Scratch-built 7x C8 reflector high-output thrower

Looks very impressive. Any chance for beamshots?

I’ll take some wall beam shots to compare beam size before and after, but probably won’t get around to outdoor shots like in post #35 for awhile.

Oh I actually didn’t see the pictures in post #35 because I was using my cell phone. Now I see it. The beam is marvelous! I wish the manufacturers would make something like that. Maybe even use XHP35 HI.

Easy B,

Exceptional minds think alike, and history repeats itself...

When I first started this hobby 15 years ago there was a one man army named Wayne and his company Elektrolumens that pioneered the formula behind your outstanding design. CNC machined head and body lights that utilized clusters of LEDs and reflectors to achieve a massive flood wall of light with extraordinary throw character as well, the holy Grail of flashlights to have them BOTH.

He made his own designs and I was awestruck with the incredible performance of geometric array lights. He started off with 5,6,7 and up to 12 in one gargantuan head. Back then using NiMH batteries because LiIon had not yet evolved fully enough to supply the insane parallel current demands.

Wonderfully you have included active cooling, a necessity for the insane waste heat your array will make for continuous duty. I suggest going a step further and going to a liquid cooling or reverse Peltier design for even greater effeciency.

Standing ovation BRAVOS for your ingenuity and skills. I can't wait to see how it evolves and WOWs even the most crusty die hard of us.....

Best of LUX :-)

“EasyB” Now that’s thinking outside the box :beer: You are gem here and I love to see people like you making their visions into reality. That is quite an accomplishment, I am very impressed.

Since you mentioned modding this Monster, you may be interested in these reflectors for maybe closer to 2 mile range. They are 2X deeper than the C8. :beer:

https://www.aliexpress.com/item/1PC-NEW-HS-802-802-LED-Flashlight-Reflector-Aluminum-Smooth-SMO-Torch-Light-Cup-Dia-41/32821558396.html?spm=2114.10010108.1000014.1.6c9c3f16PQVraI&scm=1007.13338.98644.000000000000000&scm_id=1007.13338.98644.000000000000000&scm-url=1007.13338.98644.000000000000000&pvid=7e42b83e-1f21-470a-962e-46a745da84ca&\_t=pvid:7e42b83e-1f21-470a-962e-46a745da84ca,scm-url:1007.13338.98644.000000000000000

https://www.aliexpress.com/item/41mm-D-x-60mm-H-SMO-Aluminum-Reflector-for-CREE-XM-L/32851870851.html?spm=2114.search0302.3.23.1b144e20yYymsq&ws_ab_test=searchweb0_0,searchweb201602_0_10130_10152_10151_10815_10814_10618_10059_10696_10084_100031_10083_10547_10624_10546_10623_10307_10548_10341_10065_10340_10068_10343_10342_308_10103_10620_10344_10325_10545_10622_10324_10621,searchweb201603_0,ppcSwitch_0&algo_pvid=2c89ac56-d1f2-42d4-b855-0106a4913ced&algo_expid=2c89ac56-d1f2-42d4-b855-0106a4913ced-3

Mmmkay, are you going to dare with 19x or 43x versions? LoL!

Cheers :-)

Thanks guys. :slight_smile:

I love circles in groups of 7 :slight_smile:

In 2005, my insane high-performance no holds barred wall of light WITH staggering long throw was an Elektrolumens circle-of-six plus 1 center with 27mm McR reflectors for each of 7 (3W Luxeon LEDs) ....

NOTHING but massive HIDs or Xenon short-arc lights could come close to that discrete LED/reflector array head. It had an infinite variable adjustable dimmer pot knob for intuitive output control (and ultimate runtime preservation need be) ; the analog predecessor to NARSIL) :-)

BRAVO EASYB for building, and striving to go beyond, your epic creation !! Standing Ovation my friend !!

:-)

Another emitter option for this light is the XHP35 HI, but I would need to add 2 more cells to make it 4s. 4 decent cells in series would drive 7 XHP35 very well with a FET driver, around 2.4A per emitter. This would make around 15000 lumens and 940Kcd.

I am giving up some throw by going to the sliced xhp50.2, but I think the beam will be great for 200m to 500m where it is more likely to be used. (Though who am I kidding, I don’t really have a use for this light :smiley: )

I finally got around to putting in the XHP50.2s and after just a few short bursts with some mostly charged liitokalas some of the emitters died. A connection in the XHP50.2 goes bad and half the dies go dark. I had read about this happening but I didn’t think it would happen so often. 5 out of the 7 have failed now. Pretty disappointing, but I think I have to abandon the XHP50.2s.

One option is to put in 4 sliced SST40s (2s2p configuration) and 3 sliced XHP70.2s. This should get me similar performance, around 500 Kcd and 21ish Klumens. Not sure what the beam would be like.

Sorry about your misfortune. I hope you get something in it and get it back up and running. That had to hurt though, losing those XHP50.2’s!

Just ordered some emitters to do this. I estimate with 4300mAh aspire 26650s each 2mmx2mm die should get around 5A, or around 40A from the battery in 2s configuration. I estimate it will do around 600kcd and 22000 lumens.

