Hi,
We used two bq EVMS to measure the charging current of a same battery and a same charging IC:
bq27200EVM reports average current of 367.71mA
bq27010EVM reports average current of 331.12mA
Experiments have been done repeatedly to verify the results. The design of the charger IC is to give a maximum fast charging current of 325mA, and is set by a 1% ISET resistor. fbq27010 result seems to be closer to the real value.
From datasheet, “Average Current Registers (AIL/AIH) — Address 0x14/0x15” is read-only value. They would be affected by discharge compensation and temperature rate in EEPROM. So do the different compensation algorithms resulted in the 10% accuracy difference?
Noah
Try using constant load discharge currents instead and measure with a meter to confirm. Also the bq27xxx EVMs are not calibrated for current. The EVSW makes values display in mA based on a Rs value that is entered on the top. This value needs to be calibrated to ensure that current is reported accurately.
Mike,
Thanks for the advice. We will report the results with this method after the experiment.
BTW, does the "fully differential Delta-Sigma Coulomb Counter (DSCC)" do continuous integration or discrete sampling or the current? I worry that if discrete sampling is performed the result might be inaccurate, because in case of mobile phones transmission there is 2A high and brief pulses which discrete sampling tend to lose lots of them.
The DSCC does continuous integration. The average current is actually calculated over time based on integration.
So even for 1A or above brief high current pulses during wireless transmission the DSCC result is still accurate, regardless whether the sampling theorem condition holds?
The DSCC can capture activity as dynamic as the GSM profiles which are roughly 210Hz, 12% duty cycle with pulses between 1-2.5A peak.