Re: Just another MAG Mod, SST90, Aspherical, AMC7135, 3P IMR26500
Some of the important bits arriving in the mail.
More mail. Teflon wire from eBay.
18 AWG wire soldered to SST-90. This was probably one of the harder parts of the build. Getting the wire bent down at the right angle wasn't easy.
LED glued to Britelumens heatsink using Arctic Silver Thermal Adhesive. Good stuff.
The AMC driver boards. The one on the right has the diode removed.
A driver board ready to stack. I have cut the track around the outside to give me space to drill one of the holes.
3 drivers stacked. I have left the wire sticking out the top while I was testing. They get cut off later.
The proposed electrical schematic. The pin names shown on the AMC7135 board stack should be 'OUT' and 'GND'. LED - connects to the OUT pin and the GND pin connects to battery -. I am switching the connection to the VDD pins to control power to the LED.
The first power on with 2 dead AAs. It's working!
PART 2: Added 10 May
The usual switch mods
Top
Bottom
The connection between the switch and the regulator GND pin. You can also see the connection from battery + to the LED + which bypasses the switch.
The 3 connections comming out the top of the switch. The wire comming from the center goes to the VDD pins on the regulator boards. The switch is not modded for momentary action.
POWER ON! A test using 3 parallel regulator boards. It works!
My batteries were only charged to 4.00V and I was getting 4A through the LED. That works out to 1/3 Amp per AMC7135 chip. I think that's about normal.
I tried measuring the voltage across the LED at this current and only got 3.118V? I don't think I had modded the tail spring at this stage. Also, I wasn't measuring right at the LED???
Now, along comes another 3 x AMC7135 driver boards. These will be connected in parallel to each other.
Connections between GND and OUT pins.
Connections between VDD and OUT pins.
And it will just fit inside the 3D tube.
All connected, ready to drop into the top of the mag tube.
The usual tail spring resistance mod.
Here's my parallel holder for 3 x 26500 cells.
I've started a seperate thread about it here.
And a tail spring extension. My holder with 3 x 26500 are shorter than 3 x D cells. Without this extension, the spring only just touches the batteries.
Current check with all 6 parallel regulators. 8 Amps.
Again this is 1/3 Amp per AMC7135 chip. I was only getting 4.5A using the standard multimeter probes. Almost double with some thick wire. This test was also with the batteries at 4.00V. The current seems to stay fairly constant. I haven't got any thermal paste between the heatsink and mag tube yet so I dont want to run it for too long. I'll do a longer current test someday soon.
The finished mod! Looks like a standard 3D Maglite.
Well, almost like a standard Maglite. The DX aspherical gives it away. :laughing:
The Next Step
I'd like to add some dimming functionality for longer run times. Maybe by adding a D2Flex to PWM the VDD pins on the AMC7135 regulators.
Or maybe a simple DIY pwm circuit using a PIC microprocessor
Or I may just replace the AMC7135 boards with a suitable multimode buck driver once they are available (H6Flex or sector_cleared driver etc). This will allow me to ditch the parallel battery setup.
Beam shots to come when I find a good place to do them.
Thanks for reading. :twothumbs