I am using a MCP73844 dual cell Li-Ion charge management controller to charge a pair of LIR2032 cells, which are rated at 45mAh.
I have breadboarded the circuit according to the schematic in the datasheet, however I am using a sense resistor of 2.8 Ohms to achieve a constant current of ~40mA so I do not exceed the 1C rating of the cells. The only other difference is that I am using an NTD2955 as the driver MOSFET.
I have connected a 10v input to the charging circuit, and a depleted battery, and I am getting unexpected behavior. When powered up, I experience the following.
1 – The circuit draws ~2mA for a couple of seconds. As expected in the charge qualification phase.
2 – The status LED illuminates and the circuit draws ~40mA, (sometimes, and only for a few of seconds)
3 – This quickly decreases to ~20mA, and then continues to slowly decrease down to ~12mA (Presumably as a result of having transitioned into the constant voltage charging phase)
4 – After some time has passed, the current draw drops to 0mA, and the status LED turns off, indicating the charging has finished, but after a few seconds, the current jumps back up to ~12mA, and the LED comes back on. This cycle seems to be repeating indefinitely.
5 – The battery voltage ends up being only 7v
So, the questions that I have are:
1 – Why is the constant current phase not always happening, and when it does, why does it only last for a few seconds?
2 – Why does the charge cycle keep restarting?
3 – Why does the battery not reach the expected 8.4v?
Cheers.
Best Answer
Best guess is that the misleading / verging on wrong example circuit diagrams in the data sheet have lead you to reverse the FET drain & source connections.
The FET source should connect to Vin, not to battery, despite the FET-symbol on their diagrams suggesting otherwise. This is a P Channel FET (both their one and your replacement.) The symbol used incorrectly implies an N channel FET with drain to input.
For the FET to act as a FET the body diode must be as shown on their diagrams but to do this the drain must be to battery when a P channel FET is used. If d & s were reversed the FET could not be controlled as the d-s diode always conducts.
As Vin is 10V and Vbat max = 8.4V & less if depleted then charge I may lift Vbat due to IR drop & ...