Same, I think this upgrade will make the whole project worth it, despite lower than expected candela.
It should also make it easier to focus thanks to the larger die, so who knows, it might get higher than 7.5M 
no promises thoughâŚ
Nice pics!
Have you tried varriing the height of the LED?
Genial :+1:
What is a reasonable price for the CFT-90 LED? I heard some rumours about 70-100 Euros/Dollars, so probably something youâd better handle with velvet gloves on.
For a second, I thought I was in the wrong tab of my browser⌠which is LPF 
Nice, really nice pencil beam Enderman !
Thanks 
There is no need, since I can simply vary the height of the reflector by adjusting all three screws equal amounts.
The problem with focusing isnât the z-position of the LED or reflector, itâs getting all areas of the reflector to be reflecting the LED image perfectly forward, so tilting in different directions.
Thanks!
Yeah I paid over $150 CAD for it, so about 100 eur⌠itâs not cheap.
Haha thank you
A couple of you have seen my creations and now Iâd like to try one of this type.
Would this reflector be too cheap and nasty for this application?
Someone on TLF is making a huge project with 3 of the 24â ones, but havenât seen any updates in a while.
I have no idea how accurate their surface is, or what the reflectivity of the polished aluminum is.
It seems like just a stamped piece of metal, which would be far less precise than electroformed optics.
You get what you pay for, right?
If youâre not trying to get amazingly high lux, just collect all the flux from the LED into a beam, then Iâm sure it will work fine for you.
It will! There are now beamshots from vin which show the difference nicely:
Thrunite Tn42 with Osram Black Flat
Thrunite Tn42 with Luminus CFT-90 (only 24A I think...)
One thing I notice on your beamshots is that the beam looks a bit "fuzzy" for such a large reflector. Maybe this is caused by the spcial orentation of the LED? It doesn't look optimal to me.
The outer diameters of the reflector not only reflect fewer lumens, but there is a larger area too, which causes the intensity of the beam to decrease towards the outsides of the beam.
I think thatâs what you mean by the fuzzy look.
A forward facing reflector kinda does the opposite, as you go to the outer edges of the reflector you approach ~45 degree angle which is where the most lumens are emitted by the LED.
Same for a lens, much sharper looking beam.
Even though the beam isnât as bright towards the edges of my reflector it should still contribute as much to the lux as all other areas, assuming it is focused perfectly, which currently it is not 
Thanks for the comparison Driver!
PERFORMANCE RESULTS
12am, I just got home from doing some focusing and testing with my friend, and here are the results:
98 x10 lux @ 100.6m distance
Lux measured with Dr. Meter LX1330B luxmeter.
Distance measured using UNI-T UT393B laser meter.
Thatâs 980 lux at ~100m, so 980 * 100^2 = 9 800 000 candela
6.26km ANSI
Took close to an hour of just turning screws and measuring different parts of the spot since the focus is so finicky, and 9.8Mcd is the best we were able to get.
Luckily I have friends that are willing to stand outside for hours reading out numbers to me over the phone
Itâs not close to the 13Mcd from the theoretical calculations, but Iâm very pleased with this result, much better than 7.5Mcd.
I think that 9.8 Mcd is pretty close to 13 Mcd considering that with such a small light source you are probably testing the positioning and reflector quality to its limits. The LX1330B (which does not have a great optical filter and is likely calibrated with a 2700K tungsten source) does over-read cool white leds so the actual number may be a bit lower.
The 1330B tests by the person over on TLF were very close to their high end luxmeter.
So either the 1330B you tested is a cheap clone, or there are discrepancies between high end luxmeters too.
Unfortunately with TLF down I canât check what model the other personâs professional luxmeter was.
PS- Do you know any place to find professional luxmeters at a reasonable price (under $500?)
I know Iâm not using the most professional accurate measuring equipment but the LX1330B was by far the best budget luxmeter for the price back when I got it.
Assuming worst case scenario, a 10-20% overestimate like in your tests, thatâs still about 8Mcd
That TLF test of the LX1330B was only done with low CRI cool and neutral flashlights (and an incan source), my test with a wider variety of light sources showed the deviations better. A 3000K 90CRI XM-L2 source was measured 3% low by the LX1330B, A 6000K(?) XM-L2 (from SWM D40A) was measured 9% high by the LX1330B. The comparison was a Mobilux (DIN-) class A luxmeter which is the same luxmeter as the Gossen meter in the TLF test but with a higher-classed (A instead of B) sensor unit.
I have no real experience-based recommendation for a quality luxmeter. Because my financial sources for the hobby are limited, I do not have an extensive collection of high end luxmeters to test. I was lucky to purchase my Mobilux meter second-hand for 425 euro (hardly used), and I own a 160 dollar (taiwanese) Tasi meter that is also in the test in the link above, that one is not too bad. But I would love to do some comparitive testing with a calibrated Extech meter (because many people have one and it is high regarded in the US), especially to see how the optical filter is behaving (I suspect it is ok but not great), and also a good Konica-Minolta luxmeter.
Searching the internet for luxmeter tests is useles, they are not there: what I do with my hobby equipment at home is already way more extensive than anything found on the internet.
Oh I see, that makes sense.
Iâll see if I can find something used, would rather pay $500 for a class A luxmeter than $200 for a tasi which I might want to replace later on.
I wonder why the high CRI 6500k led was so much worse than the 70 cri 6500k on the 1330âŚ
I find it pretty complicated to relate the reading error of luxmeters to the combination of the led output spectrum and the deviation of the sensitivity curve (it is interpreting 3 curves at the same time, of which the led spectrum has multiple peaks :confounded: ). In this case I think the extra reading error of the high CRI 6500K led is caused by the high red content, the LX13130B not just over-reads the blues, it also reads everything above 620nm pretty high.
I have the LX1330B as well, I find it reads low, by alot. Comparing various flashlights and known Lux values, my particular meter reads at least 15% low.
Interesting.
Even worst case scenario, 20% overestimate, still the brightest LED searchlight in the world
I canât find any high end luxmeters with A class sensors for a good price, I might just buy a TASI since itâs on sale: https://www.aliexpress.com/store/product/TASI-632A-Digital-Light-Meter-USB-DC-Power-Luxmeter-Meters-Digital-LCD-Backlight-illuminometer-Storage-128/2797144_32791440073.html
You have to remember that you didnât âbinâ your LED. It might simply not have the luminance of the best performing Black Flatâs.