Relic's DIY Build for 2013 - The Copper Triple-Triple
Saturday June 8
They say copper is expensive these days... they are correct. You are looking at a lot of money in plumbing parts. A lot. In fact, this light will be the most expensive light I have when it's done. :D
More to come, just giving a taste of what to expect from my side; a nice and shiny triple-triple ;)
Friday June 28
I got quite a bit done this evening. Lots of cutting and grinding. A lot. let's start with the main host parts, in order from tailcap to bezel:
Starting on the left we have:
- Coupling that will be sliced, flattened, and a circle cut out to fit into the end of the tail
- Two larger diameter couples for the body where the 3x18650s will go (later I opt for a single long couple I found on another HD trip)
- 1.5" Male threaded adapter for the body to head interface. I will cut off and use only the top 11mm of the threaded end
- 1.5" Female threaded adapter. The other half of the head-body interface. I will cut this off so it forms the base for the emitter platform
- Emitter platform (not shown) This will be made from two pieces for sliced/flattened couple, soldered together. Stars will be mounted on this surface
- Reflector (not shown), basically an SRK/TR-3T6 reflector.
- 2" Female threaded adapter. No cuts needed. This will be soldered to the head/body interface. The reflector and lens will go in here
- Lens (ordered from FlashLightLens.com), hopefully a 55mm UCLp lens will fit just right.
- O-Ring (not shown), probably made from two smaller ones. This will mostly prevent the bezel from grating directly on the lens. This light will not be waterproof.
- 2" Male threaded adapter, brass. A thin slice will be cut off the threaded end to form the bezel.
OK,I started the slice, but to quickly show how the cells will fit. Again, I do swap for a single couple for the body, which was about 3/8" shorter than these two combined. Saved me from shortening one and having to solder them together.
Here I have the tail piece split and flattened, as well as the head/body interface cut. I needed this to template the round shape onto the flat. The tail will go inside the body. The other one sits on top of the head interface as the emitter plate, so it is cut to outside diameter.
After cutting out the tail end piece, it is now in place. It is a very snug fit, so it will not fall out before I solder it. This part turned out better than I was thinking it would (here's the one-piece couple).
Now I have the body/head interface cut off (second from left). All the parts are laid out again. The SRK came over to see what is going on...
And now we can get an idea of the size, relative to that nosy SRK. Shorter, smaller body, bigger head.
Well, that's all for tonight. I got a lot more done than I expected. I wore out six or seven cut-off wheels today (four shown). I still have a lot to do. Soldering the body will be interesting. I will have to do all the joints in one shot. The copper is not going to allow me to heat up just one end. I also have to add cell retainers inside the tube, which I haven't quite figured out yet. There will be springs soldered onto the bottom as well.
After that there's the head, which also will be a one shot solder job. The interface, head and emitter plate will all get soldered at the same time.
Once everything is in place in the head, I can screw things down with the bezel, mark, and cut it exactly the thickness I want it.
I haven't mentioned the driver yet... and I'm not going to either :D
Until next time (hopefully tomorrow).
Monday July 1
Quite a bit of progress over the weekend. Let's start with the head. I cut the emitter plates,
which will be soldered onto the head interface.
Getting the star mounting surface perfectly flat took a lot of lapping. A lot. Eventually, I got a nice shiny finish.. at least where I needed it anyway.
OK, all ready to solder. Because the pieces are so thick and made of copper, I do not expect to be able to solder one piece at a time. So I came up with a way to use my hot air reflow station for this; household insulation. The head is already soldered here; I forgot to take pics when I did it. Basically the work piece is sitting on a thin piece of Roxul.
Then a piece with a hole in it is laid around the workpiece.
Finally a small piece goes on top to hold in the heat, leaving a small hole for the reflow tip and exit airflow.
All of the interfaces received solder paste and were set in place, held by gravity (for the head, the tail/body needed a slightly different approach).
I set the reflow station to 310 degrees C, put on the largest tip, and set the airflow to 6 (out of 10). About 7 minutes later, the paste melted. I removed the upper layers of insulation and checked what I could see; all looks good. Took away the head and set up a small fan to cool things down. 10 minutes later:
Driver end:
Emitter end, discoloured from the heat and paste fumes which is not an issue:
The entire surface would not lap perfectly (without a few more hours anyway), so I stopped when I had enough area to put down the stars.
