We went to an estate sale today to see how a company my wife’s family is considering using does their job and the first thing I bump into as they route us through the garage is a crate of old flashlights. Almost all plastic junk, but covered up at the bottom of the crate is a 3D Maglite in very good condition with dead batteries and/or a dead bulb or who knows what (no MM right now). I’d have ignored the Mag too if I hadn’t see some of Justin’s handiwork (not to mention the familiar feel of the light in my hand). First thing I start thinking of is how nice it will look with a polished bezel and tailcap and something better than the stock incan bulb. Maybe a triple led, and . . . . .
Anybody think it would have gone back into the basket without the weight of the batteries? So now I have an old dead tube for a starter project and another money hole. :bigsmile:
Well, it's got to be someone's fault. Hey, I was looking on ebay and saw some 50mm triple TIR optics. Just perfect for a triple XM-L.
I haven't been buying any 2D lately. I have been working on a deal for 20some leds for the winter. Anyhow, Wal-Mart has restocked a bunch of the old incans again and fixed their scan code errors. I just can't keep up with those guys.
Besides, there's a lot worse I have been blamed for than flashlights.
I would see if someone local would cut a 1/8” aluminum plate and get a Maglite aluminum heatsink, or again, have a local person cut you one. Put the two together to make a stepped heat sink and put it in the maglite. Then use XM-L leds on 16mm thin style stars and place them on the top. The TIR optic goes on top of that. When you get to where you put the leds in place, use AA on the stars and set them down (approximately) and use the TIR to center them, till the AA hardens. I am going to guess that the TIR has about a 20-22mm center to center on the three leds.
Here’s a simple and not to scale drawing, for an idea.
I think the TIR is going to be about 15mm tall or more. If it’s 15mm tall, it will be OK, if it’s taller, then it might be an issue. You might want to ask for the specs.
I like where you guys are going with this but as a large optics manufacturer. I want to be able to really focus this beam of light and throw it way the heck out there. I also have unrestricted access to a 3 axis CNC router and have about 4 dead 3D mag lights laying around. I’m not much of an electrician but I’ve done some small scale LED projects that were a flop. Not because of the LEDs they worked great but the inital design didn’t work as well as I had hoped lol
How would you drive a 3 light aray like this and what would you do for power/batteries?
Well, there’s the catch. Thrower or flood. That’s always the choice to make. Typically a thrower is done with a single LED using a deep reflector, or with an aspheric lens, rather than TIR optics like this. These multiple LED lights, usually do not “throw”. They may seem to reach out, only because they produce massive amounts of light, but they are not intended to be throwers. Most long distance throwers I have seen, are done with an aspheric lens and focused to a “square die”.
3 LEDs like XM-Ls? Usually the general rule is One Li-ion per LED or Three NiMH per LED, but that would be using “direct drive”, (no driver board). Using a driver board can change the battery configuration, depending on the type of board.
In this particular set-up, the usual would be to either use Three Li-ion batteries in series, such as 26650 batteries, or go with Nine NiMH batteries (AA NiMHs - eneloops). 9 in series, to get the proper voltage.
I would drive it with Three of the 8x7135 NANJG drivers, with one master and 2 slave, but there are drivers out there that will drive all three LEDs with one driver. Series is always the best way to drive LEDs. Parallel will work, but there are more possibilities of damage to LEDs when running parallel. Series multiplies the voltage needs, parallel multiplies the amperage needs.
If you use the 1 master-2 slave approach, it requires the current draw of a parallel set up but each board controls 1 led so is not subject to the problems of a completely shared parallel light. I am working on an approach to allow series LEDs to be driven with a 7135 driver from series batteries(a bit different from CPF Downloads arrangement). It works but a different switch and a resistor is needed to both control the initial battery voltage (resistor)and eliminate the resistor as Vbatt drops(different switch) along with a voltage regulator chip for the processor. If you are curious, you can read about it here
It’s still a work in progress but shows promise.
If you use the 1 master-2 slave approach, it requires the current draw of a parallel set up but each board controls 1 led so is not subject to the problems of a completely shared parallel light. I am working on an approach to allow series LEDs to be driven with a 7135 driver from series batteries(a bit different from CPF Downloads arrangement). It works but a different switch and a resistor is needed to both control the initial battery voltage (resistor)and eliminate the resistor as Vbatt drops(different switch) along with a voltage regulator chip for the processor. If you are curious, you can read about it here
It’s still a work in progress but shows promise.
