymatch=65535#

https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/1447277/bq40z50-r1-cycle-count?tisearch=e2e-sitesearch&keymatch=65535#

https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/1469188/bq40z50-r2-bq40z50-r2-cycle-count-not-updating-in-1s1p-design/5647269?tisearch=e2e-sitesearch&keymatch=cycle%20count#5647269

https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/1353995/bq40z80-the-number-of-cycles-increases-to-65535

In the fourth request listed above, the firmware team reviewed the config for the following registers as part of the troubleshooting effort.

- Coulomb Counter Deadband & Current Deadband

- Power Config

- IT Gauging Config

- Design Capacity

I have checked my configuration and confirm that I am using the default values for these registers, with the exception of Design Capacity and the CCT function within the IT Gauging Config Register (SLUUCN4B, p.70).

I have chosen to use FullChargeCapacity rather than DesignCapacity for cycle count threshold calculation - was this part of the issue?

Also, does the issue impact other parts, such as the CEDV-based BQ78350-R1A?

https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/1451347/bq78350-r1a-number-of-cycles-count-increased-from-3-to-65535

Should the same fix (applied between bq40z50R5 v5.03 and bq40z50R5 v5.05); also be applied to other parts, including the BQ78350?

Regards,

Jeremy.

0 JeremyG 6 months ago in reply to JeremyG

Intellectual 631 points

Hi Adrian,

I have been doing some further testing and can add the following observations.

It appears that the issue occurs early in the life of the pack, during our condition cycle process (Update Status 0x06 ->0x0E).

When the issue occurs, Cycle Count and Qmax Cycle Count both jump to to 65535.

Only some packs are impacted and it appears that if these packs are reprogrammed and complete the condition cycle without the issue presenting, it won't occur later.

For impacted packs, if the Cycle Count and Qmax Cycle Count registers are manually adjusted using the bqStudio Data Memory tool; after a minute or so with the Scan tool running (default 4 second refresh rate); a read error such as "No acknowledge" or "Packet Checksum mismatch" will occur, then the Condition Flag (CF) within Gauging Status will flash before Cycle Count jumps back to 65535. Note that Qmax Cycle Count remains the same and does not jump to 65535.

I can't seem to reliably reproduce the issue as the delay between adjusting DF and the "jump" varies - it may also vary with initial values, but I'm not certain about this.

I hope this helps!

Regards,

Jeremy.

Power management

Power management forum

BQ40Z50: bq40z50R5_tooltips CycleCnt jumps to 65535