LilKevin715
Enlightened
The main goal of this project was to have a torch good for the great outdoors with lots of throw. I wanted to balance runtime and output as much as possible, hence the "tame" 2.5A 3-mode driver. I originally planned on using a SSC P7, but the thought of a donut beam with a SMO reflector didn't really vibe too well with me:sick2:. The choice of using 3 x C batteries allows me to use alkaline batteries in an emergency situation.
Specs of new project:
Mag 2D Gray Rebel Deep SMO
Luminus SST-50-W45S-F21-GH400
Der Wichtel SST-50 Deep Heatsink
Shiningbeam 2.5A 3-Mode PWM Linear Driver (8 X AMC7135)
20 AWG Silver coated Teflon jacket wire
3 x C Accupower 4500mAh NiMH LSD Batteries
52.1mm Hardcoat Acrylic Lens
The hardest part of this build was prepping the emitter. It's pretty nerve wrecking soldering wires to the underside of a SMT emitter by hand
. I now have a much higher appreciation for emitters pre-mounted on stars. I didn't take pictures of the wires soldered to the emitter before mounting the emitter on the heatsink, so I guess the following picture will have to do.
I've seen most people here soldering their SST wires parallel and directly on top of the solder pads on the underside of the emitter. I wanted to be different and soldered the wires perpendicular to the solder pads. Durability/strength of this soldering method might be questioned, but as long as it isn't a cold solder joint it is probably fine. By soldering the wires this way I didn't have to bend the wires when mounting the emitter on the heatsink. Arctic Alumina Epoxy was used to mount the emitter to the emitter pedestal, while Ceramique was used as the thermal compound between the bottom heatsink base and the emitter heatsink.
After the epoxy was dry I filled the holes for the wires in the bottom heatsink with some Fujik thermal adhesive. I did this for two reasons: to prevent the wires from moving around too much while I prepped the driver, and later on when I mount the driver.
Next I prepped the driver. I unsoldered the stock wires (26-28 AWG?:shakehead) and soldered on the battery wires on the opposite side. I then soldered on the emitter wires at a parallel angle to the driver board so they would lay down flat when it was time to attach the driver to the heatsink.
I originally planned to mount the driver directly to the bottom of the heatsink. However due to the wire thickness and the layout of the components on the driver board the driver would not lay down flat on the heatsink. I could have used a 2-3mm thick layer of Fujik to attach the driver to the heatsink, but that would have been the sloppy way IMO. In most cases I am a firm believer of less is more when it comes to thermal paste/adhesives. I decided to use some Akasa aluminum BGA ram heatsinks (13x13x4mm) as a way to mount the driver. I figure the 4mm aluminum BGA heatsink is a better thermal conductor than a thick layer of Fujik of about the same thickness. I used Fujik to attach the BGA heatsink to the bottom heatsink.
After the Fujik dried on the BGA heatsink I then proceeded to rotate the driver 90 degrees clockwise as shown in the picture and mount the driver to the BGA heatsink with some Fujik. I had to clamp the driver down for about an hour to ensure that the driver didn't pop up or rotate while the Fujik was drying.
At this point the emitter side of the driver is finished in terms of heatsinking, but the battery side still needed to be dealt with. I then proceeded to modify some more square BGA heatsinks by using a metal file to change the shape to match the round walls of the bottom heatsink. The results are excellent and I don't have to worry about the possibility of the AMC7135 chips thermally throttling down. FYI on Der Wichtel's heatsink the depth of the area to mount a driver is 12mm deep. The shiningbeam driver I used was about 5mm thick. Combined with the two layers of BGA heatsinks the battery-side BGA heatsinks pop out around 1mm from the bottom heatsink. This is important to note as there isn't a lot of room in between the heatsink and the snap ring of the maglite. The "engine" of the new torch is now complete.
Now that the hard work is over it was time to modify the stock switch by removing the stock rebel module and cutting off the stock pedestal with a hacksaw. I then used a metal file to file down the plastic. The battery wires were then soldered to the contacts on the mag switch. Afterwards I modified the stock reflector by cutting off the cam and filing down the plastic until it was close to the same level as the opening of the reflector.
It's time to place the modified switch housing, snap ring (PITA to get out BTW), and completed "engine" inside the mag host. I purposely left about 2 inches of wire on the battery end of the driver as I knew the extra length would be needed at this point. The switch goes back in, then the snap ring, and then the heatsink that is lightly coated with some Ceramique thermal paste. I rotated the heatsink so that the wires would coil up slightly before going in. This would ensure that the extra wire length wouldn't prevent the heatsink from going down all the way.
The choice of 3 x C batteries in a 2D Mag host requires a tailcap modification. I used Liquid Plumr Gel to remove the anodize inside the tailcap. The stock spring was also trimmed slightly. A piece of 1 inch ID/1.25"OD PVC pipe cut to a length of 5.25 inches was used as a sleeve for the 3 C batteries.
Not much to look at except for the business end on the flashlight with the hardcoat acrylic lens installed; it's a good sleeper light:naughty:. The new deep rebel reflector truly is a thing of beauty compared to the old incan reflector. I'd estimate around 30-40/180-200/450-500 OTF lumens (0.14A Low/1A Medium/2.5A High).
