1

I have a LiPo battery connected to my device and to the MAX17048 chip. Battery has a cut-off voltage 2.75V. Unfortunately, the IC does not detect the SOC level properly, see the example read log of the chip below

Sat Sep 24 23:40:01 CEST 2022
Battery Voltage: 3.61V
Battery percentage: 2%
SOC rate: 0%/h
-----
Sat Sep 24 23:50:01 CEST 2022
Battery Voltage: 3.61V
Battery percentage: 2%
SOC rate: -1%/h
-----
Sun Sep 25 00:00:01 CEST 2022
Battery Voltage: 3.61V
Battery percentage: 2%
SOC rate: -1%/h
-----
Sun Sep 25 00:10:01 CEST 2022
Battery Voltage: 3.6V
Battery percentage: 2%
SOC rate: 0%/h
-----
Sun Sep 25 08:10:01 CEST 2022
Battery Voltage: 3.4V
Battery percentage: 0%
SOC rate: 0%/h
-----
Sun Sep 25 08:50:01 CEST 2022
Battery Voltage: 3.37V
Battery percentage: 0%
SOC rate: 0%/h

My device has a battery life of about 24 hours up to 3V, and the system from 3.6V shows discharged battery. My battery model LP805080 4000mAh LiPo.

What can be the cause of wrong SOC calucating?

dmvs
  • 51
  • 1
  • 3
  • Which battery? If you have a datasheet, please edit and link it, thanks. – rdtsc Oct 04 '22 at 20:03
  • This charge monitor uses battery voltage (and history) alone to determine SoC. It can get confused if checked too soon, or externally powered before connecting a battery, or many other things. Can you power cycle the board (disconnect all sources, reconnect battery)? How about charge and discharge the battery while the monitor is watching? – Tim Williams Oct 04 '22 at 21:10
  • I am observing very similar behavior with my PCB. Have you verified with a multimeter that the voltage is actually correct? Mine is off by 0.5V which is quite a lot. Using the same reference circuit provided by Maxim. I reached out to them as it seems a common occurrence. Also what's the reported chipID/version? – Paperino Nov 05 '22 at 05:57
  • Also, what size is the capacitor between VCELL and GND you picked? We picked what's in the specs but we are getting wrong readouts as well... – Paperino Nov 05 '22 at 20:44

1 Answers1

0

I am getting convinced that, like me, you hit a batch of defective chips. There's still stuff that I can't understand very well and I destroyed my defective sample, but I am pretty sure that the voltage drop across VCELL and GND was not the same observed at the battery terminals. But it was the same on a working chip.

Paperino
  • 205
  • 1
  • 7
  • I couldn't find a solution so I estimate the correct battery SOC from the polynomial equation. – dmvs Nov 08 '22 at 14:49