Hi There,
I have been using and investingating the bq34z100-g1 for a little while now with a bench test setup (EVM + EV2400) with a SLA replacement LiFePO4 battery (UltraLife URB1270 - https://www.ultralifecorporation.com/Go/5-148-URB1270+Technical+Datasheet.aspx). It has an internal cell configuration of 4S5P, 7.5Ah capacity, 14.4V charge voltage and 10V cut-off. I'm not totally sure of the internal cells, with that configuration they must be 1.5Ah cells. Hoping someone can answer some of my questions.
I couldn’t find the right cell when I started, and to get going I used a chemID for a K2 18650P LiFePO4 cell (ChemID: 0401) and this appeared to work ok and I performed a successful learning cycle after a few failed attempts – this seems to be the hardest bit!
I’ve performed a fair bunch of testing for our end application and read lots of documentation (such as SLUA664): however a lot is not clear to me about specific uses of parameters in the bq34z100-g1, but it is the most thorough and informative documentation I could find.
1) I wanted to check if there was a ChemID more suitable for this cell already than what I used, or if I need to send some batteries to TI for profiling.
2) Because the batteries are sealed, we cannot sense the temperature of the cells, in fact we only sense the ambient temperature that is in the same compartment as the batteries, how much of a problem in terms of SOC error should we expect from this? I have observed up to around 20C self-heating on the batteries (mainly occurs during charging), that is not observed by the bq34z100-g1 device.
3) In the application, the user/service can replace the battery with a new one after a few years if required. The bq34z100-g1 is not replaced with the battery. How long will it take the bq34z100-g1 to ‘adjust’ to the new battery before it is accurate again? Will the battery profile be completely different from a new battery?
I have performed some limited bench testing but I don’t think that is representative of a transition from aged cells to new cells.
4) I have tried unsuccessfully to test/simulate the SOH measurement feature. When/how often does the SOH get computed? What use does the ‘SOH load current’ parameter have?
5) Does a new learning cycle need to be performed if I want to test different ‘Load mode’ or ‘Load select’ settings? I would like to see the effects different usages of these parameters have on my application, especially constant power mode, but haven't observed any differences yet.
6) When is factory programming going to be supported in the multi program/calibration setup from TI, or should we build our own program/cal setup for use on the production line?
7) The parameter ‘TCA set %’ does not appear have much documentation, is this important? SLUA664 refers to a termination alarm, but I haven't seen this elsewhere.
8) The use of the parameters 'max/min sim rate' is not very clear, should these just be configured to be the minimum and maximum expected loads? Is there a tradeoff with setting them too large?
9) SLUA664 mentions that parameter 'Ra Max Delta' should be updated to '15% of the grid point 4 Ra value after an optimization cycle has been completed.' It then says under normal setting: 'must be calculated and changed for the golden DFI file to be 20% of the Ra 4 grid point'
In this case does optimization cycle mean learning cycle? Can you please confirm if it should be 15% or 20%?
In bqStudio, Ra Max Delta is shown as a percentage (15% in my example - which i haven't changed from defaults). How do you set this to a non-percentage value as per SLUA664?
10) Hopefully you can provide some comments on the attached learning cycle log and the settings file.
(Please visit the site to view this file)
Thanks!