The Sceptre - PScal's quest for a superthrower

In 2014, I picked up a chrome Elephant host and turbo head. I built it up with an xp-g2 s4 2b, taskLED H6cc at 5 amps, 3 series battery holder, 78mm DX aspheric, and a wavien collar. I nicknamed this light The Sceptre due to its royal appearance.

For larger photos go here

BLF GT for reference

This light was good for about 1m lux. I never got what I would consider to be an accurate lux reading. It went back and forth beating and losing to the GT, but it appeared brighter to the naked eye.

Around the same time, I read that a worthy LED upgrade has finally come along. Time for an overhaul.

Parts purchased for upgrade:

  1. Synios LED (Best of 10)
  2. 75mm Optolife lens (A298)
  3. Mag D Heatsink slug
  4. Noctigon 32mm XP mcpcb
  5. Convoy Biscotti driver (8x 7135 but I removed 2 to make it 6x 7135)

Design constraints:

  1. MCPCB must be electrically isolated from the host
  2. Wavien collar must be re-aimed for the Synios LED
  3. The LED module shall be capable of being inserted and removed from the Mag intact. (In my previous build, the collar was glued on after the slug was secured with a set screw.)

The Build:

MCPCB mods and LED mounting:

The MCPCB was drilled to allow access to the set screw hole in the heatsink slug. I also jumped the thermal pad to a nearby solder pad to solder the cathode connection.

Arctic Alumina epoxy was used to electrically isolate the MCPCB from the heatsink slug. Since both the MCPCB and the heatsink had holes drilled into them by me, careful deburring was required.

Old pill vs new pill

Wavien collar aiming:

I scratched my head about this one for quite some time. Once the parts arrived, I used a stack of post it notes to roughly aim the collar. Then, I found a piece of 1.5mm thick prototype board. I sanded it down to the perfect height, cut it into a circle, drilled a hole in it, cut relief for the solder pads and set screw hole, and sharpied it black. I made the circle slightly larger radius than the wavien collar but still small enough to fit inside the mag so that I could hot glue the collar to this piece directly. This piece is the one I spent the most time and thought on.

Heatsink prep:

I drilled a small hole for the wires, then I drilled and tapped a set screw hole aimed such that it would be accessible with the wavien collar on the light.

The driver is simply epoxied to the back with arctic alumina.

Speaking of the driver, it needs to be powered by 4v instead of 12v. I happen to have the 3p battery holder for this host also, so it was taken out of the parts bin and put into use.

I used 22ga teflon insulated wire due to its small diameter. I did not want the wires to touch the collar - only the custom spacer. I used 18ga silicone wire between the mag switch and the Biscotti driver, and made the wire run extra long for ease of installation.

Lens upgrade - or not:

The Optolife lens fits beautifully in the FM3X head. Unfortunately its focal length is a bit short and it does not quite focus the die as well as the DX 78mm lens. I plan to do testing on both lenses with a lux meter fairly soon. For now, visually, the DX 78mm seemed better so I am using it.

The light went together great. I am thrilled with how it came out. I like how the Synios LED uses 1/3 the power of the XP-G2 and does not require a buck driver.




Here are 2 shots from its previous life with xp-g2 s4 2b. I know they are crap. I’ll get better ones with the new light and my BLF GT.

Thanks for reading!

Got some quick shots on my patio this evening.

ISO400, f/8, 1 second exposure.

The Sceptre with Synios LED



Foggy with Synios LED (2sec exposure)

Beamshots with Flat Black (had to adjust exposure to 2s due to atmospheric conditions, but they still turned out darker than they looked on the camera)


The Sceptre


Shooting the Moon

First lux measurements:

The Sceptre (with depleted batteries)

Initial lux reading at 80 feet was 1,460 lux with a 10” spot. The aperture of this light is 3”.


Lux reading at 80 feet was 1,640 lux with a 30” spot (hard for me to measure). The aperature of this light is 4.75”.

Round 2 measurements with Oslon Flat Black

at 25 meters (using the “peak hold” function

The Sceptre
2260 lux

1950 lux

At 20 meters (observing the lux meter in real time)

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

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

That is looking great, and I’m looking forward to the result! I have a few of these leds and plan some modding with them as well.