This was my first time assembling a light from scratch and it was an interesting experience to say the least. I've learned that it takes a lot of patience and proper research is of the utmost importance. It's nice to buy a light from a manufacturer as it saves you a lot of time. But the thought of making your own light how you want it is much more satisfying and adds that special feeling that you can be proud of.
lovecpf
Specs of new project:
Mag 2D Gray Rebel Deep SMO
Luminus SST-50-W45S-F21-GH400
Der Wichtel SST-50 Deep Heatsink
Shiningbeam 2.5A 3-Mode PWM Linear Driver (8 X AMC7135)
20 AWG Silver coated Teflon jacket wire
3 x C Accupower 4500mAh NiMH LSD Batteries
52.1mm Hardcoat Acrylic Lens
The hardest part of this build was prepping the emitter. It's pretty nerve wrecking soldering wires to the underside of a SMT emitter by hand
. I now have a much higher appreciation for emitters pre-mounted on stars. I didn't take pictures of the wires soldered to the emitter before mounting the emitter on the heatsink, so I guess the following picture will have to do. 
I've seen most people here soldering their SST wires parallel and directly on top of the solder pads on the underside of the emitter. I wanted to be different and soldered the wires perpendicular to the solder pads. Durability/strength of this soldering method might be questioned, but as long as it isn't a cold solder joint it is probably fine. By soldering the wires this way I didn't have to bend the wires when mounting the emitter on the heatsink. Arctic Alumina Epoxy was used to mount the emitter to the emitter pedestal, while Ceramique was used as the thermal compound between the bottom heatsink base and the emitter heatsink.
After the epoxy was dry I filled the holes for the wires in the bottom heatsink with some Fujik thermal adhesive. I did this for two reasons: to prevent the wires from moving around too much while I prepped the driver, and later on when I mount the driver.
Next I prepped the driver. I unsoldered the stock wires (26-28 AWG?:shakehead) and soldered on the battery wires on the opposite side. I then soldered on the emitter wires at a parallel angle to the driver board so they would lay down flat when it was time to attach the driver to the heatsink.
I originally planned to mount the driver directly to the bottom of the heatsink. However due to the wire thickness and the layout of the components on the driver board the driver would not lay down flat on the heatsink. I could have used a 2-3mm thick layer of Fujik to attach the driver to the heatsink, but that would have been the sloppy way IMO. In most cases I am a firm believer of less is more when it comes to thermal paste/adhesives. I decided to use some Akasa aluminum BGA ram heatsinks (13x13x4mm) as a way to mount the driver. I figure the 4mm aluminum BGA heatsink is a better thermal conductor than a thick layer of Fujik of about the same thickness. I used Fujik to attach the BGA heatsink to the bottom heatsink.
After the Fujik dried on the BGA heatsink I then proceeded to rotate the driver 90 degrees clockwise as shown in the picture and mount the driver to the BGA heatsink with some Fujik. I had to clamp the driver down for about an hour to ensure that the driver didn't pop up or rotate while the Fujik was drying.
At this point the emitter side of the driver is finished in terms of heatsinking, but the battery side still needed to be dealt with. I then proceeded to modify some more square BGA heatsinks by using a metal file to change the shape to match the round walls of the bottom heatsink. The results are excellent and I don't have to worry about the possibility of the AMC7135 chips thermally throttling down. FYI on Der Wichtel's heatsink the depth of the area to mount a driver is 12mm deep. The shiningbeam driver I used was about 5mm thick. Combined with the two layers of BGA heatsinks the battery-side BGA heatsinks pop out around 1mm from the bottom heatsink. This is important to note as there isn't a lot of room in between the heatsink and the snap ring of the maglite. The "engine" of the new torch is now complete.
Now that the hard work is over it was time to modify the stock switch by removing the stock rebel module and cutting off the stock pedestal with a hacksaw. I then used a metal file to file down the plastic. The battery wires were then soldered to the contacts on the mag switch. Afterwards I modified the stock reflector by cutting off the cam and filing down the plastic until it was close to the same level as the opening of the reflector.
It's time to place the modified switch housing, snap ring (PITA to get out BTW), and completed "engine" inside the mag host. I purposely left about 2 inches of wire on the battery end of the driver as I knew the extra length would be needed at this point. The switch goes back in, then the snap ring, and then the heatsink that is lightly coated with some Ceramique thermal paste. I rotated the heatsink so that the wires would coil up slightly before going in. This would ensure that the extra wire length wouldn't prevent the heatsink from going down all the way.
The choice of 3 x C batteries in a 2D Mag host requires a tailcap modification. I used Liquid Plumr Gel to remove the anodize inside the tailcap. The stock spring was also trimmed slightly. A piece of 1 inch ID/1.25"OD PVC pipe cut to a length of 5.25 inches was used as a sleeve for the 3 C batteries.
Not much to look at except for the business end on the flashlight with the hardcoat acrylic lens installed; it's a good sleeper light:naughty:. The new deep rebel reflector truly is a thing of beauty compared to the old incan reflector. I'd estimate around 30-40/180-200/450-500 OTF lumens (0.14A Low/1A Medium/2.5A High).
This was my first time assembling a light from scratch and it was an interesting experience to say the least. I've learned that it takes a lot of patience and proper research is of the utmost importance. It's nice to buy a light from a manufacturer as it saves you a lot of time. But the thought of making your own light how you want it is much more satisfying and adds that special feeling that you can be proud of.
lovecpf
