The Infamous XX-Ray: 5x neutral white MC-E (pic heavy)

Techjunkie

Enlightened
Joined
Nov 16, 2007
Messages
943
Location
in the brightly lit suburbs of NYC (Long Island)
Introducing... The XX-Ray

So-named for its 20 LED emitter dies (XX = Roman numeral twenty)

----====::::EDIT #3::::====---- (June 27, 2009)

The replacement Sharks from Wayne came in today (Thanks, Wayne!). These make Sharks 7, 8 & 9 to enter this torch.
Well, they say the third time's a charm (knock wood), and now finally, three matched, tested blue sharks running in parallel are running this torch at just the right current.

Sharks789.jpg


Therightcurrent.jpg


---===:::Edit of the Edit:::===--- (June 5, 2009)
The increased current as measured below was due to one of the sharks malfunctioning. Tested alone, it was pushing 1.5A instead of 1A. I thought it would continue to only push that much more and was expecing 3.5A total. The total 4.89A current was measured after it and the other two were permanently affixed to the sinks with AAA. Unfortunately, after an "extended" run of about three minutes constant, the one bad shark took the other two down with it and all three fried. Lesson learned: don't fully assemble if ANYTHING is out of spec.

The XX-Ray is "in the shop" again, awaiting replacement sharks and now I also have to replace one of the MC-E emitters as I damaged the dome while unassembling/reassembling to get to the 2nd course of toasted shark. The replacement will be a 4A tint - I couldn't find any more 4B.

-=:Edit:=-
I replaced the sharks I torched by reversing polarity. For reasons unbeknown to me, they crank out more than 3A combined. I'm not complaining though, 'cause the LEDs are liking the extra juice. Here'sa quick photo insert and then some new beamshot photos down around post #20.


(Photos removed, see update above)


-=:End Edit:=-

---===:::End Edit of the Edit:::===---

----====::::END EDIT #3::::====----



Specs:
  • 5 x CREE MC-E neutral white 4B tint emitters
  • Solid Aluminum reflector and base (both modifid for fit and focus)
  • Silver Mag 2.25D Dual and Quad bored, cut and rethreaded, head modded inside for fit
  • Carrier-less battery design will accommodate 4 IMR or LiFePO4 18650 batteries or a 12AA NiMH pack
  • 3 Blue Shark drivers (remora removed from original design)
  • Emitters wired 5s4p with stranded 20ga copper
  • Estimated > 3000 Lumens of neutral white LED light!
Currently test running in direct drive mode until Remora issue is decided. (Measured only 840mA from Blue Shark with Remora on High vs. 970mA without Remora). Once that's settled, the 3 Blue Sharks will be mounted on Led Zep shark sinks in the neck. Head assembly to tri-sharks and tri-sharks to Mag switch might remain fitted with detachable 30A connectors for maintenance or future modification, provided there's room after the Sharks are installed.

This one's a bit of a sleeper, considering the stock Mag head and short body. It started out on a budget too, the only exravagence being the modded body in favor of a 4C or 3D, until the need for 3 sharks became apparent.

After modding the distance between the base end the reflector bottom, focus gives a much tighter hotspot than the 20mm narrow boom reflcetors I've used on one of my 3xMC-E torches. There is a bit of a small donut in the hotspot, thanks to the smooth reflectors, but the rotation of each emitter purposely varies to avoid lining them up and eliminate plus sign in beam pattern.

Some pics:

TheXX-Ray.jpg


TheXX-Rayopen.jpg


BusinessEnd.jpg


Tailstand.jpg


Emittermounting.jpg


unassembledpre-sharkdisaster.jpg


I got tired of searching for a 2x2x18650 battery carrier and building my own with only 16mm to spare was a daunting task, so I took an old shortcut. The nylon washer insert takes advantage of the fact that the milled out battery channel is conductive by mating the negative battery termianal to the channel (and the positive terminal to the spring. At the tailcap, another custom fit washer is used to make a contact board to bridge the two stacks of batteries. Yet another insulates it from the low pressure modded tail spring. No electrical contact is made to the spring when running 4 x 18650 in series this way. The Sharks which have to push 18V @ 3A to the emitters can accept either 3.7V IMR batteries or 3.0-3.2V LiFePO4 batteries. Take the washers out and there's just enough room for a 4x3 12AA Eneloop pack (if I could ever find one).

