Franken Mag Shorty --XML King of Throw --127,000 lux @ 1 meter

So far to date, the Reflector based King of throw has been the SR90 Intimidator, at 110K to 125K lux based on various reports. I had one at one time, what I was intimidated the most by is its size, it's freakin HUGE .

So I set out to build a powerful yet compact XML mag to beat the throw of SR90 intimidator, using reflector not Aspheric lens. This not your average XRE R2 pin spot thrower, it's got power too, estimated about ~1300 lumens @ 4.2Amp

The host is a Mag 2D to 1D Cut down by Human Lathe method mated with DX Turbo shower head previously described here:
http://www.candlepowerforums.com/vb...ag-D-like-a-Human-Lathe&highlight=lathe+human

Under the carbon fiber neck, is a big chunk of Copper heat sink recessed to mated with DX 3" SMO reflector. The reflector bottom is trimmed a bit to go inside the neck.
The Reflector fits loose, So I made a centering ring ( the blue ring), which is made out of 73mmOD-66mmID Hub centric ring from after market Car wheels (Leftover parts from my previous hobby)

The Front window is 73mm UCL glass from flashlightlens.com. A again I had to make a centering ring for the glass window.


XML-U2 on 20mm cutter star mounted on Copper heat sink.
Driven by 4.2Amp 3 mode 7135 based DIY driver (Shinning beam 3-mode 8x7135 2.8A + 4 extra 7135 1.4A slave)

More info on wiring more 7135 boards in parallel to the 8x7135/3 mode board:
http://www.candlepowerforums.com/vb/showthread.php?313593-help-on-making-12x7135-3-mode-driver



Left. XML T5 Neutral in mag Rebel SMO. Right XML-U2 in 3" DX SMO.


10 meter white wall shot, Auto exposure:



In door shoot out:
left----------- Mag 1C Deep Rebel SMO, XML T5 Neutral 2.8A Max.
Middle-------- Mag 1D 3" DX SMO, XML U2, 4.2A max
Right--------- SSK-33 35W/28W HID HA III




The following beam shots are manual exposure, ISO 200, 1/50 of sec. 10 meters.
Low exposure were used to avoid over-exposure to show spot intensity differences.

XML T5 Neutral 2.8A (~930 lumens) in Mag Rebel SMO: 41,000 lux @ 1 meter


XML U2, 4.2A (~1300 lumens) in 3" DX SMO: 127,000 lux @ 1 meter


SSK-33 HID at 35W(~3500 lumens) : 135,000 ,lux@ 1 meter

Very nice!! Question for you. How do you manage to clean your lens and reflectors so well? Photoshop?

I have a hard time cleaning my lens and reflectors. There are always some fingerprints and dust.

Clean it with the cleaning cloth of my eye glasses.

No Photo shop, just 30 yrs of experiences cleaning eye glasses

Nice work! but eww a cutter MCPCB? You could see a further boost just by having the U2 on a better MCPCB, especially overdriving it at 4+A. Did you see the results I posted of comparing the cutter MCPCB to the KD MCPCBs in the " Why a good thermal path really matters" thread?

You already have it epoxied down, but if you want to try a different one PM me and I'll slip a KD MCPCB in a padded envelope and send it your way, so you dont have to wait 3 weeks. If you already dont have a favorite method for reflowing LEDs onto boards, if you have a flux-pen or bottled rosin flux for solering, you could probably move the LED over with a skillet that wont be used for food, and doing it on the stovetop while SWMBO isnt around. Or with a hot air gun.

Thanks CKOD,

The key is not necessary to have the "best" any given component,
the key for optimizing any given system is to eliminate "weak links" that forms a "bottle neck".

The lux number speaks for itself, which means there's no "bottle neck" in this system, the PCB is not a bottle neck here & improves upon "non bottle neck" = improves nothing.

