Thanks! I quite like how it’s turned out. Here’s some more pics:
On the left, a picture of the top of the bezel with aspheric and AR coated glass lens on top of the aspheric. Bezel fully retracted. To the right is the same thing with the bezel partiall extended so you can see a bit of the magnified die. The fresnel on the bottom is glued in with Norland NOA61. The glass lens is press-fitted with a bit of Kapton tape at the (that’s the yellow visible at the edges of the lens). Enough of a press fit that the glass shouldn’t come out, but if I decide to do without the glass, or if I need to replace it, I can pry it out with small hook. I’m still deciding whether I want to glue the glass lens in permanently with Norland, do without, or leave as-is with the Kapton tape.
Here’s a pic of the top of the light with the lens off and the bezel fully retracted. Note how few threads are visible above the top of the pill and how thin the lens is. The optic consisting of lens holder, fresnel lens and glass AR lens is only about 3.5mm thick. Getting the lens to sit as close to the LED as possible in flood mode is the key to getting a super-wide mule-like flood … and it also results in a more compact light. In this light, I have the lens siting so close to the top of the emitter, that if the emitter weren’t dedomed, the dome would be sitting inside the back of the lens.
I don’t understand why you get so much improvement with the new lens. The throw ratio should be about 20*2 / 18*2, which is only around 23%. The shorter focal length might decrease throw if it is really really short but should not increase it. The gain should be in the size of the spot.
I should look again at Edmund as a source. I did actually pay around that much for one aspheric. I have lenses left from when I was a kid, from them.
Really nice work. I am glad you knurled the bezel and tail. It fixes the cheap mass produced look of this type of light, as well as giving better grip. A bright pocket zoom 18650 is really an accomplishment, and you also have lots of throw and unusual flood.
I suspect the large difference has more to do with my crappy skills using the luxmeter than anything else. Other times I repeated that test I got anywhere from 40k to 43k lux. I was only able to get 48k once. This is a new luxmeter and until last week I’ve never owned one before.
There’s also some possibility that the quality of the plastic in the 20mm lens is just better than what’s in the stock lens. That might account for some of the unexpected difference.
The Edmunds lens is way better than the Fresnel Factory lens.
With the fresnel factory lens there are distinct white lines separating each ring. The white part constitutes dead area where the light scatters instead of being focused. With the Edmunds lens, the rings are finer and there’s no visible white lines between them. It seems to transmit much more light and produces a much sharper image of the die.
Whew. Got some additional work done on this light:
1. Replacement bezel installed. Using a thin coated glass lens attached with Norland NOA61 and then Edmunds fresnel press-fit behind. It works well. This time I didn’t sand off quite as much of the bezel so there is a bit more rim protecting the lens.
2. Swapped in different switchboot. I chose one that projects up a bit more. Still stiff enough that accidental pocket turn-on is highly unlikely, but makes finding the button much easier. I originally designed this light to use button boots taken from a cheapo plastic duracell flashlight. But I found that I get a more comfortable and grippier button boot just by cutting down a standard 14 or 16mm silicone button boot.
3. Added ribbed rubber sleeve around the body taken from a cheapie Nebo flashlight.
4. Filed out the inside of the tailcap. One of the problems with the initial mod was that the tailcap had the male threads. Previously, there had to be a space behind the battery for the threads to mate. Now with the tailcap inside the male threads filed out, the battery goes right inside between the threads and gets much closer to the back of the light. This let me file off a considerable chunk of the back of the body. Result is the entire fully assembled light is now just 91mm long…… shorter than a Sipik 68!
My most compact light is an AkoRay AK-12. It is a usual reflector light, except that everything is optimized for small size: square threads, intricate knurling that maintains crush resistance, metal thinner where it is supported inside. I was not able to find a second one; maybe other collectors snapped them up.
!!! This is insane! With a side switch! So awesome. The main reasons for my interest in led flashlights to begin with was robustness and small size. You have just raised the bar considerably! Thank you for showing all this. I know it must take a while to do the writing and posting too. TY
I had a lot of fun shrinking this light down. Became a game to look at the light and see what I could remove or change to shrink it, without compromising performance.
At this point there’s not a lot left to remove:
Current length: 91mm.
Places which could be modified to further shorten the light:
1 mm - The lens is recessed 2mm behind the front of the bezel. Reducing this to 1mm would shorten the light by that much. Downside is less protection for the lens.