Carrier-lesssolution.jpg


4x18650nocarrier.jpg


Tailend.jpg


As it will be a while before the sharks are added, I've added some preliminary beam shots in post #17, with the XX-Ray underdriven at ~1.6A or about half power. (Or just checkout the slideshow on photobucket.)
 
Last edited:
Re: Introducing... The XX-Ray (pic heavy)

woooooow, a dream comes true :grin2:
 
Re: Introducing... The XX-Ray (pic heavy)

hi,

really nice build. i also test a mc-e setup with those dx reflectors, but it produce a big donut hole or you have a bad efficiency when you move the emitters a little outside the focus point. also the thermal management of this very thin alu heatsink is miserable so a build my own heatsink that will transfer the massive heat wall direktly to the mag body and head:
5q6c-a.jpg


and here is the thread to this flashlight:
http://www.candlepowerforums.com/vb/showthread.php?t=225234

creative battery holder, i will take this idea in back of my mind for my future mods :twothumbs

markus
 
Re: Introducing... The XX-Ray (pic heavy)

wow:faint:

I want one!
 
Re: Introducing... The XX-Ray (pic heavy)

Holy C#@*, 5 MC-Es. Outdoor beam shots please.
 
Re: Introducing... The XX-Ray (pic heavy)

hi,

really nice build. i also test a mc-e setup with those dx reflectors, but it produce a big donut hole or you have a bad efficiency when you move the emitters a little outside the focus point. also the thermal management of this very thin alu heatsink is miserable so a build my own heatsink that will transfer the massive heat wall direktly to the mag body and head:

markus

Markus,

The trick to both issues, I've found, is trimming just the right amount in just the right places. The base of the MC-E is thicker than that of an XR-E, but less than an XR-E mounted on a PCB. Mount the MC-E on a MCPCB, and you've to got to create space. Mount the MC-E bare and you've got to decrease space. In this case I mounted the emitters directly to the base with Arctic Silver Thermal Adhesive after I filed down the wall at the edge of the base ever so slightly to close the gap between base and reflector bottom. I had considered replacing the center neutral white MC-E with a warm white XR-E to ensure no artifacts in the hotspot, but didn't feel it was necessary after testing focus of the MC-Es mounted this way. The edge wall of the base still mates entirely to the outer diameter of the heavy AL reflector for a very strong thermal coupling to the base for excellent heatskinking (not to mention the two screws that hold them together).

The fact that the reflector and base module assembled are way too tall to fit in the Mag head made an opportunity for even better thermal coupling. I beveled the inside edge of the Mag head and trimme the threads off of the top of the reflector to make those parts mate over more surface area. I also beveled the outside of the base and the inside bottom of the Mag head (at the top of the threaded socket for the neck) so the base and head would mate more directly as well, and also allow the base to sit low enough in the head for everything to fit. Add to that the fact that the narrow bottom stem of the base also mates directly to the neck of the Mag body when the head is fully seated and I've got direct flush contact between the base and reflector, base and head, base and neck, reflector and head and to top things off just for good measure, what little cavity was left between the parts within the head and the head interior has been filled in with crushed Aluminum foil. All thermal contact points get a thin coat of Arctic Silver thermal grease before mating. To put it plainly, heat sinking is not a problem. I ran the 5 emitters on the IMR batteries at 16.9V fresh off the charger (and these things don't sag) for ~15 minutes straight last night. The head almost instantly became warm, but never got hot. Thermal conductivity is excellent and there's significant mass to distribute the heat.