I have used the cutter XMLon 20mm star up to 4.6 Amp, see this shoot out thread, at 4.6A each, it beat out fully driven quad P7 & 5x MCE. If this xml PCB tops out at 3.5Amp, this wouldn't have been possible.

http://www.candlepowerforums.com/vb/showthread.php?314308-5-Lights-51-LED-Dies-%28and-a-bit-of-wire%29-300W-and-17-800-Lumens!-Big-Mag-Shootout

I was being conservative, thus drove it to 4.2 Amp. If I add another 7135, I could have made it over 130K lux. I have experimented several PCBs on this including Dx copper star XML & cutter XML on 14mm PCB before setting down to the 20MM cutter XML on PCB.


Also, which cutter PCB did you use? Was it a cutter XRE PCB?
If you have not tested XML on 20mm Cutter XML PCB (purchased as a ready-made item), you should get one & give it a test. I don't think all the cutter 20mm PCBs are the same. Cutter XRE PCB are not meant to mount leds & drive to over 4 amp.

Cmon now, apples to apples I wouldnt say cutter XM-L stars suck because I tested XR-E PCBs

Did a comparison of the Cutter XM-L boards and KD XM-L boards. Mounted a U2 2S XM-L to one of each PCB (reflowing onto pre-tinned pads with RMA flux)

Screwed the pcb to a heatsink I pulled from an old mobo for the mosfets, using thermal compound(one at a time of course). Ran each LED at 3.73A for 10 minutes to reach a somewhat steady equilibrium. I recorded the Vf at the power supply and light output on a lux-meter 6.5" away, neither the clamp for the heatsink nor the lux-meter moved between tests.

Even soldering the wire leads to the PCBs with them mounted to the heatsink, I noticed the KD board was harder to solder. I also noticed the heatsink became hotter much faster with the KD compared to the cutter PCB (hotter faster is good!)

Results:

Code:

                KD     Cutter
Initial lux    11.65k   9.17k
T+10 Lux       10.31k   7.47k
Lux drop       11.5%    18.5%
Initial Vf*    3.78v    3.73v
T+10 Vf*       3.70v    3.63
Vf drop        .08v     .1v

*The Vf was at the power supply, not at the LED, so there is ~6ft of wire that is probably a touch small for 4A between the LED and the voltage measurement.
I think the results pretty much speak for themselves, the KD boards are the better of the two

Click to expand...

I didnt use their premounted ones, I reflowed it on myself, but I have a hot-plate and temp controlled hot air gun, so while it wasnt an oven with a profile to follow, it wasnt exactly a hackjob, and did them both, so its not a variable between them. It was two different LEDs, I didnt move the one to the different circuit board, I felt dismounting and remounting it would be a bigger variable then using 2 from right next to each other on the cut-tape.

Now, both being U2 bin allows for a 6.6% variance from 300lm@ test current, upto 320 lm @ test current, which doesnt fully explain the difference in the initial, and this only a sample size of 1, so I'm not sure how much to read into the initial difference. but the lux drop is very telling imho.

Do you want to test Cutter pre-mounted XML U2 vs. your advanced hack job? :naughty:

I have on extra U2 on 20mm star, you can have it for $15 shipped.
Hack, I think I paid close to $15 just for shipping/handling to get a couple leds in from cutter.

But the way,
Why the big difference in initial lux?
It should be about the same as nothing is heated up at time zero, could you un-solder & swap the LED with the stars & test reciprocally again?

How do you know the difference aren't from the two leds?

Nah, I'm good in the XM-L department for now.

Along with ~50 KD MCPCBs for mounting.

EDIT: and regarding the inital lux and T-0, I'm not using a data logger, just grabbing a number off the meter once it settles somewhat, and I'm also watching the power supply for a second or two to make sure it enters constant current mode and is staying there correctly, and I think the 3-5 seconds or so it takes to get a reading is enough time for the die/ceramic to heat up enough to make a difference. I would swap and test, but I already have pulled the LED from the cutter board, and have used them for other testing. However, I did have one of my test pieces, ran it with my finger on the back of the KD MCPCB, and at 3.73A again, it was quite warm after 3 seconds, and uncomfortably so after 5.
We'll just have to hope that Cree's binning is correct. Otherwise the 2nd LED was ~21% less the the other, and since lux and lumens are proportional (I.E. 2x the lumens at the same distance = 2x the lux... right?) That would make the dimmer of the 2 a T3 bin :sick2: (300 *.787 = 236)

ma_sha1 said:

By the way,
Why the big difference in initial lux?
It should be about the same as nothing is heated up at time zero, could you un-solder & swap the LED with the stars & test reciprocally again?

