I did some detailed diagnosis on the fried driver from my Explorer E84 (previously reviewed here ) and determined that it was the voltage boost circuit that had failed, just like both my UniqueFire G10 lights. Connecting the battery to the output of the boost circuit restored complete functionality.
Taking a look at the battery side of the driver:
You can see two chips. The one at 11 o’clock, next to the “E” is the power transistor switching the LED current for PWM. The one at lower left is, as far as I can figure out, an ACT6305 “Low Input High Efficiency Syncronous Step-Up DC-DC Converter” configured for 5V output. “Vin” is its input, connected to one side of the inductor, and Vcc is its output, driving both the microcontroller and the LED. The other side of the board has the inductor, the microcontroller (SOIC8), and three magnetic sensors.
Connect Bat+ to LED+ for a single-mode direct-drive flashlight. Connect Bat+ to Vcc to bypass the failed boost circuit. Or, as I did after a couple of unsuccessful attempts to solder a wire to the battery terminal, bridge Vin and Vcc on the ACT6305 with a tiny bit of wire (all current will flow through the inductor, but it does so with the boost circuit operational anyway).
Update:
Now that it’s dark, I’ve been able to give the fixed light a proper workout. It looks like bridging Vin to Vcc is not a good choice. I’m seeing flickering coming and going over a period of a few seconds in low and medium modes. I figure the varying current draw from PWM is setting up oscillation in the inductor.