Once I add the shark-sinked sharks, thermal output will increase, both from the emitters being driven harder (~18V out) and from the Sharks themselves, but from my test runs last night, there's still plenty of headroom for additional heat. It was concerns about heat and runtime that motivated me to purchase the Remora in the first place but at this point, I'm leaning heavily toward not including it.
 
Last edited:
Re: Introducing... The XX-Ray (pic heavy)

Wow, you completed THE build that was just a pipe dream to me when I saw that reflector/base on DX. :eek::thumbsup::clap: That is the one that is on DX right? Correct me if I'm wrong. I think there was a comment about p7's being to big to fit in that reflector setup. That made me think about using MC-E's. You've also solved the power source issue that didn't exist till now. This is really an inspiring build. I hope to complete something like your build someday. This thing must be crazy bright. Awesome build! :hitit:
 
Re: Introducing... The XX-Ray (pic heavy)

Markus,
The fact that the reflector and base module assembled are way too tall to fit in the Mag head made an opportunity for even better thermal coupling. I beveled the inside edge of the Mag head and trimme the threads off of the top of the reflector to make those parts mate over more surface area. I also beveled the outside of the base and the inside bottom of the Mag head (at the top of the threaded socket for the neck) so the base and head would mate more directly as well, and also allow the base to sit low enough in the head for everything to fit. Add to that the fact that the narrow bottom stem of the base also mates directly to the neck of the Mag body when the head is fully seated and I've got direct flush contact between the base and reflector, base and head, base and neck, reflector and head and to top things off just for good measure, what little cavity was left between the parts within the head and the head interior has been filled in with crushed Aluminum foil. All thermal contact points get a thin coat of Arctic Silver thermal grease before mating. To put it plainly, heat sinking is not a problem. I ran the 5 emitters on the IMR batteries at 16.9V fresh off the charger (and these things don't sag) for ~15 minutes straight last night. The head almost instantly became warm, but never got hot. Thermal conductivity is excellent and there's significant mass to distribute the heat.


mhh, i still can´t belive this. 5 x mc-e each with 10 watt = 50 watt and the head get only warm after 15 minutes ? not possible - imho.
all my powermag´s with a comparable setup get warm after 5 minutes and hot after 10 minutes. these (and your´s) high power flashlights are like a formula 1 car. high output but only for short time.
i also did some tests with this dx reflector, 3 mc-e´s and two r2. cut down the mag head and the heatsink with my lathe so it fit perfect into the mag head. filled up the inside with liquid aluminum and much copper powder for best heat conductiona and also this setup get very warm.
i also tryed the mc-e without a pcb, but i never was happy with the beam. still think that you had more andurance to find the perfect setup :goodjob:

markus
 
Re: Introducing... The XX-Ray (pic heavy)

Wow, you completed THE build that was just a pipe dream to me when I saw that reflector/base on DX. :eek::thumbsup::clap: That is the one that is on DX right? Correct me if I'm wrong. I think there was a comment about p7's being to big to fit in that reflector setup. That made me think about using MC-E's. You've also solved the power source issue that didn't exist till now. This is really an inspiring build. I hope to complete something like your build someday. This thing must be crazy bright. Awesome build! :hitit:

Thanks. Yes, that's the one from DX. Amazing how a $7 part turns into a multi-hundred dollar project. If DX had never come out with that (back in December), I never would have dreamt this up.

mhh, i still can´t belive this. 5 x mc-e each with 10 watt = 50 watt and the head get only warm after 15 minutes ? not possible - imho.
all my powermag´s with a comparable setup get warm after 5 minutes and hot after 10 minutes. these (and your´s) high power flashlights are like a formula 1 car. high output but only for short time.
markus

Admittedly, 5 MC-Es in direct drive off of 4 batteries only pulls half the 50W total (I measured draw at ~1.55A). So at this point, I'm only cranking out half the heat (and maybe 2/3 the light). I made one more heat-sinking modification today, I applied AAA to the emitter wires and solder joints to both improve heatsinking, protect from shorts, and immobilize the wires.

