BQ40Z50-R1: Learning cycle failure - RDIS, Negative Scale Factor Mode bits set during discharge

We are working through the process of determining ChemID and subsequently running a learning cycle. During the learning cycle discharge phase, RDIS activates along with the NSFM bit of GaugingStatus, and thus we are unable to successfully complete RA table updates. I am looking for assistance in understanding

• Why RDIS is set (presumably due to NSFM - but why does this occur)
• What to change to get a 'clean' learning cycle

I have read in other responses that this may be due to incorrect ChemID. We are using ChemID 1450, determined using the GPCCHEM online tool. The log file submitted to GPCCHEM was:chemID_cycle.zip

And the report received: chemID_cycle_report.zip

Which specifies a best-fit ChemID of 1450, with max DOD error of 0.97% - well within the 3% threshold I've seen referred to.

We have attempted the learning cycle twice; first attempt with a discharge rate of C/5, second attempt using C/7 (as seen that referred to in some answers as a good value to use). Both logs attached, along with *.gg exported when the log was saved: learning_cycle_log_gg_c5_rate.ziplearning_cycle_log_gg_c7_rate.zip

The second attempt (C/7 rate) was curtailed after RDIS bit observed to be asserted. The log and *.gg file shows that the RA table was at least partially updated (Ra flags set to 0x5555). However, update status remains at 0x05, would expect 0x06 after a successful learning cycle.

Any suggestions as to what is going wrong?

Thanks,

Hamish