OK, on to the body. I had a couple of different ideas on how to keep the cells in the right position within the body. They would need to stay over the springs. I needed to add something to the inside to provide guides, similar to how an SRK has built-in cell holders. I thought of bending some copper into a V shape and soldering it there (more cutting and grinding). I had a suggestion from Comfychair to cut copper circles with three cell holes in them and put them in the tube, but the cells are too close together to allow any material between them. Then I thought about some 1/2" copper pipe...
sliced in half...
Now for the dry fit... perfect!
Now, I need to hold them in place when I solder them... maybe I can make a spring from copper wire...
OK, now everything should stay in place while I heat up the body. Similar to the head, add solder paste to all interfaces and repeat like the head.
The moment of truth; will the cells fit? Yes!
I forgot to mention; that body interface piece on top? It needed a lot of grinding to open it up enough for three cells to fit through. Also, this light will not accept protected cells which is fine for me.
Oh and it does screw together. This is one compact light for how much it weighs. I thought about adding an in-hand shot, but I'll save that for another day ;)
Tuesday July 9
Not a lot of progress in the last few days, but it is an important step. The emitters are now in the head.
I'm waiting for the small spots of Fujik to set. The majority of the thermal compound is silver thermal compound. I add a dab or two of Fujik so the emitters will not move around when I take the reflector out to wire the emitters.
I started with a plan to use XM-L2 U2 1C emitters, but I figured that was a rather cool tint for a copper light. So, these are XM-L2 T5 5C1 75CRI. I've put one in a P60 Drop-in and I really like the tint. I cannot wait to fire this thing up. Getting close. Needs emitters wired, a contact plate and driver, and the lens to come in so I can cut the bezel. Then, a lot of cleaning and polishing.
Friday July 12
IT'S ALIVE!!!!!!! Finally, I have it all together, and it lights up. Boy, does it ever light up! WOW! 3x105c drivers, the master is a QLite. The modes; perfect! Couldn't ask for better. The moonlight mode is super low, I mean firefly low. I guess the PWM at 15kHz is not driving hard enough for 24 chips, so they barely turn on. I like.
It didn't take me long to crank it on to High and throw it on the lumen-box. I measure 2440 lumens at start, 2400 lumens OTF after 30 seconds. And get this; the plastic protective coatings are still on the lens! :D Could not be more pleased. Pics later tonight! Probably the biggest twisty light you have ever seen ;) Edit: Not quite; I forgot about the Electrolumens Bruiser lights :O
OK I took the protective lens coatings off and it jumps to 2530 lumens after 30 seconds. This is well above the target I was shooting for. Sweet! (Update: very slightly higher output measured later on, see measurements below)
(Added July 14)
Here are a few shots of the driver. I got a bit anxious to get it going, so I forgot to take more pictures. The driver is basically three NANJG style drivers with one being a master.
The driver ground rings are connected with 14ga copper wire. This was mostly to hold them together, since each driver was getting its own ground connection anyway.
The Master is a QLite set to four modes; Moon-Low-Med-High. The master will control all 24 7135 chips with one Vcc output. Each driver Out- will control one emitter. This is not the same as putting the emitters in parallel and sinking them with one big Out-. By dedicating each Out- to one emitter the current to each emitter is known (in this case, about 3.04A each).
The driver attaches to the backside of the emitter plate with Fujik and a dab of super glue (so I do not have to wait hours).
There is very little driver space, so my contact plate will sit right on top of the drivers, held in place with Fujik and super glue. In order to do this, all of the wires need to sit down lower than the top of the 7135 chips. Now I add the wires and prepare to add the contact plate.
I didn't grab a picture, but the contact plate does get wired to the drivers. This was the tricky part. I drilled three small holes in the ground ring and one larger hole in the center of the positive center. Each driver ground wire was stripped down to almost the bottom, tinned, and fed through the holes in the ring. The emitter positive and master Vcc was stripped and fed through the center hole. I slowly worked the contact plate down the wires until it sat on the drivers that had the adhesives already on top. Hold for a minute or two and then solder the wires that poke up through the holes. Cut them off as flush as possible. The central positive solder bump needs to be small in order to not interfere with the cells. And it has to be dead-center.
That's it! Driver done! Only polishing left to go.