Emitter upgrade:
I changed out the emitters for 4 sliced SST40s and 3 sliced XHP70.2s. The original 7x XPL version of this light worked well, but with all the very-high-output lights coming out my desire for power motivated me to redesign the light. 7x sliced XHP50.2 would have worked great, but when most of the XHP50.2s failed I looked for other options. I predicted the SST40 + XHP70.2 arrangement would give similar high intensity and output as the XHP50.2s, but how would the beam look? The idea was to have an intense beam center that would reach out far and provide the viewer with lots of detail of the subject/target, then the XHP70.2s would provide a wider, less bright, view of the surrounding scene.

I had never mixed different sized emitters in a light so I didn’t know quite what to expect in the beam. The beam from the XHP70.2s would have about twice the diameter as the SST40 beam. My concern was whether the XHP70.2 beam would be too dim compared to the SST40 beam. If this happened it would not really help with illuminating the surroundings far out. The output of the flashlight would still be high, but the two beams would not be working well together to illuminate the same scene. The SST40s are arranged in 2s2p configuration and the XHP70.2s are 6V. The forward voltage of the 2s2p SST40 arrangement is similar to an XHP70.2 so it should get similar current to each of the XHP70.2s (i.e. each 2x2mm die should get similar current). With this knowledge I estimated that the 4 SST40 reflectors would provide around 400kcd combined and the 3 XHP70.2 reflectors would provide around 150kcd. So the larger XHP70.2 “corona” would be around a quarter of the intensity of the beam center. I figured this was close enough to the beam center intensity to effectively illuminate the surroundings of a far away scene.

On the left is the beam of the 7x XPL original light and on the right is the new configuration. The beam center intensity in each case was adjusted to be the same at the time of the pictures and the exposure settings were the same. The 3-petal pattern in the right picture is apparent because the SST40 beams are brighter than the XHP70.2 beams. As I mentioned in the first post, the plastic lens flexes and actually causes the outer reflectors to point slightly outwards, which effectively widens the beam. The beam divergence from a single reflector is greater than the divergence angle of the reflectors, so far away the beam loses the 3 petals and becomes round.

Now for some measurements. With charged Aspire 26650s I measured 35A at the tail using a 1.6 mOhm shunt in place of the tailcap (the actual tailcap resistance is 1.0 mOhms). Using ceiling bounce I estimated 23,000 lumens at start and 20,000 lumens at 5s. I think this quick output decrease is mostly due to the battery voltage falling rather than mostly heat. Based on Mooch’s test it looks like each cell’s voltage drops by 0.2V in the first 10s. Beam intensity at ~7m was 530kcd at 5-7s. With Liitokala 5000mAh 26650s at 3.75V OC I measured 19A and 14,000 lumens.

So far I only have shined it 50 or 70 meters, and the beam looks very smooth. I’m excited to try it at 300-400m. :slight_smile:

Okay, which do you like better in real world use - the original configuration or the new one? I personally prefer a wider spot size. I don’t need to see things hundreds or thousands of meters away, so a super thrower with a pencil beam is only something I’d show off to my friends and never touch it otherwise. The 3-petal pattern would probably bother me if it showed up in real-world use outdoors. But I suspect it only really shows up in beam-shots against a flat surface. So, for me, the new configuration would probably be better.

Well, honestly I haven’t had a lot of real world use. :person_facepalming: The new configuration has close to the same beam intensity but with over twice the beam area, so it will be more effective at illuminating a scene. But it also takes a lot more power, so it’s a trade-off with high output throwers vs narrow beam throwers.

The 3 petals disappears after 10-20m. Those beamshots are on my ceiling which is only ~10ft away.

I just wanted to chime in and say that this build is AWSOME!!! has me on the edge of my seat and is inspiration for a upcoming project I will be doing.

Very very cool. Keep up the great work!!!

Congrat’s on a magic torch. it’s a beauty.
There are a lot on here not capable of, but would love to have one.
Hint. Hint.

Thanks guys. I think this will be its final form as it is practically maxing out the power capability of 2x26650 cells. Total power use is 2x4.15V(35A)=291W. Heat dissipated in the cells is (35A)(35A)(2x0.018ohms)=44W. Heat dissipated in other circuit resistance is approximately (35A)(35A)(0.015ohms)=22W. Which leaves 225W going to the emitters.

I did a sustainable output-temperature test. For this test I powered the light with my power supply and held the light with my hand by the tube with no batteries in it. I held the light pointing horizontally. It was indoors with no airflow or fan in the room, but I was sweeping the light slowly from side to side (maybe 1ft/s speed).

The ambient was 22C. I started the test at 10A and ended up at 7.5A. I measured the head temperature using an IR thermometer pointed at the anodized black heatsinks.

At 7.5A the temperature stabilized at near 57C (temperature was stable for 4min). The whole test took 25min. Output at 7.5A was 6000 lumens. This was with constant current so in actual use (FET with PWM dimming) the output would be slightly lower, like by 10%.

Edit: the difference would be more than 10. I was thinking of the efficiency difference between PWM dimming and current limiting linear dimming which is around 10, as measured by me and Texas Ace. Using the battery source and PWM instead of the power supply would cause additional heat production in the circuit resistances, mostly in the cells themselves.