BQ76930: What is bq76930 current monitor accuracy?

Hi Jimmy,

The simple answer is yes.  Ignoring the INL and offset specifications and the setup value of the CC, 30 mA/2000 mA is 1.5%, so it is at the maximum gain error specification if the CC is perfectly linear, has no offset and is at nominal temperature. It does seem unusual that what would seem to the reader to be a single measurement would be at the specification limit.

The more complicated answer is probably since there are several effects which can influence the resulting CC (ADC) measurement.

The CC_CFG register address 0x0B must be set to the 0x19 value for best performance.  Without this setting errors are often observed at low values.

The gain has a temperature coefficient which can increase the deviation.

The CC is specified with best fit INL having a gain and offset.  This keeps the INL number smaller than using an end point INL specification.  The deviation is in LSB, so the amount of error changes across the range.

The part is on the board measuring current so the sense resistor and board resistances will have an effect.  The sense resistor will have some some temperature coefficient and will self heat, the effect becomes more dramatic at higher current & power. Unless the design is using a 4 terminal sense resistor which is unusual, there will be effects from the sense line connection and the solder.  Calibration of the system can also have an effect, calibration of the offset and gain at a point should make the error small at that point and should be very good based on the measurement instrument.  Moving away from the calibration point the error may increase based on the INL of the converter.  Generally the converter is expected to be very linear near the middle of the range (near 0) and have curvature near the end, so calibration low in the converter range is expected to be good for most of the range as long as the resolution is suitable.