Taking heat into consideration, I'm reconsidering the Remora board for three reasons. For one thing, it will keep heat under control while on a very bright medium. For another, the difference between the 2.9A single mode or the 2.5A Remora hi-mode would likely not be detectable in terms of brightness but probably will be in terms of heat. Finally, when low mode actually looks different than medium mode, then I'll know it's time to recharge the batteries.

+1 for beamshots..

And if it had 30 emitters, would it then be the Triple-X Ray? :p

Don't give me any ideas :p. I'd need a Collossus and a private investor.

---------

Now for some bad news. Today while battling a cold and rather distracted by constant interruptions, I made the worst possible newbie error...

I had individually tested and installed each of the three sharks after wiring them and then again after mounting them to the shark sinks. After installing the sinks into the mag tube, I opted to temporarily use some power connectors instead of directly soldering all the wires together, just to allow me to measure current draw to the sharks and current output to the LEDs using my multi-meter.

Now for the blunder - I reversed the polarity on the shark input connector and didn't discover it until it was too late. In a single half press of the mag switch, all three blue sharks fried! (No reverse polarity protection on these, unfortunately.)

It may be some time before I can afford to replace the three sharks and try again. (They're the most expensive part of this project, after all.)

For the time being, I can use the torch in DD mode. Pulling less than 1.6A from 1600mAH batteries, I've got a full hour of runtime, and it's still as bright as my 3 x cool white MC-E torches pushed to the max, but it will be some time before I ever see what this XX-Ray can really do on high. :(

I was going to wait until the sharks were operational to take beam shots, but as that will likely be a long time, I'll try to grab some with the torch as-is this weekend.
 
Re: Introducing... The XX-Ray (pic heavy)

hi,

why did you not make your own regulator? it is very easy if you can solder (i think you can :whistle:) and it is very cheap - <10$ ?

here is the shematic for the regulator that i use:
http://www.atx-netzteil.de/pwm_mit_ne555.htm#Schaltung 2

it is capable of very high current with low heat. problem ist that you must put out the mag switch and take a pot with switch for on/off and regulate. but the benefit is that you can dimm from near 0 to full power steppless :eek:

markus
 
Re: Introducing... The XX-Ray (pic heavy)

hi,

why did you not make your own regulator? it is very easy if you can solder (i think you can :whistle:) and it is very cheap - <10$ ?

here is the shematic for the regulator that i use:
http://www.atx-netzteil.de/pwm_mit_ne555.htm#Schaltung 2

it is capable of very high current with low heat. problem ist that you must put out the mag switch and take a pot with switch for on/off and regulate. but the benefit is that you can dimm from near 0 to full power steppless :eek:

markus

Neat circuit, but strictly speeaking that's just a PWM controller allowing you to vary the average power to the LED, not strictly a regulator. The current in the LED is controlled by the resistors, so the current will vary with battery voltage for a fixed PWM setting (a regulator would attempt to maintain constant drive independent of battery voltage). The 5s4p configuration mentioned in the first post requires the LED voltage is greater than the (nominal) battery voltage, so that requires a boost converter (which the suggested circuit can not do).

Generally, I would be wary of simply paralleling the driver circuit. If you want all drivers to give the max current, they will all need to have matched sense pins. Const current drivers usually measure the voltage across a low valued resistor (resistor is usually between the -ve pin of the LED and -ve supply) and vary the drive to keep this constant. The voltage set point on this sense port depends on the driver can be as low as 0.15V for some or as high as 1.25 for others. Either way connenting the sense pin across multiple drivers forces the sense voltage to be the same for all. Now if the drivers are slightly mismatched, the one driver may be trying to regulate to, say 0.16V and another will be trying to regulate to, say, 0.15V, then the latter will think the drive level is too high and it will attempt to switch off. Only if the devices are well matched will you get full output from all drivers.