How do you know the difference aren't from the two leds?

Click to expand...

I was wondering the same thing. It seems that the difference may not be because of the boards, but possibly because of variation between the two emitters. In order to provide convincing evidence, you need to re-test and eliminate the emitters as factors in the equation.

Ok guys, making me get all messy with thermal grease again :nana: I find your demand for scientific testing and valid comparsions refreshing compared to some communities where whatever defacto standard is thought of best is "the best" and is "common knowledge" without any real, valid proof.


I dont have the time to do the 10 minute thermal test again, but I did check out the initial numbers

I pulled an XM-L off the reel, put it down on the cutter PCB, put it on the same heatsink as before, set up the lux meter, power supply on 3.66A this time. 11K lux and droping quick from turn on. I moved that XM-L to a KD MCPCB, mounted it on the heatsink and put the heatsink back in my clamp, same position as before. Fire it up and its at 13K lux and dropping from start up.

Both times I had to let the power supply settle down for ~1-2 seconds for there to be a steady reading on the volt and amp display (I.E. it stays in CV and once its past its current limit by a little bit it clicks over to CC and settles in on the set current.


I found it wierd too but considering its size and power input... its gotta heat up quickly. I just weighed one at, and its pushing the lower limits of my balance, but I got .0025 oz, lets convert that and round it up to .1g and lets make the generous assumption that the various materials that make up that .1g have a specific heat of 2 J/(g*K) Even though in reality its probably alumina or similar which comes in at .8 J/(g*K) (silicon comes in a .75 in the temperature ranges we're working in) that makes .2 J/K or .2 (W*S)/K. invert that , and you get 5 K/(W*S) (kelven per watt-second) and we're inputting 3.4v and 3.66A, making for 12.4W in, making the temperature rise at 62K/second. That means in 1 seconds it will jump from 25C to 87C, unless its cooled by conduction into the PCB.

Another nice build.:twothumbs So with 127K lux from a reflector, who needs an aspheric or a Rebel reflector?

BTW, is that DX sku.47996?

Nice build, great beam!
Almost makes me go back to reflectors...

Great builid man!Though, are you sure your numbers are right? How did you measure 1300 lumen?Or just according to Cree graphs?
I have the same reflector,Cutter U2 2S(used one of the best of the best i bought, i measured their performance),reflowed them (perfectly i think) on a kd 21mm board,4amp driver(not 4,2 that you have) and i get max 85.000 lux! How did you reach so many lux? Will continue the tests though!!

Al Combs said:

Another nice build.:twothumbs So with 127K lux from a reflector, who needs an aspheric or a Rebel reflector?

BTW, is that DX sku.47996?

Click to expand...

Thanks! Yes, it's 47996.

Walterk said:

Nice build, great beam!
Almost makes me go back to reflectors...

Click to expand...

Thanks, I am also very happy to be off the Aspheric limitation finally.
This XML on reflector is slightly higher lux than the SST-50, 5-Amp 3" Aspheric Franken
MagDEFT (Clocked in at 115 K lux) and Franken Mag Dragon SST-50 6 Amp 3" Aspheric
(Clocked in at 125K lux).

ergotelis said:

Great builid man!Though, are you sure your numbers are right? How did you measure 1300 lumen?Or just according to Cree graphs?
I have the same reflector,Cutter U2 2S(used one of the best of the best i bought, i measured their performance),reflowed them (perfectly i think) on a kd 21mm board,4amp driver(not 4,2 that you have) and i get max 85.000 lux! How did you reach so many lux? Will continue the tests though!!

Click to expand...