Wavien collar were the groundbreaking tech to double intensity quite easily and it stopped… such a shame.
Anyway, really nice aspheric light you got there !

Nice! Thanks for the write up.

Excellent aspheric build, I will have to add it to the list of most power flashlights over 1Mcd :slight_smile:

Very nice build!

Nice work on the upgrade. :+1:

Very impressive light! So, are you happy with the new emitter? Higher beam intensity but a smaller beam size.

I think your measurement is high (relative to other peoples’ measurements) not because of your lux meter but because of the method you used to calculate the candela. While the method does make some sense, I’m not sure it is correct. See this thread for discussion: The proper way of calculating lux / cd / candlepower

Thanks for the info about measuring lux. I’ll read up on that. I do like the new emitter. This light is all about lux, so the smaller but more intense beam is perfect. I also like the lower power consumption / heat output and longer battery life that this LED offers.

Yeah I agree, see this post specifically: The proper way of calculating lux / cd / candlepower - #35 by EasyB
I made that thread to explain the concept that I read about on CPF, but after some further research it seems to be incorrect.

The light coming out of a lens is not a single cone of light that diverges from one point, it is an infinite amount of light cones coming from the surface of the lens.
See image 5: Light collimation with an LED flashlight
Based on this, the total lux of the flashlight is simply the sum of all these cones of light.
Since each cone behaves according to 1/d^2, then the sum of all cones also behaves proportional to 1/d^2

So in fact the light converges at the lens, not behind it.

Very nice! I like that you can actually see the shape of the LEDs package in the beam. 1.77Mcd is great!

You could probably still optimize it a bit though. With 75mm lens diameter with no ar-coating and this LED it should be possible to get up to 2.8Mcd with a 315cd/mm^2 LED and perfect placement of the collar. The collar can increase the luminance of an XP-G2 LED by up to 120%. This effect might actually be greater with an even smaller LED.

I would think about modifiying the light so you can use the more high-quality lens. It should make a difference with such a tiny LED because of its more accurate shape.

I wish those wavien collars were more readily available. Pretty sweet build you got there

The initial lux readings are probably inaccurate. I will work on getting better long distance readings.

Due to the easy access I designed into the light, it is definitely possible to spend time optimizing focus of the collar. I will spend more time on that and rig a lux meter to give me real time info while focusing.

The optolife lens was a disappointment. I will take photos to show the difference. That should explain everything. At this point I plan to stick with the dx lens. Lenses get so expensive so fast. Also, part of what I enjoy about this light is the beauty of the elephant host. I don’t want to change the looks to fit a different lens.

Also, keep in mind that I am not driving the LED at 2.5 amps. With 6x 7135 chips, the LED sees 2.28 2.1 amps. I may upgrade the driver at a later time, but I am very close to the top of the curve and diminishing returns are in effect.

Thanks to everyone for the encouraging words and tips on furthering the performance of this light.

I spent some time with the lux meter and the wavien collar. I observed something unexpected due to the asymmetrical shape of the LED. I get the highest lux reading when the rectangle is perfectly centered over the rectangle. However, the die is not in perfect orientation over itself. My thoughts are that I should focus on the die itself and not the entire package / highest lux reading. I’d like to hear others’ thoughts, too. As an experiment, I can glue the collar in place both ways and take measurements, but my free time is a bit limited and I think some discussion could be interesting.

I also realized that my batteries were almost dead, so I charged them up. That should help a bit with lux. :person_facepalming: I’m going to officially scratch the initial lux readings until things are worked out.

Here are photos to illustrate the difference between the Optolife and the DX lens. The optolife lens scatters light. These photos are taken with the same camera settings, stepped down so as not to overexpose. In person, the scattering is much more pronounced.



Wow that is bad.
My optolife lenses had almost none.
What is the product code of the one you bought?
Are you sure it was plano-aspherical?


Ok well that’s unfortunate.
Must be a defective one, or it’s bad because of how low the F number is.
My A380 an A037 are higher F.

I would recommend this one for people who are looking for a precision 75mm lens, it’s pretty cheap: