I modded this light in 2010 and it still works today and the epoxy has never separated from the heat sink.
I would use an XPG2 from Illumination supply in 5000k which is a nice tint.
Need any other help with this you can ask me here or PM me. Good Luck!
Here is a part list to get you started. The glass lens is a couple of MM small, but the o-ring can be used to center it.
Glass lens
Reflector for XPG
Good driver for single mode with 2 batteries
Multimode driver for 2 D setup
You can build your own light for less than that and it will be much better! Heck, I could build out someones anglehead for less than that.
It looks to be genuine.
Things to look for would be rivets on the switch and clip, vs screws on a knockoff. The name fulton industries MX-991u is usually in a raised letters in the circle on the top head of the angle.
Thanks for this
No problem! Post your progress.
Thank you, Nitroz!
Old type 2D cell lights are the original flashlight configuration and are still available for well under $5. D cells are a really good source of power if you don’t mind the size and weight. The problem is that they have incandescent bulbs, and therefore no attempt at cooling. A low power led replacement bulb gives a light with not much output but a very long battery life. To get something like a 1 to 3 amp led to work in one, I am thinking of filling one with oil or alcohol or glow model airplane fuel. The circulating, and possibly boiling, liquid would take the place of the missing aluminum for cooling. Two alkaline D cells supply the current and nearly the voltage of an 18650 and several times the life. Water would almost work but it hydrolyses at around 3 volts and promotes electrochemical reactions of the metal parts.
One of the incandescent bulbs I have says 0.8 amp, 2.8 volt, so this is what was expected of 2D cells a couple of decades back. It would be better not to use the conduction strip and spring, as these are made of steel. I suppose I can kludge waterproof seals with epoxy and silicone rubber cement.
I have a spare 18650 XM-L light I can use parts from. I think I can leave the pill alone and just epoxy it to the truncated reflector. The throw should come out better than my small 18650s if I get the focus right.
But wouldn’t heating a volatile substance like oil or alcohol or especially model airplane fuel in an enclosed space create the equivalent of a small bomb? The tint should be REALLY interesting!
That might come under the heading of “It seemed like a good idea at the time!”
Just be careful, and be safe. 
There isn’t that much heat, and these liquids are not like gasoline or acetone. Oil is what has actually been used in led bulbs. The alcohol and model airplane fuel, which is mostly alcohol, are more of a question, but they have lower viscosity and maybe more heat capacity. It is hard to see how the liquid could be ignited, unless the battery is lithium and is crushed and ignites, which would already be a big problem. My main worry would be leaking. That would be messy and could create a fire hazard.
I tested 70% isopropyl alcohol for insulation. I got 0.1 mA at 4.1 V with some salt sprinkled in and no measurable current without it. On the ohm meter, I got 100 K ohm with salt and 1 M ohm without it. I don’t trust the ohm meter for this because I don’t think it uses realistic voltages on these scales. So 70% isopropyl is marginal, after it picks up ions from other components, but 90% should do. Methyl alcohol is more polar and might hold more ions, but I could get it dryer and it might cool better.
Alcohols boil at less than 100 C, which is in the operating range for leds, so we might get streams of bubbles like in a Cray II. This would work even better if the led were de-domed, but I don’t plan to go that far. I am not that big on military angled flashlights and hope to try it with a regular $3 2D cheepo. The Crays used freon. It may have better thermal properties, but probably the main reason is that they consumed a lot of power, so flammable liquids would be a fire hazard. An led will burn out long before it gets hot enough to start a fire. If one used Li batteries and they burned up, the liquid might add to the fire hazard, or if it spilled in the presence of a flame.
I’m sure the very similar, brit version was made by Crompton, it didn’t have the switch guard though, and had a set of red/green lenses in the bottom as well as a spare bulb, I may still have one somewhere.
tabs
Petrol doesn’t explode, just burns, and only when the ratio is 9-18% meth/eth/prop/but/pent/hex/hept/oct/non-ane and the rest air. Oh god my organic chemistry is coming back to me now…
Anyway if the pressure builds until it leaks, then it is possible it will burn if there is a source of ignition such as a spark or an open flame, or it can auto ignite if it’s like 600 degrees celcius+.