1300 lumens is emitter lumens not OTF measured lumens, just an estimate based on a cree data sheet & an over driving graph from saabluster, assuming heat sinking is adequate.

Only 85K lux with 3" SMO & 4Amp sounds like there is a bottle neck somewhere, I don't trust self re-flow, tried a few times, I either killed or damaged the led while doing so. You can't really tell if you damaged it but didn't kill it. I no longer attempt self re-flow.

For reference, there was a set-up similar to mine earlier (BTW, I already started my build back then but trying DX copper XML twice has delayed this project by 2 month, DX copper XML failed me twice before I gave up on that)
http://www.candlepowerforums.com/vb/showthread.php?312461-TMR-73-SMO-reflector-Flashlight-82-000-lux

This build by luminator is XML T6, 3" DX SMO driven to only 3 Amp & he got 82K lux. So your U2, 3" SMO, 4 Amp for 85K lux is under performing. I have on extra Cutter pre-made XML U2 on their 20mm star, if you want to try swap it out, PM me.

Also, there might be meter to meter difference. My meter is Extech EA30, one of the most accurate & consistent meter on CPF. Used to have DX meter that was under reading by 17-20%.
I read at 10 meter distance & convert to lux @ 1 meter.

Also, I have the reference light XML T5 Neutral 2.8A/Rebel SMO 41 Klux & T6/Rebel SMO 2.8A 46K lux, 35W HID 135K lux, all are right where they are surpose to be. So I am confident with my number on this over achieving shorty.

One last thing, Measure the Amp at led & see if your driver really deliver 4Amp or not,
if you are using DX SST-50 4A driver, they are known to deliver only 3.5Amp.

Nice build!!! :thumbsup:

For my mod, I make peak beam about 82000 lux, same reflector same bezel but drive at 3A

THanks for the details of your setup! I will try everything from the beginning to see where is the bottleneck. My setup@4amp is performing somewhere 960 OTF measured, so it is quite a good and logical number i think, the reflow process i am doing is quite successful have done it several times. I have to change glass for sure, it doesn't seem to have ar coating. I measure my data at 4meters, will try longer distance as you do. Also, the current is measured on the led, i have the driver that you said, measured the current and was really 4amp, though i may have to let it play for a bit more seconds to see if it is regulating at lower currents.

Anyway thanks for all the details will try more things to do!

ergotelis said:

Also, the current is measured on the led, i have the driver that you said, measured the current and was really 4amp, though i may have to let it play for a bit more seconds to see if it is regulating at lower currents.

Click to expand...

I first tried the Dx SST-50 4A driver, it measured at 3.5A @ led, there are post on DX confirming that the drive delivers 3.5A, not 4A. What amp meter are you using?

I then tried the DX SST-90 6.5A driver, it delivered 5.5A, gave up on that as its too high. Would be perfect for SST-50 though.

So I set out to make 12x7135 4.2A driver instead. Since I never made one before, I posted a help link at the CPF driver section: http://www.candlepowerforums.com/vb/showthread.php?313593-help-on-making-12x7135-3-mode-driver&highlight=7135+ma_sha1

It measured 4.21 Amp at led, perfect! But it was a pain to solder, I made 2 just in case, one of them didn't survive soldering & gave less than 1 Amp after I was done & lost all the modes

Really cool build Ma_sha1

I like how you used the Deft bezel and found a perfect lens from DX, that fits it perfectly. I may now have to try and coppy this masterpiece.

Maybe a copper to LED bonding and an XM-L at 6A with your set-up...hopefully no blue.

bigC

Thanks, Big C,

The lens is 73mm UCL lens, not from DX, it's from flashlightlens.com.

6A is insane, if 5.5A is good enough for you, I still have the 5.5A driver, it was bigger than mag tube, So I grind it down about 4mm OD, (cut 2mm all around from the fat 3mm copper rim and re-build the rim electrical path with some Teflon wires).

The mod is necessary so that I could fit into Mag D tube. PM me if you want to try it,
it's 5 mode 12V driver, so you can do a 3x17670 in a Mag D shorty.