A little update - I tried it out on the bench supply and grabbed some numbers for the initial current at first turn-on, and after it had warmed up for a few minutes and settled into a steady state:
| Volts | 7.60 | 8.00 | 8.40 |
|
Initial Amps (Watts) |
3.69 (28.0) |
4.36 (34.9) |
4.96 (41.7) |
| Settled Amps (Watts) |
1.81 (13.8) |
2.15 (17.2) |
2.39 (20.1) |
(Keeping in mind those numbers are taken from the supply, so the actual power through the LED will be lower)
Yes, it is. It has three major characteristics to recognize a real XHP70.2:
What’s the pencil trick?
Make a dark line between each end of a resistor or capacitor. The graphite is conductive. For a capacitor, it is useful as a substitute for a bleeder resistor to discharge the capacitor more quickly - whether to affect how the light’s memory behaves or enable a lighted switch or something. For a resistor, it will decrease the resistance, so on a sense resistor you can trick the light to give more current.
Thanks, handy to know.
Unfortunately I don’t think this even has current sensing - I can’t see where it would be in the circuit, and the brightness varies a fair bit with input voltage changes.
I’m not sure reducing the resistance across R8-R11 from 50mΩ even to 25mΩ is going to do much in terms of allowing more current to pass either.
There’s a voltage divider made up of RT and R3 that I could probably tweak to reduce the temperature throttling, and that might be the most straightforward way to boost output. Increasing R5 might also boost output, but it’s harder to test that with a pencil 
Yeah, those are just a pile of parallelled ballast resistors to limit current when DDing.
That 9926 is a dual-FET that just crowbars the LED across the mains but through those resistors, PWMed by that 3-lead “transistor” (µC, actually).
Same crappy design like on the hateful AT-01 and similar lights’ cheepcheepcheep little cheeken drivers.
Good info!! thanks all!!
hahaha on the rubber pencil thing.