• TI Thinks Resolved

BQ40Z50-R2: RSoC drop at end / change in FCC value between charge and discharge state

Prodigy 160 points

Replies: 6

Views: 84

Part Number: BQ40Z50-R2

Hello,

I use the BQ40Z50 in an application where the charge current is quite high (800mA) and the discharge current is low (10-30mA usually, with some short peaks in 100mA range).

Even after many golden images I have some drops in capacity at the end of discharge, as example the one below that drop from 10% to 0 :

I can't share publicly all the parameters of my application, but here is what I set :

3300mVpc should be 0% of RSoC, I set the options to be at 0% as soon as one cell reach this value. I do this to be sure to report 0% before reaching CUV at 3000mVpc
The problem is we are not going down smoothly....

One thing I have seen is  the FCC change quite a bit between Full charged state at 757mAh to 691mAh at the end (this was reported at 13%, but I don't have a frequent report of it)
So if we have 66mAh change between charged and discharged (almost 10%) I think this is not great to have a precise gauging.

I already use the 3mR sense resistor, which is the max recommended. Can this be problematic at low current ?


I share the IT configuration part, if you need another information let me know. If you need the whole file, please tell me how to send it privately.

Thanks in advance for help !

  • Hello Emilien,

    I've assigned this thread to an applications engineer who will follow up with you via DM.

  • In reply to Kang Kang:

    Hello,

    Thank you. Hope I have some news soon, I haven't been contacted yet.

    Thanks for help !

  • In reply to Emilien Chappatte:

    Emilien

    Sorry for the delay getting back to you. When you say you created the golden file, do you have update status as 06 after learning completed successfully? Are you using the correct chem id identified using our online tool gpcchem? If you did all these, you might want to consider using user rate for your load select.

    thanks

    Onyx

  • In reply to Onyx Ahiakwo:

    Thanks for your reply Onyx.

    I made multiple chem ID identifications, sometimes with varying results, depending if I use C/10 discharge current or a bit less to fit more to our application.
    But results are repeatable when using same parameters. I don't know for sure if it's better to use C/10 or something more close to our application (in the range of C/30) to define the right chemistry ?

    About the use rate, I'm afraid that will not fit our need. If I understand correctly, I define the average discharge current and the BMON do prediction using this instead of coulomb counter.
    As our load can change a lot (like double compared to lowest current ) based on condition and current mode, I think this will not be very reliable ?

    Thanks and best regards

  • In reply to Emilien Chappatte:

    Hi Emilien,

    The using the user rate will allow the gauge to be conservative and will prevent SOC drops. From the load profile you describe, I would suggest a user rate of 50mA. Can you verify if learning was run successfully on this unit.

    Chem id identification using c/10 suffices to determine the right chemistry  for your cells.

    thanks

    Onyx

  • In reply to Onyx Ahiakwo:

    Hello Onyx,

    Thanks for continuing providing help.

    Learning was ok on this device.

    I'll try user rate with suggested value and I keep you posted in a few days.

    Thanks again !