1/2 mm - Filing down the back of the tailcap. To make the tailcap I sawed off the back near the base of the switchboot and then filed it down. I kept the rim that held that switch boot and used it to hold an aluminum disk that forms the back center of the light. I could file that rim thinner. Downside is there is a slight divot on the inside of the ring caused by me being overzealous with the electric drill when I removed the switch. Filing it down further would shorten the light slightly, but would reveal this divot and could make the tailcap look worse.
1.5 - 2mm - shortening the pill by removing the copper heatsink. The stock light came with a hollow pill. I installed a copper baseplate and heatsink composed of 1 larger copper disk and 3 smaller copper disks all soldered together. Removing the 3 smaller copper disks would open up some space and possibly allow me to move the driver up. The downside is this means a lot less heatsinking under the emitter so isn’t a good idea.
That’s about it. There’s pretty much nothing else I could do to shorten it further. I think rather than shortening it further future modifications to this light will be mostly cosmetic. I’m thinking the following:
1. file channels into the fluted portion of the body to match the rubber sleeve. The purpose would be so the rubber sleeve recesses at the sides and can’t slide out of position. I’m undecided if I want to try this. It looks and works pretty well without them. And if I do this I’d no longer have the option of using the light without the sleeve.
2. file and drill a lanyard hole at the edge of the tailcap. Not sure I want to try this though. First, I never use lanyards. Second, there may not be quite enough material left in the tailcap to do this. There’s a risk I could damage the aesthetics of the tailcap if I try and fail.
3. LED replacement. I’m waiting for arrival of a shipment from International Outdoors of different XML2 emitters. Hopefully these will give better tint when dedomed.
4. Duracoat. I’m considering duracoating this light. I’ve never used the stuff before, but have read up on it. Looks like it would work. Would be a pain setting the light aside and not touching it for a month though.
The instructions suggested using Trustrip degreaser to clean the metal first. I sprayed it on and then tested it with a drop of water. I noticed it still beaded up. Perhaps the trustrip failed to remove the coating from the metal polish I’d used when I first finished the light.
I then soaked the bezel and tailcap in Greased Lightning for 20 minutes. I scrubbed the body down with it too, but because I didn’t want to redo the switch, I couldn’t immerse it. After the soak, I then washed the pieces in hot water and scrubbed them with liquid handsoap (the clear kind without any moisturizer in it) using a toothbrush. Then I rinsed them again.
After this second cleaning, testing showed water no longer beaded up. Hopefully my cleaning didn’t leave any other kind of reside that will interfere with the Duracoat.
I then sanded everything down with 600 grit sandpaper, masked off the switch button and inside of the telescoping portion then sprayed on the Duracoat. I chose Blue Titanium color, and the Duracoat was in an aerosol can so no airbrush was needed.
I have to wait till tomorrow before I can touch it. And then wait another 4 weeks for it to cure enough for pocket EDC use. Hope I did it right! Next month, I’ll get to reassemble the light and see if it still works and the paint doesn’t chip.
It looks a bit glossy. More like paint than like anodizing. I was hoping for a flatter look. The instructions said the sheen should fade somewhat as it cures over the next 6 weeks. Hopefully, this will be the case.
I used the “can within a can” aerosol Duracoat. After releasing the hardener, it’s supposed to stay good for 48 hours. I found this not to be the case.
Day 1: I released the hardener and sprayed it on: Looks glossy.
Day 2: I decided to add an additional coat to part of the body. Also since the can wasn’t empty I decided to coat a second flashlight - a modified Sipik 58 with triple Nichia 219B. The duracoat was still good. However, surprisingly this second-day spraying did not give a a shiny coat. The second flashlight and repainted part of the first light look more matte. The matte looks quite a bit better actually.
Day 3: A few hours after spraying the second flashlight I decided to recoat more parts of the first light in the hopes of getting more matte finish. Unfortunately, the spray that came out of the nozzle was spotty… no longer a mist. I aborted the attempt before spraying the light. Several hours after that, I noticed that the mixer balls no longer rattled…. the can was officially dead within 30 hours or so of releasing the hardener.
I cured all parts in the oven at 110 degrees F. That’s supposed to speed up the initial overnight cure, but supposedly they’ll still need 3-6 weeks to fully cure. Until then I don’t plan to EDC either of these lights. The finish on the Sipik 58 seems quite strong, but I’ve already rubbed off a small portion of the finish on the bezel of the Aleto. Hopefully it will get stronger with more cure time.
I’ve now EDC’d this light for a month in my pocket with my keys with the duracoat added.
I’m not impressed with the duracoat. It’s definiitely chipping off. In both durability and looks this finish is far worse than anodizing. Even type II anodizing is better. Of course the problem might be my surface preparation rather than the paint, but this still doesn’t seem that great. On the upside even with worn paint, I think it still looks better than bare aluminum.
