I made a mellow S2+ triple shorty with gen.2 92CRI 3000K Oslon Square leds a while ago with a 2x7135 driver (700mA) as a reading light for my girlfriend, and it occurred to me today that with its very flat output it is a good light to measure what the actual light loss is from this common BLF-mod. The typical Carclo 10507 TIR triple optic claims 10% loss, but that is measured with an XP-E size die, and there's the rim at the front that blocks some light too. And I like to use the glass lens to protect the optic from scratching.
I found a piece of flashlight head in my junk pile that had the wrong threading but could conveniently be forced between pill and battery tube so that a bare triple head was constructed for maeasuring the primary amount of light in my integrating sphere. Both set-ups have exact the same regulated current and I suppose the leds put out the same amount of light.
So here are the numbers:
Bare leds: 241 lumen
Assembled S2+ with clear Carclo 10507: 188 lumen. So there's a 22% loss
Assembled S2+ with frosted Carclo 10511: 175 lumen. 7% loss going from clear to frosted
Assembled S2+ with frosted Carclo 10511, stock glass lens added (no AR coating): 164 lumen. 6% loss from the lens
So going from bare leds to an assembled S2+ with clear triple optic and non-ARcoated glass lens will cost you 27% output. (edit: Nichia 219 leds and probably XP-G2 as well do a bit better in this optic, see post#20)
I do not think if measuring the same for a black S2+ host will obtain any different numbers. Using other leds will give slightly different numbers I guess, because the light will enter the optic somewhat different but I doubt the difference will be much.
How much overlap is there around the inside of the bezel (over the edge of the optic)?
I’d imagine, with such small optics, that any little overlap would make a massive difference.
Nice test anyway!
The overlap is there but it would not differ more than a few percent, would it? Mind also that the center part of the little TIR’s is brighter than the reflector part.
link to djozz tests
Very good data collection djozz. I always wondered how many stray photons were serving a life prison sentence inside thes optic heads.
Wow, almost a quarter of the lumens lost from the optic. That’s significant.
So does this mean that a reflector is more efficient since most of the light is reflected forward?
But the outer edge is responsible for more throw, so that will be slightly compromised. I didn’t think the reflector brightness varied across it’s radius. I may be wrong.
I’d say that’s more than a few percent, but there’s only one way to find out!
I will try to think of a way…
link to djozz tests
Ok, the BLF A6, stock with 5A led, mode 3 of 7, so 100% 7135 controlled and fairly constant:
Bare led: 80.9 lumen
Bare led + white plastic centering ring: 79.7 lumen. So 1.5% loss by the ring
Assembled BLF-A6 with non-AR lens: 67.2 lumen
Assembled BLF-A6 with stock AR-coated lens: 69.1 lumen. So the AR lens gains 2.8%.
So going from the bare led to the assembled stock BLF-A6 with AR-coated lens will cost 14.6% of the light, with non-AR coated lens 17%. Note that the coated lens is 1.5mm, the non-coated lens is 2mm, this will not so much absorb more light but it may cause some extra light loss to the side.
If you calculate back the effect of just the reflector (well, almost, plus loss from the bezel which is as good as completely out of the way in this light), I get 10.3% loss (so 89.7% efficient)
So using this 24mm head tube-style host, in single led form with a reflector it will extract more light (83%) than in triple led form with Carclo triple optic (73%)
link to djozz tests
Thanks Djozz for the data.
So what you’re saying is that we need to find about 30% more power for our Fun Build triples to make up for this loss…
I’m wondering what the results would be for dedomed or XP-L HI emitters?
- LEDs & Other Stuff - - UI CheatSheets - - CRX Flashlight Builds - - CRX Stuff for Sale -
..or set the trees on fire with them for the extra lumens
link to djozz tests
- LEDs & Other Stuff - - UI CheatSheets - - CRX Flashlight Builds - - CRX Stuff for Sale -
Thank you djozz, appreciate the testing and info!
You could always try the triple with a widened bezel (with a file or penknife)…
I would do it myself, but my XPL-HI triple is disassembled for repair.
Either way, 3 XPLs make up for the slight loss in output.
Triple + black electrical tape = fire. Be sure to set it to low.
Care to tell that story?
I’m with Caleb here. Wow. Thanks for the testing djozz.
djozz quotes, "it came with chinese lettering that is chinese to me".
"My man mousehole needs one too"
old4570 said "I'm not an expert , so don't suffer from any such technical restrictions".
Old-Lumens. Highly admired and cherished member of Budget Light Forum. 11.5.2011 - 20.12.16. RIP.
Yes, I thought this was a quiet sunday afternoon little pastime project avoiding the things that I really had to do (but were less fun). But since I'm at it anyway here we go, electrical tape, last set of measurements on this subject:
You can see in the top view photo that the electrical tape is not touching the little reflectors (undisturbed yellow reflection)
Here's how it looks in the dark:
The numbers (the flashlight from the OP was disassembled again
) :
Bare leds: 241 lumen. I remeasured it and the output is consistent
Carclo 10507 added, the side packed with electrical tape: 194 lumen.
This is a 19.5% light loss from just the optic so just 80.5% efficiency. I looked up the specsheet for the 10507 and it claims 88.7% for the XP-E and 86% for the Rebel. It sounds logical optically that the bigger the die, the less efficient the optic because a bigger part of the object is not in the best imaged middle.
