The Sceptre - PScal's quest for a superthrower

You should calculate if this lens actually makes sense with the emission angle of the Wavien collar. If the focal length is too short th lens wont be used completely which leads to less throw.

I think this is the lens used in the Mjölnir from Vinz.

Absolutely. The dx lens is right at the sweet spot for focal length and the collar.

You’re right, the F number is too low for us with a collar.
The DX lens looks good enough.

I did a quick experiment without the collar

BLF GT 1610 lux at 80ft
The Sceptre with Synios (no collar) 1170 lux at 80 ft
The Sceptre with XP-G2 S4 2B (old style) (no collar) 1200 lux at 80 ft

So, right now the xp-g2 is winning over the synios. I’ll build up a module with a Flat Black and give it a shot, too.

Hey @the_driver, could it be possible that the high cd/mm^2 from synios LEDs is because of the white “cup” area around the LED?

Then when we divide the luxmeter reading by the die area we get a higher number despite the die having lower intensity and the reflected light from the sides helping increase the number?
:frowning:
You can see from the beamshots that there is clearly a lot of light coming from the area around the LED:

This doesn’t happen with the black flat because the surface is 100% flat around the LED.

Maybe message koef3 and ask him to redo his intensity measurements after blacking out the area around the die!
That would be really unfortunate if this is the case, and our 300cd/mm^2 LEDs are false readings.

I just had the same thought, Enderman. This method of measuring the cd and dividing by the die area does not work for many LEDs for this reason.

This is really unfortunate.
If pscal is getting lower than the old XP-G2 then the black flat is still highest intensity, and CFT90 is second!
That basically means I wasted like $100+ on LEDs and drivers :frowning:
Hopefully koef3 can retest then and we can find out the true numbers.

At least now we know that for any future tests of LEDs, the area around them definitely needs to be blacked out, especially if they are not flat.
I think it may still be possible that the flat white area around the black flat could be artifically increasing the lux a little, but not anywhere near the amount for the synios.

Yes that is possibe, as I have now learned together with Köf3. It is indeed very unfortunate and I have not found a way to predict this effect berforehand.

The standard way of measurering the luminance has worked for years. :frowning:

I appreciate the effort the community has made to find new brighter LEDs for our projects. Without the members who go above and beyond, none of these impressive lights would perform to the extent they do. Even with a potential measurement error, the Synios performs almost as well as the previous “king” of throw. That is something!

Just yesterday I was actually wondering about this beamshot:

The GT seems to have more throw in the pic, the spot seems more intense.

Yeah. The battery was low! Oops! Even with a new battery, it still didn’t measure as well as the GT.

I really enjoy when folks share these builds, I’m curios what size wavien collar you used.

Thanks! The wavien collar is size small.

Did you buy it from wavien years ago or did someone sell you one?

I think it might be possible to get them from the marinebeam flashlights still, pretty sure they have the small collar in them:Marinebeam Ultra Long Range CREE RLT Illuminator

I bought it from Wavien in 2013. I have eyed that light and may pick one up if funds allow. I’d love to have 2 collars to test different LED modules without worrying about swapping / aiming the collar.

Marinebeam now has the patent.

I re-tooled the pill to use a Flat Black LED, and changed the driver to an H6CC I had lying around. I set the drive level at 4.5 amps and re-installed the collar.

First lux reading at 25 meters

BLF GT 1950 lux (1,218,750 @ 1m)

The Sceptre 2260 lux (1,412,500 @ 1m)

Stoked that I finally out-measured the GT with a 3” light!

The measurements themselves may not be accurate. They still seem high even after a simple back calculation. The thing I am excited about is having a higher measurement relative to the GT.

Nice!
For a 75mm uncoated lens I calculated a luminance of 348cd/mm^2. There is still room for optimization in your build (assuming the DX lens is perfect). 450-500cd/mm^2 should be possible with this LED.

Your measurements seem more accurate now that you’re calculating the candela from the lens :slight_smile:
All GTs are between 1.2 and 1.3Mcd.

Very good performance with the black flat.
I drive mine at 5.5-6 amps but that’s because all my lights have active cooling.
4.5-5 is about the max in a regular flashlight body.

Can you post beamshots with the black flat? :smiley:

Will definitely post beamshots once the weather clears up.

I did some more lux readings. I did 20 meter shots instead of 25 (for 25 meters in my house I have to put the light against the extreme wall of the house, 20 is easier.) This time I put the light on a tripod and walked over to the lux meter while lit, and moved the meter around within the beam. I noticed the “peak hold” feature on the lux meter (which I had used for previous measurements) is more of a “last value” than “peak.” The results I got are a bit unexpected.

BLF GT 2930 lux (back calculated to 1,172,000)

The Sceptre 5800 lux (back calculated to 2,320,000) (is that even possible?)

Not believing the measurement, I tried it a couple of times, The meter consistently read 56xx-57xx lux.

My lux meter is pretty cheap. It’s one of the LX1330B meters. If the reading is accurate, holy moly!

P.S. is there a standard methodology for collecting lux measurements? I’ve not done much measuring before.