The final light is 91mm long. It’s shorter and thinner than a Sipik 68.
I’ve done a good amount of durakote, last year I purchased a lot of 1930’s era mossberg SA-22’s and Some bolt .22’s that were all in really bad condition and wanting working guns, not show pieces I ended up durakoting them. Because they’re old enough to not have serial numbers (totally legal to buy / own) I had issues mailing them to my usual cerakote place since mailing firearms requires a SN so I elected to DIY with the durakote. I’ve also used the stuff on a few non-serialized AR lower receivers I’ve done as 30% and 80% builds.
Its not as good as the real stuff (cerakote) but if done right it holds up decent. The number one thing is surface prep followed by application technique. One trick I was taught, heat the parts to 180*F before spraying, that way the second the spray hits it it will dry (not cure, just dry) so there are no runs/streaks to bubble up and chip later. I also bake the stuff after the final coat as well to cure it then I wait 7 days before handling the parts without gloves (to keep oil’s off). The longer you can leave it set before use the better it holds up.
When I duracoated the light, I put gaffer’s tape over the switch boot. The result was that the paint was a little uneven around right around the switch. The button was also a little hard to see because it didn’t have the silver ring around it.
I ended up removing the switch cover and adding the silver ring. The ring is made from a strip of aluminum glued to the edge with super-glue gel. I then filed it thinner with a handfile and cut down the boot so it would fit the slightly smaller opening.
Then I needed to touchup the area around the switch button … and I didn’t want to buy more expensive duracoat and wait another month to do it.
I took some acrylic hobby paint intended for model miniatures and mixed it up to get the correct shade of grey. Then I painted it on around the metal switch cover with a hobby brush. Being water-based acrylic, the stuff dried in minutes with no odor.
Of course acrylic hobby paint isn’t durable enough for a high traffic area that gets constant wear from my thumb and pocket. So after the paint dried I used a qtip to brush super-glue gel on top of the touched up area. Then I let that dry.
This technique worked surprisingly well. The area that I touched up with the hobby paint and superglue hasn’t worn at all and it looks indistinguishable from the duracoated section.
Pulled the light out of my pocket yesterday and it didn’t work. This surprised me because it had been very reliable.
I did a bit of checking and noted:
My Panasonic NCR 18650b 3400 mAh inside was completely dead. It read zero volts…. not good for an unprotected cell. My i4 charger wouldn’t even attempt to charge it when I briefly put it in. I discarded this cell.
The fresnel lens had popped out and was loose inside the head.
I built the light with 2 lenses. The outer lens is AR coated glass and is held on with Norland N0A61. It’s there to provide an easy to clean external lens and protect the fresnel. Immediately behind it was the PMMA fresnel lens, held in place with a relatively tight press-fit. When I unscrewed the bezel I noticed there was a considerable gap on all sides of the fresnel. It was no longer possible to press-fit it without a spacer. Either the aluminum bezel had expanded (unlikely consider the outer lens and Norland glue was intact) or the fresnel had shrunk maybe 1/4 to 1/2 mm on all sides.
I didn’t notice any heat in my pocket, but my guess is the light accidentally turned on and ran the battery down. The heat from the light running at max power maybe caused the PMMA fresnel lens to shrink slightly and pop out of its press-fit mounting. This is the only thing I can think of that explains both the dead battery and shrunken lens.
The focal point of the fresnel was also slightly off. Fortunately, this light is still fairly easy to mod. I had an aluminum C-ring spacer inside the sliding mechanism to provide focus. A bit of trial and error later and I made a slightly thicker one to restore the proper focus in spot mode. A bit of kapton tape around the edges of the fresnel allowed it to press-fit again.
I’ve decided to only use protected cells in this light since it seems the driver maybe doesn’t have sufficient low voltage protection.
I’m also in the process of testing dedomed XPLs in this light. I ordered 3 different cool white tints from Illumination Supply. Unfortunately, they sent them to me unlabeled and all in the same box, so it’s impossible to tell which tint is which emitter. I’ll just have to try them all.
I wasn’t too impressed with the first dedomed XPL I tried. The tint was ok, but still quite greenish, and it felt inferior to the dedomed 3A tint XML2. Even worse, lux went down from 27k to 21k, suggesting that it wasn’t performing as well as the XML2.
2 more XPLs to try. Second one should be ready for dedoming and installing tomorrow.
This is insane! With a side switch! So awesome. The main reasons for my interest in led flashlights to begin with was robustness and small size. You have just raised the bar considerably! Thank you for showing all this. I know it must take a while to do the writing and posting too. TY 
