Hello,
I am confused by the description of “6.4.2.2 Discharge Detection Threshold” on page 35 of bq27421-G1 Technical Reference (sluuac5) on its “independency” from FLAGS.DSG bit.
We know that bq27421 has integrated current sensor, so the basic understanding is that FLAGS.DSG bit is determined by whether the integrated resistor has positive or negative voltage across it, and I guess this bit should follow the current direction instantaneously without any hysteresis/deadband effect. Is it the case?
Secondly, the default “Discharge Detection Threshold” is 167mA. 167mA is a significant value and is at least dozens times larger than parasitic/leakage current due to any poor PCB design, then I wonder why the threshold is set such high. Does the device actually use the instantaneous voltage to determine charge/discharge state, or the discharge current must be greater than 167mA to be a qualified “discharge”, and the charging current must >100mA to be qualified as “charging”? Actual device might have “sleeping” modes in which under 50mA load current is not uncommon, then are they going to be ignored when “discharge detection threshold” is 167mA, thus render gauging inaccurate? Conversely, when charging current tapers to below 100mA, are they completely ignored by the device and result in corresponding inaccuracy?
A grave implication following the above confusing interpretation is that the gauge seems to be incompatible with any device having multiple states‼ unless the “discharge detection threshold” is set to the minimum of currents of all states. However, it yet has to be above board noise/drift. As for the “charge detection threshold”, there seems no cure for the inaccuracy with tapered low current.
Hui