Saturday July 13 ----- IT'S DONE!!!!!!!!!!! ----------------------------------
At long last, I am done! I have some more build shots to add as well, they will be there by tomorrow night.
Here's some shots of the result. I spent a fair amount of time sanding and polishing this thing, just to have it return to normal in a day or two. At least the shots came out OK. Enjoy!
Relic's DIY Handmade Scratch Build for 2013: The Copper Triple-Triple
Summary:
Battery: | 3x18650 (Unprotected only) |
Switch: | Twisty |
Modes: | H 100%, M 30%, L 3%, Moon, with memory
|
LED Type: | Cree XM-L2 T5 flux bin, 5C1 tint bin |
Lens: | UCLp Plastic |
Tailstands: | Yes |
Price Paid: | Estimated parts cost: $110 |
From: | Parts from various stores |
Started/Completed: | June 9 / July 13 |
This has been quite a learning experience for me. I've never made anything significant out of metal before. Back in high school I turned a cannon from aluminum and built a few decent boat anchors. Nothing even close to this, and especially nothing by hand with only a Dremel and a Drill. I have quite a sense of achievement right now. I'm very happy with how it turned out. Let's have a look.
I left all of the hints that this is made from plumbing parts.
Tailcap isn't perfect, two small gaps, but that part is water-tight (the light as a whole isn't waterproof).
This was polished 15 minutes before the shoot. It doesn't take long for copper to turn. Already has some fingerprints. I'll let it return to normal, takes too much time to maintain a shine.
The size is surprising, it's quite small for a triple.
The grip is good, my hand almost covers the whole body and head. It's easier to hold than an SRK, a little slipperier though.
Speaking of the SRK, here they are side-by-side. I put an SK68 in there too. Almost everyone has one or the other (probably both).
One final shot of the brass bezel.
BEAMSHOTS!
All I have are wall beamshots for now, I'll add some outdoor shots tomorrow.
The beam is buttery in smoothness and tint. I really like it, and it matches the light really well.
Moon:
Low:
Medium:
High:
Outdoors:
Mouseout CTT, Mouseover SRK XM-L2 (3000 lumens)
The SRK takes it in apparent brightness and throw, but to me the CTT wins the tint race, hands down. It was windy for the SRK beamshot, so things are blurry.
The mode spacing is just about perfect. The moon mode is a barely there firefly mode. And sometimes it really isn't there! If the light is a bit hot, the 7135's seems to ignore the tiny bit of on time they get. Once things cool down you can watch the LEDs turn back on to moon levels. It is so low, I cannot even measure it.
Measurements
Dimensions:
- Overall Length: 112.1mm
- Bezel Diameter: 69.4mm
- Head diameter: 57.0mm
- Body Diameter: 44.4mm
- Longest supported cells: ~67mm
Weights (without batteries):
- Overall: 574g
- Head: 400g
- Body Tube: 174g
- With cells: approx. 710g
Performance (stock, 3x18650 @ 4.2V supply, uncalibrated measurement equipment):
- Light Output: ~2611 lumens at start, ~2588 after 30s
- Beam Intensity: ~27.3kcd
- Med: ~711 lumens
- Low: ~45 lumens
- Moon: <1 lumen (could not measure)
Power Source Options: 3x18650, unprotected only
Switch type: Twisty
Modes: High(~9.0A), Med(~2.7A), Low(~0.27A), Moon (0.004A)
Mode Memory: Yes
Tools Used
- Dremel, with cut-off wheels, grinding disks, sanding drums, sanding pads, drill bits, buffing wheels, wire wheels.
- Cordless Drill
- 5mm round file
- 4lb Maul
- 8oz Hammer
- Block of hard wood
- Pliers
- Locking pliers
- Needle-nose pliers
- Side cutters
- Digital calipers
- Sandpaper, 400, 800, 1500, 2000 grits
- Soldering iron
- Hot air reflow (for emitters, and in lieu of blowtorch)
Closing Comments
I really enjoyed this project. I want to thank Old-Lumens for organizing this, the judges for their time, everyone who is participating, and everyone watching too. I also want to thank our sponsors (to date, Illumination Supply and Lightscastle) for their support and prize donations. Maybe next year some sponsors can help out our members by sponsoring them with some DIY parts.
This has been a blast. I'm already thinking of some ideas for next year.
Thanks for reading! SearchID8935