This also implies that the flashlight head, blocking part of the optic, just eats up 3.1% of the light, so that is not as dramatic as it may look like.
I had another thought: the flashlight from the OP has gen.2 Oslon Square leds which have a funny-shaped dome that perhaps messes with the optic?:
I have another (coincidentally also red coloured) S2+ triple mod with 3000K Nichia 219C leds. I wondered if the 10507 has a different efficiency for those. So I did the same electrical tape measurement with the 219C light as was done with the Oslon Square light. I chose also a steady low mode for this.
Results:
Bare leds: 82.5 lumen
Carclo 10507 added, the side packed with electrical tape: 69.4 lumen
So that is 84.1% efficiency, quite a lot better than the 80.5% from the Oslons! And actually the beams from the two lights would have given a clue already, the 219C beam has a narrower tidier hotspot than the Oslon Square beam (left Osram Oslon Square beam, right Nichia 219C beam, the Oslon beam was a bit brighter, but you can still see the difference, and better in reality) :
Could it be that the XP-L Hi with its messy beam through the 10507 has an efficiency even under 80%? Well, I'm not going to find out, I've done enough of these measurements now
Edit: aaarrggh, messed up the last beamshot, forgot that the Oslon Square light had its frosted optic back. Had to change that for a clear 10507 and here is the real comparison:
So the Oslon actually gives a tighter hotspot than the Nichia 219, but still more light is lost through the optic. It is getting too complicated for me now
link to djozz tests
Great testing DJozz.
That said, there are a number of things that can be done to reduce the light loss:
The combination of not using an outer lens and using a more revealing bezel should reduce the light loss from almost 30% to around 20%…. much closer to the 14% loss seen in a typical reflector light with AR coated lens.
Thanks a lot for this work! I always wondered how much light gets lost in optic and the overlapping bezel.
BLF Q8 Janus --- Emisar D4 RGBW
Convoy S2+ Multi Color --- Convoy S2+ Multi Color 18500 Shorty
6th Annual Light Contest Entry --- 7th Annual Light Contest Entry
Same here, thanks for the numbers djozz!
Well that’s gross….
I was getting away from 20mm triples now just because I want more throw out of the little lights, but man, that’s major. Explains a bit of how hot they get too.
My Favorite Modded Lights: X6R, S8 , X2R , M6, SP03
Major Projects: Illuminated Tailcap, TripleDown/TripleStack Driver
Thanks for the test!
I wonder what’s the case with XP-L HI and XP-G3, in my opinion the light loss with the HI is massive, as my triple DD copper sinner 18650 gets host faster where the optic is than where the pill touches the body. Most likely light loss and scattered to the sides but have no way to test it unless I build a “lumen tube”, ceiling bounce won’t be precise for this test.
It seems like searching for good reflectors and hosts will become a top priority from now on, rather than stucking carclo triples and quad into everything.
USB power meter/tester thread
I made a mistake, just edited the beamshot in post#20
link to djozz tests
Yes. Of coarse we will have light loss through our triple optics. You will have light loss with single emitter reflectors too. There will always be light loss of some sort. Trying to keep photons all rounded up is like herding cats.
Are triples worth it. Let’s compare the same light with single emitter with reflector against a triple optic.
I think it’s more what you want the light to do.
From Djozz’s testing it sounds like a well-built single-emitter light may have around 15% light loss into the reflector and lens compared to maybe 20% loss into the bezel and optic of a no-lens triple. In my opinion, that 5% difference in efficiency isn’t enough to make a significant contrast between the two types of lights.
It will ultimately still come down to mid-lumen throw vs high-lumen flood. But as usual, it’s been a pleasure experimenting for the sake of knowledge. Thanks Djozz!
Yeah, especially when it’s completely overshadowed by three times the emitter lumens.
This test is very interesting but let’s not forget about efficiency gain with lower currents with multiple LEDs. According to data from Cree PCT:
Single XP-L HD V6 driven around 10W with Tj=60C (attainable with copper DTP) produces 1165 lm (@10.174W to be precise).
Triple XP-L HD V6 driven at 9.94W produce 1583 lm at the same Tj. Quad XP-L HD V6 driven at 10.01W will output 1681 lm.
That’s around 36% more for triple and 44% for quad. Higher efficiency means less heat and I assume here exactly same junction temps. So even if TIR waste most of this there’s still plenty of gain over single LED with reflector. As to perceived more heat for triples it also can be explained with the LED count – three LEDs have 3 times the contact area so will be more efficient with heat transfer to the radiator so it will heat up faster. This also means LEDs could run slightly cooler compared to single LED so will gain still a few percent of efficiency. If there’s a good cooling all of the above should be quite true and the difference towards multiple LED setup could be even higher. But there’s too many dependent variables to accurately predict in specific case, and with small hosts heat sinking may not be adequate. Someone with required equipment at his disposal would have to test that.
I LOVE THIS FORUM.
Just when I think I have an understanding of the variables at work, more information is shared by another member from another part of our planet, that sheds more light (pun intended).
Awesome!
Good point. Assuming your light is driven at the same current and is using the same type of emitter, a multi-emitter light should always be more efficient than the single emitter. This is because LEDs are inherently more efficient at low currents.
I think the issue with heat and multi-emitter lights only comes up when they’re heavily driven. You can’t drive a single XPL or Nichia 219C light at 15 amps… the LED simply isn’t capable of pulling that much current. But you can drive a triple or quad at those currents.
Pages