Build thread of sinner's Tri-EDC

Flashlight Modding and DIY Parts
1

I recently received one of sinner’s Tri-EDC hosts, and was planning to use a triple XP-G2 on Noctigon and a BLF17DD as the driver. Here’re some notes and pics:

The Tri-EDC:

Disassembled (from left to right: head section, tailcap section, and body:

Front view of the tailcap:

The pill. That’s the battery end in the first shot. You’ll notice that there aren’t any holes on the edge of the pill. I recommend drilling two holes, on opposite sides of the pill, so that you can use them with tweezers or needle-nose pliers to screw the pill into the head. Otherwise, you won’t be able to do that, and you have to screw it in almost the full length/depth of the head:

BTW, the threads may not be exactly perfect, e.g.:

I took the pill, after drilling the holes I mentioned above, and screwed the pill into the head and out of the head several times, which, although it doesn’t clean up the threads totally, helped a bit (I think that sinner said the copper is kind of difficult, I guess because it’s so soft?).

After assembling the light with the BLF17DD and the triple XP-G2 and a Carclo:

Notice the gap between the tailcap section and the body, and this was after screwing it so tight that I dented the top of my batteries :(…

I spent the next day or so, getting this resolved/fixed, with a lot of help and suggestions from Dale/DBCstm and from RMM.

Here’re my notes about what I had to do to get this build done without the gap. Note that I’m not saying this is a definitive list, but this is just what I did/noticed

- I found a smaller/shorter spring, and replaced the original spring on the BLF17DD. The BLF17DD from RMM comes nicely done, with solder wick on the spring and all, but I mainly did this because the original one got bent over.

- The wires on the BLF17DD were kind of thick, and I think that they might’ve been part of the reason that the driver wouldn’t seat properly into the pill, but both RMM and DBCstm kind of suggested avoiding going to smaller gauge wires.

The way the driver came originally, both the red and black wires were soldered such that they naturally routed around the side of same chip, which is probably ok if the holes in the pill are on the sides, but, in my case, I wanted to run both wires through the center hole, and I think in this case, the thickness of the wires, on top of the smaller chip, was part of the reason for the seating problem.

So, after discussing with DBCstm, I decided to unsolder the red wire and resolder it such that I could run it around the other side of the chip, i.e., so each wire was routing around different sides of the chip.

- Even after re-routing the wire, I found that while I could seat the driver better into the pill, it still felt like it was “rocking” a bit (sorry - technical term), so I looked at the driver and pill for awhile, and noticed that the solder point for the positive lead was very close to the edge of the driver, and also, the “ledge” inside the pill did not go very deep, so I theorized that maybe the soldered point was sitting on top of the ledge. So, I took a small diamond file, and kind of cleaned up the solder joint, so that was no longer so close to the edge and also “lower”.

After that, the driver sat stably inside the pill. But, note that, at least for the BLF17DD, which I think is kind of thick, the battery side of the board is still not level with the edge of the Tri-EDC pill. So, I had to do some (really ugly) solder bridging between the pill and the driver edge (sorry, no pics for that).

Even after ALL of that, there was STILL a gap. More PM’ing with Dale, and he suggested filing down the brass stud in the tailcap and also the nylon retaining ring. It took a while to do that, because I didn’t want to go too far, but I eventually got it to the point that that gap was gone.

Note about the tailcap: The nylon ring unscrews. Then, push on the brass stud, and the switch and stud piece will pop out of the other end of the ring. I followed Dale’s “technique” for the filing: 5 strokes, then turn quarter turn, and repeat.

Seemed to work :)!!

So, thanks to DBCstm/Dale, RMM, and also, sinner, for my new “hand warmer” :)!!

2

That is a great looking setup! Any idea of the actual output?

3

Nice effort. Due to the thousands of different component combinations that could fit into this light it does not surprise me that a few mods were necessary. Copper is a real pain to machine and what I could see looked ok. Again you have a lovely little double custom light.

Make that a triple custom if you include the help from your fellow blfers that helped you.

4

Hi JM,

Thanks. I was kind of a sceptic about how fast this thing heats up, but seriously, it starts getting really warm within seconds of putting it in turbo :)!

I don’t have a way to measure lumens, but FYI, here’re some tailcap currents:

- With red laptop pull: ~6.85 amps

- With VTC4: ~8.57 amps

  • With Eagletac (protected - using meter so no tailcap, so length isn’t a problem): ~7.2 amps

The above was with batteries I had been using for testing while I did the build, so none of them were fully charged (they’re in the charger now).

Jim

Edited description of Eagletac testing.

5

Thanks, it looks like I’ll have to do some of the same things to get mine to fit together.

Actually, I already did the thing with routing wires around both sides instead of just one. I still need to reduce the spring size though, and I might try to reduce the negative contact post a little too if necessary.

I can probably also save part of a millimeter by using a de-domed XP-L instead of the domed one I’m using now. I was hoping for better tint by using domed 3D emitters, but the tint turned out a little weird anyway even with an optic which almost fits. So, I may as well use the de-domed 2C emitters (which look distinctly yellow but only a bit more than the squashed 3Ds). Or perhaps later I’ll go for triple Nichia 219B. I’m on the fence about whether tint or lumens are more important on this light.

6

It worked!

Mostly, I just had to reduce the spring. Details are in the Cypreus thread (link assumes you have the default number of posts per page):