I'm in the process of making more potatoes, and I'd like some deep information on the driver board for the A6. I'm trying to think of a way of powering a 6200k 4up XPG (this) and I wonder if the A6 board would have a happy time doing it? I'd potentially be powering the system with 14500s or rcr123s in parallel, with all 4 dice in series (for a 4A draw, max... okay maybe 6). I just thought that seeings how the A6 driver is so small, and available, it'd be easier than trying to make my own driver unit.
You can use the a6 driver with the cells and leds in parallel. Running either in series will require different components. There is no limiting the current on an fet driver so it will pull whatever amps the leds like. 4x xpg2 should pull 10-12 amps total from an average cell. It will get nice and bright
I'm pretty squared up on MCPCBs and heat management for overdriving dice (thanks to a study in lightsaber building), but I'm trying to square away as much cool white light for as little residual heat as I can so I can maybe mount it into a plastic housing (mainly because I don't have tooling to make aluminium, steel or titanium housings, and I was considering seriously getting a 3D printer). I'd also assume for a forward current of total 4A, and a supply voltage of 3.7V, the most the unit will see is 15W. I just wanted to know how the A6 driver drives to see if it would be suitable for sending the angry pixies to the futon emitters. I mean simply having a plug-in-and-go driver is going to be easier and cheaper and smaller than trying to make my own...
I guess this is going to end up with me and the board out of the light I just ordered and my chinese multimeter... again.
You could get a Nanjg 105C with 3 brightness modes and 8x 7135 = 2.8 Amperes.
They’re not expensive and have high PWM frequency (no strobe effect).
Just connect 4 LEDs in parallel, each will get 0.7 Amperes.
The difference between 2.8 and 4 Amperes is not big, it’s hardly noticable.
15W driven LED on longer runtimes than a few minutes is a serious issue for heatsinking in a 3D printed light
Even a Convoy S2+ with 2.8A doesnt handle about 7-8W heat even if the body is from Aluminium
it gets too hot left alone reports from died lights that way are not unusual
and even held in your hand it gets uncomfortably hot to hold after a few minutes
I'm actually more worried about the case than the LEDs, an increase in temperature will increase the forward current over the die, but a constant current source will adjust for this. As long as enough of the heat is dissipated to keep the leds below their thermal limit it should be fine...
If you want a known, fixed maximum current then either a 7135 driver (i.e. 105C,D or other) with the appropriate number of 7135’s (approximately 11-12 for 4A) or one of Led4power’s LD-3 drivers with a 4A sense resistor. The latter is the best off the shelf choice if you’re not into stacking 7135’s. Alternatively to stacking you could slave the output of a 1A 7135 board to a 3A 7135 driver.
The quad board with the old generation XP-G’s that you link in the OP seems rather expensive to me, I think I would not grant Ledsupply the pleasure to get rid of their old stock for this amount of money.
