Yes i was trying to do that, but im not good at writing code. Is there something to know about torroids? something about PWM? I updated my convoy to NLITE, thats not problem. I ordered some attinys…
and button is from ebay, it had to file flat at two sides to fit.
That's pretty cool Comfy. I haven't seen that anywhere.
Jeepster,
Unfortunately, I don't know the answers to your question. I would guess the inductors are not needed with such a high PWM as the Nlite, but that is just a guess. I was planning on leaving them in place at first to see what happens.
I haven't really thought it through yet as far as how to control the Attiny. I was thinking of trying to use a normal reverse clickie kind of switch that connected between battery positive and the VCC pin on the Attiny. Wouldn't need to be a high amperage switch since it would only drive the mcu and the PWM signal that the MCU outputs.
EDIT: Post 14 for the following thread has a Attiny wiring schematic. You can see power is also applied to pin 7 for voltage monitoring. I don't know if it has to get power, but it would be easy to duplicate the ciruit of the Nanjg 105C so that you have low voltage protection.
This is on the back of my XML2 Quad. The braid is polished to remove high points on top of springs.
!
!
> and button is from ebay, it had to file flat at two sides to fit.
I'd love some more details... :)
Most of these buttons seem to big for the hole.
So here's my plan (only requires 4 wires):
I'm numbering the MCU pin pads counter-clockwise starting on the bottom left. Nanjg should not need any heat sinking. So it will just float.
See any potential problems with this?
Switch would need to be high current clicky… Nanjg drivers don’t work with momentary switches. Dr Jones has firmware that does…
Why? it would only be powering the MCU.
How would you turn the light on/off. Button would have to be held down to keep the light on. With driver circuit powered and CPU off, no telling what would happen.
The electronic switch would need to be replaced with an on/off type switch.
The rest of the SRK driver always has power in the stock configuration. The FETs don't allow power to flow unless positive voltage is applied to the Gates. With positive power cut to the Nanjg, no positive can flow to the Gates.
EDIT: I'm aware DrJones sells an electronic switch version (though he has not posted prices I'm aware of) and I'm looking forward to your driver when it comes out. I'm doing this just because I can. I don't see it being a solution I will keep in the long run.
Ah, now I see, you bored up the hole it seems :) Thanks for the pic!
The SRK drive PWM signal does not go to a FET. Also, if there is no pull-down on a FET gate, it will float to who knows what level and may or may not turn on. You can’t tell what the output of an unpower processor pin will be at.
Oh, now that does sound like a problem. Thanks for the input. Back to the drawing board.
dang it I don’t have any 30K resistor… I got 47K… would it work if I solder this 47K parallel to 15K? what kind of gain will I get?
I have 10K but removing those 153 resistors is PIA 
What would it do for me if I direct drive 3 XML’s, running the four batteries in two parallel sets ie. 8.2 volts? would that smoke the XML’s or give crazy light?
Only if you have really crap high-internal-resistance cells that sag like crazy, and you get extremely lucky. If you try it and it works, go buy a lottery ticket ASAP.
I've done something similar, but with more LEDs to share the pain. Five XR-Es wired in parallel, and two crappy 18650s in series, direct drive. It still pulled around 12 amps, melted several switches and turned the springs into carbon (the springs would glow like the filament in an incandescent bulb). I don't think it would work with only 3 LEDs, and doubling up to 2S2P.
Yes 
Nice job O-L and Pavithra! 8)
How many resistors did you change in total? Was the resistor change the ONLY thing to do to get amps -> 3A?
Thanks for saving me again Comfychair. I gotta get you on speed dial before I hurt myself.
Mouseover to make all the components disappear...
I'm still trying to figure out why it's suggested to mess with the tiny 15k resistors instead of the plain ol' sense resistors. Sub-1 ohm resistors aren't hard to find, they're bigger and easier to get to on the board, there's plenty of room to stack multiples, and using multiple .02-.03 ohm parts instead of the original .01/.1 ohms will give finer resolution for fine tuning and balancing the channels...
Also, if you look at the traces from the sense resistor pads to whatever chips they lead to, and then go back and look at one of the earlier pics with copper jumpers, it looks to me like the jumpers would do squat. The original traces are plenty fat enough as-is.