The OptoFire Searchlight | BEAMSHOTS on page 3 | record-breaking aspheric LED thrower

I think that even better cooling does not give you extra output, within the parameters of your build this output is about it. The heat transfer bottleneck is already inside the led and even if it was (theoretically) possible to keep the die at room temperature, there is a maximum at what light it can produce regardless of temperature.

Thanks :slight_smile:
Itā€™s a pretty big coincidence that me and the_driver both finished our Oslon Black Flat throwers at the same time xD
His Excalibur is also an amazing build, more ā€œflashlight-likeā€ than mine.

Looks interesting :+1:

Thank you! :slight_smile:

Great project and story! :+1:
More expensive than I thought it would be, but you really optimized the heatsinking.

Did you ever measure it without the collar? It would be interesting to see how much gain is possible with this LED since itā€™s smaller and more cool-white compared to de-domed XP-G2s.

What is the tint like?

Thanks :slight_smile:
Most expensive part was the collar, second is the overkill battery :stuck_out_tongue: luckily I had that stuff from previous projects so I didnā€™t have to pay $400 up front to build this light.

Never measured without the collar, however the lens is farther away and doesnā€™t use anywhere near the full 60 degrees of light after the collar, so Iā€™m working on some calculations to see what the benefit would be (for future flashlights) to have a collar with a smaller exit hole.

The tint is pure white and a little blue, likely due to me overdriving it at 600% itā€™s rated current xD

Itā€™s rated for 1.8A :wink:

Concerning tint: that does seem a bit weird. Have you tried reducing the current to see if the candela go up (while using the collar)?

Calculations:
I get a luminance of 442cd/mm2 for your light. Thatā€™s high, but sma measured a XP-G2 at 470 with a perfectly alligend collar.

Itā€™s tested values are for 1A
ā€œStatus
Product Type
Ordering No.
full production
LUW HWQP-8M7N-ebvF46fcbB46-8E8HQ65111A8023
Product Family
Color of Emission
Wavelength
Viewing angle at 50 % IV
Brightness
Grouping Current
OSLON white Cx,Cy 0.33/0.34 120Ā° 107 cd 1000 mAā€
it says 1.5A max on the datasheet, 2.5A surge.

How do you measure cd/mm^2?
Just measure the lux at 1m and divide it by the LED die area?

Also no I havenā€™t tried it at 5 or 5.5A in the flashlight, the driver I have just does 6A, the next mode is like 3A or something.

Your right, I had the wrong value in mind, itā€™s 1.5A.

Itā€™s always the same equation for luminance:
Luminous_intensity[cd] = luminance_LED[cd/mm^2] * area_of _reflector_or_lens_as_seen_from_hotspot[mm^2] * -reflectivity_of_reflector]- * transmission_of_lens[

Since it is an equation you can can calculate any of the values if you have all the other ones. I calculated the luminance using 120mm diameter and 92% transmission (normal glass).

Oh ok thatā€™s why itā€™s low.
The lens actually has a clear aperture of less than 120mm,
I would say maybe like 110-115mm because the outer edge is not properly aspherized, and just makes a big spill around the hotspot.
92% sounds accurate, since itā€™s not AR coated, and itā€™s just regular N-BK7 glass.
I get 487cd/mm^2 with 110 as the diameter.

Ok.
Why did you buy 10 LEDs if you only tested 2?

Price is cheaper per LED if you buy 10.
Shipping was gonna be like half the cost anyway, so I got 10 to have spares (I only killed one during reflow) and possibly use more in the future.
Maybe later if I feel like it and have some spare money I will buy more PCBs and see if thereā€™s any that have better performance.

Originally I thought maybe just use the same PCB over and over to test all of them, but I didnā€™t have enough time or patience to do that.
Also it seems like the PCB gets damaged after a long time under high heat, with the top unsticking, so IMO itā€™s better to only reflow an LED onto it once.

For a real test you definitely need to put each LEd on itā€™s own PCB and reflow them all in identical fashion. Otherwise the test might be influenced by how good they were soldered.

Yup. Gotta buy more of those 32mm PCBs :stuck_out_tongue: they arenā€™t that cheap!

dang a light bazooka :smiley:
great detailed build log
very impressive!

Thanks :smiley:
PVC sewer pipes make great flashlights!

Very cool build Enderman! :sunglasses:
Thanks for walking us through the entire project. Very well done. :+1:

Tips to save money:
I would try to buy the displays, the heatsink and the pvc tube on ebay and replace the battery with an 18650 battery pack from some local shop. I would also replace the blackout material with something much cheaper, especially since you stil have visible spill.

Thank you :slight_smile:

I already had the battery from other projects, and the blackout material was from the previous build, it was pretty cheap too.

Cost isnā€™t really a big issue until it starts getting into the hundreds/thousands for a single part.
Which might happen with the next build I doā€¦ :sunglasses: who knows.

Crazy light Mr. Enderman. Crazy light indeed. :+1:
It reminds me on N10Sivern Zombie Slayinator build which unfortunately canā€™t be seein this thread: N10sivern's Zombie Slayinator - 3rd Annual BLF Contest Hand Made - Grand Finale Posted! cause photobucket images disappeared but you can see his video forwarded at 3.09 here: The Zombie Slayinator - YouTube