Hi all,
When the BLF Q8 came out I threw one in my Amazon cart with full intention of buying one very soon, probably will eventually but the real excitement was when the Q8 quad reflector and matching MCPCB went up for sale on the Sofirn store. Single,triple and five-up mags are great and fun to build but my very favorite are quads, probably cause they work great with both single and double cell applications of FET drivers. In this build we use the Q8 reflector and MCPCB to upgrade a 2C mag to around 5000 lumens with Samsung LH351D emitters. Just for kicks we’ll use a D4 ramping driver.
We start by re-flowing the emitters onto the rather large MCPCB. I do these with a soldering iron but that sort of thing becomes a chore with large multi-emitter boards. Might be time for a hot plate. The centering rings don’t come with the reflector or the MCPCB but fortunately Sofirn c8F kits come with way too many. As a slightly related side note, these MCPCBs can be easily modified to 2S2P configuration with careful xacto knife work and a couple copper jumpers.
The next step is to devise an appropriate heat sink. The lens/reflector/centering ring/MCPCB stack leaves about 7.5mm in vertical space for the heatsink. Hacksawing off a slightly over-large chunk off a 1.875” diameter aluminum rod and sanding to fit does the trick. A hole in the center for wires works just fine as we won’t be screwing the MCPCB to the reflector in this build. I don’t show it here but a partial wrap of 30 gauge copper sheet will make it a tight press fit. Also not shown is attaching the emitter board to the heatsink with thermal epoxy.
The reflector doesn’t sit far enough down in the head for the bezel to be fully screwed down so we need to alter either the reflector or the mag head. It was fairly easy to sand about 3-4mm off the height of the head at the threads and then use a metal cutting bit to chamfer the inside edge.
The reflector now fits in the head like they were made for each other.
As mentioned above I wanted to use D4 firmware so I put together a D4 driver with the 17mm FET+1 driver board from Mountain Electronics. I hand build all my drivers but I’m thinking there must be a quicker way. Might be time to get a hot plate. The 18 gauge wires will go to the emitters while the 24 gauge go to the momentary switch.
Another spot of necessary adjustment concerns the length of the battery tube originally designed for a pair of C-cells versus the lengh of the 26650 cell we will be using for the modified light. Without getting into the math, we’ve got to account for around 22mm of extra length in the tube. Options are either cut the tube of create a spacer. It’s work cutting a mag tube and I dislike dummy cells so I decided to make the spacer integral with the switch body. Bonus is I finally found a use for plastic 18650 to 26650 adapters! Simply cut the adapter to length, cap it with a piece of plastic lens dremeled from the original mag lens and epoxy it to the back of the switch body. At some point in the process it is wise to create a positive battery contact connected to a thick wire which runs through the switch body to the front of the switch.
The wire from the battery positive contact is soldered to the back of the driver while the negative ring is connected to the switch anchor for ground.
The two switch wires are run through a pair of holes drilled horizontally through the switch body and are soldered to the momentary switch contacts.
After a function check it seems like a good idea to anchor the driver and switch wires with hot glue.
We’re about ready for final assembly but of course the battery we want to use doesn’t actually fit into the battery tube. Last step is to sand the inner diameter of the battery tube to fit the cell. I use a large drill bit wrapped in duct tape and then wrapped in adhesive sandpaper. Chuck the bit in a drill press and prepare to be bored with a hot hand for about a half hour or so.
A couple notes: I used the AR lens from flashlightlens.com. Also, while the mag tailcap spring is big and heavy, one can realize 15-20% in lumen gains by bypassing the spring with 18 gauge wire. The high forward voltage of the Samsung emitters limits tail current to around 14 amps with a full Liitokala 26650. That combined with ceiling bounce lux numbers that are only relevant to me suggest OTF lumens are slightly below my 5000 lumen target.
All finished!
Here it is next to a five-up XP-L2 with shortened battery tube.
Beamshots! About 60 yards to the sugar shack.
Sale thread coming up soon! If it sells I may buy a hot plate…
Thanks for looking,
Brian
Love your work.