The flashlight that never was... (pics, poll)

Funny how getting a good deal on some extra batteries on eBay turned into building another expensive flashlight I didn't need. Introducing the flashlight that never was... the XVII-Ray (17 emitters).

I thought adding a center CREE XRE (which was to be warm white) instead of a fifth MC-E might give a better hotspot that fills in the bright ring generated by the MC-E emitters in these reflectors.


You can tell by the tint of the center die that I temporarily installed a cool white emitter. I'm still waiting for the warm white ones I ordered from KD to arrive.




Here's the wiring inside the head. MC-Es are 2p2s relative to each other, their dies altogether are 8p2s. Total MC-E power is 7.2Vf @ 5.6A. The XRE is wired separately, run by a single emitter buck driver.




Here you can see that driver fixed to the inside of the base with AAA. The hot connection is made through a banana plug. Metal-to-metal contact between the base and the neck make the ground connection.




Some serious honing had to be done to the tube to fit the Sony 26650VT batteries, which are as fat as AW's IMR 26500's. The tailspring had to be stretched to make contact. (I have no tools for cut and re-thread.)




Here's the guts of it. Those Sony 26650VT Lithium Nickel Cobalt Manganese, safe chemistry, 50A capable, 3.6V batteries that I got for a song are what inspired this and two other Mag 3C projects. They simply do not sag under this 6A load. After a very short use with no resistance, the neutral white MC-E emitters began to turn blue. Not having any of my favorite Radio Shack .47 ohm 5W resistors on hand, I found some NTC soft-start thermistors I had left over from my old hotwire days. Viola! It started out dim and as the thermistors warmed up it got brighter.

Time to take it outside...

CRAP!




Funny thing about going outside with a torch 5 minutes off the production line is that outside it's much harder to tell that you've made a grave error until it's too late and it self-destructs.

Here's what I think happened... after running for a minute or so, with no detectable color shift of the LEDs (at least not detectable outside in the dark), in a brief instant, the NTCs got so hot, their resistance dropped to nil. Without warning the MC-Es were baked. These were neutral white MC-Es too, harder to get and more expensive.

I'm a dumb*ss for taking a shortcut at the last minute and using the NTCs instead of the usual resistors that have served me well on other similar projects.



Desoldering the leads and testing with a multimeter showed that only the first MC-E in each pair was damaged. The last two at the ground connection survived.



Some careful lateral pressure and I was able to snap that micro thin coat of Arctic Silver Thermal Adhesive without damaging the good MC-Es.

I've run that lone XRE by itself and was surprised to discover that the hotspot from that center position is not that tight at all. In fact it's about twice the diameter of the MC-E hotspots in the outer positions.

So now what? I've got the two good MC-Es I salvaged from this wreck, and three more left besides those, but I had plans for those individually, or possibly three together if I found the right three-flector.

I've also already made two other 5X neutral white MC-E torches, The XX-Ray and the LED USL. I've also have another Mag 3C and 5-way reflector+base module like this prepped to receive 5 warm white CREE XRE emitters and a DX 5*CREE 5-mode boost driver, so...

What do you think I should do with this host? Single mode 5x WW XRE comes to mind. So does center MC-E surrounded by 4x XRE. Not sure what else might make sense...

a center MC-E surrounded by 4 neutral white XRE sounds good

that's some cool work you do, too bad the first try died. :/

+1 on both counts.

The center reflector isn't shaped to take advantage of the XR-E, put the MC-E there. Perhaps 4 K2 TFFC's would do well for you in the other positions. I think they would pair really well in there.

It occurs to me that if I go with the popular choice, one center MCE surrounded by four XRE, I'll still have enough MCEs left to also do one of the other options, build another tri-MCE. Now which reflector for the tri-MCE...?

Techjunkie said:

Time to take it outside...

CRAP!

Click to expand...

Maybe you should rename it the Titanic. I think you had a brilliant idea with trying to fill in the gap left by the MC-Es. So much so that I will donate 2 neutral MC-Es so that you can bring it to life. The idea was great but the execution needs work obviously. Although not symmetrical the one XR-E should be in one of the larger reflectors with the other spaces filled in by the MC-Es. This should accomplish your idea much better. I'd love to see the results of this setup so if you want to pay the shipping I will send the LEDs to you.

saabluster said:

Maybe you should rename it the Titanic. I think you had a brilliant idea with trying to fill in the gap left by the MC-Es. So much so that I will donate 2 neutral MC-Es so that you can bring it to life. The idea was great but the execution needs work obviously. Although not symmetrical the one XR-E should be in one of the larger reflectors with the other spaces filled in by the MC-Es. This should accomplish your idea much better. I'd love to see the results of this setup so if you want to pay the shipping I will send the LEDs to you.

Click to expand...

Generous offer.
Interesting idea with the placement of the XR-E.

saabluster said:

Maybe you should rename it the Titanic. I think you had a brilliant idea with trying to fill in the gap left by the MC-Es. So much so that I will donate 2 neutral MC-Es so that you can bring it to life. The idea was great but the execution needs work obviously. Although not symmetrical the one XR-E should be in one of the larger reflectors with the other spaces filled in by the MC-Es. This should accomplish your idea much better. I'd love to see the results of this setup so if you want to pay the shipping I will send the LEDs to you.

Click to expand...

Wow, that's a very generous offer. Thank you! I can't accept it though. After experimenting with my 5x neutral white MC-E torches and a single warm white XRE torch with an 18-20mm OP reflector and 1000mA driver, I've come to the conclusion that even in one of the corner spots, a single XRE isn't bright enough to stand out against multiple MC-Es. It will take multiple XREs in reflectors this small to generate a hotspot strong enough to properly fill in the depression in the MC-E ring shaped hotspot.

For this particular host, I've decided to reverse the original design and go with 4XRE and an MCE in the center. Instead of resistors, I'll use two 5-8.4Vin 3A buck drivers. One will power the 4 XRE in parallel and the other will power the center XRE (4p). I've trimmed the edge of one to fit against the underside of the base, the other will go in the pill spot and they'll be connected by top-side traces to the Vin and Grnd, so it will look nice and clean on the outside of the light engine.

I'll have switch the modules between this project and the 5XRE project (which, luckily, I haven't assembled yet) because I can't pry the PCB from the single XRE driver off of the bottom of this reflector (too much AAA). Luckily, the 5XRE driver in the pill ring will clear the leftover (scraped clean) PCB, so I'm just swapping reflectors modules evenly.

It is my hope that this will turn out to be a nicely regulated torch with a strong hotspot and lots of bright flood, without use of a giant head.

As for the idea of filling in the donut of multiple MCEs in small reflectors with multiple XREs in small reflectors, I'd love to see someone use 7 equal sized reflectors and alternate XRE/MCE around the outside and place one XRE in the middle, for a 3MCE:4XRE ratio. Finding the right driver and battery combination could present a challenge. I suppose they could wire up all 16 dies as 4s4p and run the whole chain off of a 3A buck driver and 6 Li batteries in series. I think something like that would VERY nicely achieve what I set out to do here originally, but at a much greater cost for the host, heatsink and reflectors. I had considered converting this ultrafire concoction but the tailswitch is a twisty and the config I described would have forced me to use 6 IMR 16340 batteries which would have provided crappy runtime.

Reversed the emitter config, added regulation, dubbed it The Mark VIII.

Will likely make another for lack fo better ideas and better parts.