Actions
Forum Thread Details
  • State Not Answered
  • Replies 2 replies
  • Subscribers 34 subscribers
  • Views 238 views
  • Users 0 members are here
Related

MAX17263 SOC Changing on Power Cycle

nolannirsense

Hello,

I've been working with the MAX17263 battery gauge recently via the MAX17263 EVKIT, and I'm connecting to it and reading the SPI outputs via an NRF5340 dev kit. I also have the MAX17263 EVKIT connected to a battery and a charger. I have the device set to print out the state of charge by reading the RepSOC register and dividing it by 256. My device polls the SOC every two seconds. When I first turn the device on, it gives me an initial SOC reading, let's say 55%. After a few minutes of charging, it'll go up to 57%. If I then power cycle the device, the percentage I will see is often a few percent higher after the first power cycle, so maybe 61%. Then, if I power cycle it again, the SOC returns to 57%, or sometimes even lower than the initial value, such as 56%.

I notice this inaccuracy trend when reading the RepCap register as well. My battery has a total capacity of 250 mAh, so the DesignCap register is set to 500. After charging for a while, the capacity will be around 324 before power cycling, 337 after power cycling once, and 323 after power cycling a second time. This pattern is pretty consistent, and I've checked a lot of settings and different registers, but I'm not sure how the RepCap and RepSOC registers are calculated. What could be causing these variations? My device needs to be accurate within one percent, and I was using the RepCap to give a more detailed view of the battery state.

What controls the RepCap and RepSOC registers within the MAX17263, and what could be causing this inaccuracy?

Thanks!

Parents
  • 0

    This part uses both voltage measurement and coulomb counting to try and figure out arbitrary battery status. They claim about 2% -3% accuracy.

    During charging and discharging they are mostly using coulomb counting, but presumably on power cycling voltage measurement is needed to get an initial baseline.

    Have you precisely measured the battery during power cycles to see if its voltage shows a corresponding behavior to your readings?

Reply
  • 0

    This part uses both voltage measurement and coulomb counting to try and figure out arbitrary battery status. They claim about 2% -3% accuracy.

    During charging and discharging they are mostly using coulomb counting, but presumably on power cycling voltage measurement is needed to get an initial baseline.

    Have you precisely measured the battery during power cycles to see if its voltage shows a corresponding behavior to your readings?

Children
  • 0 in reply to dougw

    I have not tried precisely measuring the battery via a voltmeter, but that's a good idea. Do you know if there's a way to measure the voltage that the MAX17263 is seeing so that I can print that alongside the RepCap and RepSOC reports?

    Also, thank you for telling me about the 2% to -3% accuracy margins of error! Where did you see that detail? I was looking around in the datasheets and couldn't find anything about that. Is there another place I can look to find relevant information like that? That might save me some future questions as well.

    Thanks so much!