I would have gone for either 4 drivers with 4x 5s1p strings or two drivers with 2x 5s2p strings.
 
Re: Introducing... The XX-Ray (pic heavy)

Only if the devices are well matched will you get full output from all drivers.

I would have gone for either 4 drivers with 4x 5s1p strings or two drivers with 2x 5s2p strings.

That's good advice. Wiring would have been much more difficult in either configuration, but with thinner guage wire, I could have managed.

With 4 sharks (and no pots), I'd have a whole amp going through each emitter, instead of 700mA. While I've done that with a single MC-E (and 2 2A drivers) and been able to manage the heat, I don't think I could say the same for 5 of them at once. I suppose I could have ordered 4 with the pots still installed and then set each string by hand, but cramming all 4 into the Mag D tube on 4 shark sinks would have been very tight.

With 2 sharks, the torch would be underdriven. In the end, I think 3 was the closest to making mathmatical sense.

Led Zep has done this on a lot of his Moby **** builds in the past, so I've tried to reproduce his "proven" methods. Time will tell how well the sharks hold up. I'm hoping that being the (expensive) quality product that they are, that they're well matched.

Maybe by the time I'm financially ready for replacement Sharks, there will be a better 'Super Shark' solution, or a 60W cchipo that will fit in the Mag D neck.
 
Last edited:
Re: Introducing... The XX-Ray (pic heavy) - Beamshots added

Here are some preliminary beamshots with the Shark-less XX-Ray underdriven at ~1.6A (will be 3A when sharks have been replaced).

Rather than clicking on each photo below, checkout the slideshow


The torch specs from left to right and in order of appearance are (ignore the salt & pepper shaker lanterns):
1. MC-E cool white 40mm OP reflector DD 3x NiMH C cell
2. Lowe's Task Force 2C (35mm collimator) w/ CREE Q5 WC @ 1400mA
3. MC-E cool white w/ Fraen 34mm reflector
4. MC-E neutral white w/ Fraen 34mm reflector
5. 3x MC-E cool white with 3x 20mm Ledil Boom spot reflectors
6. 3x MC-E cool white with Chinese OP tri-flector
7. The XX-Ray: 5x MC-E neutral white with Chinese SMO penta-flector (underdriven w/o sharks)

All shots taken at ISO 200, F/4.0, 1/4 sec, WB 6000K

As evidenced by the outdoor floodlight shot, this fixed ISO setting and shutter speed make the shot appear much darker than perceived by the human eye. I should have opted for a higher ISO or slowe shutter.

Distance to tool shed from torch is exactly 40 ft. (camera to shed is 54 ft.)


control, lights off control, floodlights on

1. 2.

3. 4.

5. 6.

7.
 
Last edited:
Re: Introducing... The XX-Ray (pic heavy) - Beamshots added

Some long exposure beam shots, all taken with the same settings.

From left to right:

1. TrustFire SSC-P7 HA-III 2x18650 (same torch as in my avatar) running on 2x18650 LiFePO4
2. MC-E Neutral White, 52mm OP reflector, 4 Amp driver in Mag 2C with 2x18650 LiFePO4
3. The XX-Ray under-driven

1. 2. 3.
 
Re: The XX-Ray Completed! 5x MC-E neutral white (pic heavy) beamshots!

Here are a few new pics now that the sharks are installed. I've only compared this 5x MC-E neutral white torch to my 3x MC-E cool white torch, which was formerly my brightest LED. This one is now clearly brighter and a much nicer color too :whistle:

Control


3x cool white 5x neutral white




http://s255.photobucket.com/albums/...completion/?action=view&current=65xMC-ENW.jpg
This stressful, nerve racking, expensive, time draining, emotionally draining experience has been my last big project. Now that I have this trophy to keep, I think I'm done with big complex custom jobs like this one, at least for a very, very long time anyway.

Thank you all you enablers. See you in the dark.
 

Latest posts

Top