It can’t get that hot. The only danger I can think of is that the case leaks or brakes from handling or dropping. Maybe that would argue for using a military version that is tough. At least it is no more dangerous than when they used to store alcohol in glass bottles.
Added: The thermal situation will be that it is about 85 degrees C around the led, what XM-Ls are rated at I think, and close to the boiling point of alcohol. Whether or not it boils, the alcohol will expand with the temperature enough to circulate and spread the heat at least around the plastic reflector and window (lens) and to some extent around the whole flashlight. The plastic case is very large area and thin, so the heat is conducted to the outside with good efficiency. It is then carried away, just like it is with an Al light. Clearly it won’t boil enough to increase the pressure significantly. The case will expand at a similar rate to the alcohol, and the case is flexible, so the pressure won’t vary much.
If it boils, it will be a tiny stream of tiny bubbles as with the Cray II I saw.
For my 4 D cell light with good Rayovac alkaline cells I get 5.25 A across the battery pack with a meter that I think has around half an ohm in the leads. The cells are a little over 1.5 V each. So it would take about 4 cells to drive an XM-L at a little under its rated 3 A. That is half of the 5 or 6 amps and half of the 6 volts. That would be direct drive, but a driver could not improve on it much, because that is about their maximum power output. A 2 D cell light needs a driver to drive the XM-L at half power, with alkaline cells.
I plan to start with the parts from the light I am cannibalizing, but may try these if all else works.
That leaves out the question of more conducive glue.
The army surplus store has Fulton angle heads for $30 or 40. At the drugstore I got two Eveready lights for $3.60. These are identical to my old orange one except already converted to led with a drop in bulb and a step in each of the 5 internal ribs to hold only one D cell. Each came with one Eveready Super Heavy Duty dry cell. The spring and side current path are steel. The part that moves as the switch appears to be copper plated. It contacts a steel plate that holds the bulb.
Clearly I can’t use any electrical part, but the body and bezel seem sturdy and the reflector looks good and throws a tight spot almost as small as my old one does. I am testing what adhesives will stick to the body.
After thinking, I came up with a better name for this entry.
One of my first custom builds, I used 32 5mm LED’s wired in series
Direct drive off D alkalines, I ran it non stop for 230 hours when I screwed up and Turned off my phone (that I was timing it with). It never dimmed, I stopped the test and put in a new set of batteries and to my eyes it looked exactly the same brightness as when I stopped the test, I would put money in it lasting >500 hours on a fresh pair of alkalines! I don’t use it, ever, but its in one of my vehicles somewhere, I think maybe the horse trailer, the only thing that worries me now that I think about it is the alkalines leaking, maybe time to upgrade to some LSD NiMH D’s or even just L91’s in adapters.
Since I am not using an angle head, I started a new thread for my liquid cooling experiment, Liquid cooled XM-L Eveready 2D sleeper, here Liquid cooled XM-L Eveready 2D sleeper .
Don’t you mean parallel? That would take a lot of D cells.
I think isopropyl alcohol lights with a match because the heat vaporizes it but its flash point is higher than room temperature. The ignition temperature and heat produced are probably between those of gasoline and methyl alcohol.
The led is nowhere near the ignition temperature nor can I imagine any way to light alcohol with a malfunction of the 3 volt electric circuit, so there would have to be an outside source of ignition for it to burn.
Anyway, there are only a couple of ounces of it so there is no danger if I see it. Making it safe for mass market is a different question but will be solved if liquid cooling turns out to be useful.
Added: It appears that the diode itself may fail by becoming more conducting. The runaway current is limited by the copper leads melting. Their melting point is above the ignition temperature, but like 3 volt inductive sparks their size is too small to cause sustained combustion.