Wavien collars are SOLD OUT [limited production]

And as Marinebeam elegantly explained, in a standard reflector based torch the on-axis light that doesn’t hit the reflector turns into spill, whereas the hotspot is created by the off-axis light that does hit the reflector.

Meaning if you want a thrower that is practical (with some spill) you might be better looking at extreme reflector designs.

Edit: the deeper the parabolic reflector, the better.

If you want to focus an image of the die with zero spill (wasted ? light), then an aspheric with a collar is the way to go.

Only the blue light that the LED emits to the side, is actually used by the collar to increase the converted part going forward in a meaningful way.

The Marinebeam guy is not the first guy I would go to to get real information regarding this… The light he offers doesn’t need the collar to get it’s performance. Adding the collar makes that light way more expensive than it should be.

Of course. But all of the other light from the phosphor that goes sideways is also “recycled” by simple scattering from the surface of the phosphor. No re-stimulation possible for this.

Which is super inefficient! Most of it will be absorbed and some scattered back into the reflector and then back into the LED… I find it hard to believe that this has a meaningful impact.

You will probably get higher performance by using a collar which only reflects blue light and absorbs the rest of the spectrum. This would keep the phosphor cooler and you might reflect 3-4% more of the blue part.

You could try proving your point by using the collar with a very warm (<3000K) led. I predict a very small gain.

So let someone make a glass collar with a dichroic mirror coating tuned to the LED blue spectrum, then we might see :wink:

Thats what mem (he claims to be the inventor the collar) was doing before he disappeared from the forum. He developed a collar v2.0 :D.

I will add some links to this post…

Here men talks about the RGB-ratio of the collars mirror coating and the tint bin of the LED having an effect on the final performance-gain of the collar.

More talk: Wavien ceased operations - #47 by MEM

Here it gets interesting. He talks about boosting the XP-G2 to a luminance that would equate to 2900 lumens if it were this intense without the collar. The actual maximum of the XP-G2 S4 2B (the good old one) is 6A and 1120lm (see koef3s test ). This means that mems “v2.0” collar achieves up to ~160% gain. Thats about 30% more than the Wavien collar (120% gain). I find that believeable.

Here he mentions that collars made out of stainless steel don’t work very well because this material only reflects around 30% of blue wavelengths.

So he said, but he burned off a few people, who sent money, with his claims (Dale for one) and never delivered.

That doesn’t mean they don’t work though. Everything I have read from him makes a lot of sense.

Either the collar works primarily by re-cycling the blue using re-stimulation, or mostly by bouncing around all the spectrum and scattering it back.

To make a x2 gain just from the blue (what proportion of it is present in the original LED+ phosphor output ?) surprises me, but I can be convinced, by facts and data.

50 of you have bought them, so hopefully we’ll hear some results.

Isn’t BLF great ?

The collar has the same effect that a blue laser has when shined onto phosphor. A blue laser also makes the tint more yellow/green.

These pics both show an Osram Black Flat. It is running at the same current in both pics. In the pic where it’s brighter, a high-powered blue laser is being shined at it. You can see that the tint of the LED becomes much greener.

Now we are back into remote laser-phosphor technology, which works splendidly. With suitable phosphors.

Edit: not just shining a laser at an LED, but with designed remote phosphor pieces (not LEDs). And optics (mainly reflectors) to maximise the output.

Coming to a car to you soon (maybe not in the USA, the regulators are very conservative). But has been lighting up the roads in premium EU cars for a few years now. At a price.

A cool-white LEDs phosphor produces yellow-green light, not orange or reddish light. It will never be truly “warm” like an incan. Only a sickly yellow-green which is warmer than standard cool-white, but not “warm”.

Here is a diagram from sma’s test . He measured the tint before and after. The clear shift of this de-domed XP-G2 towards yellow can be seen. It was probably a 6000K variant before de-doming.
“mit Kragen” means with collar:

I use hot glue but if you want something more permanent maybe use some silicone stuff which is also flexible and shock absorbing.

Yeah it’s some combination of blue light adding excitement to the phosphor and white light bouncing off the die and re-scattering.

Nothing is assembled but just showing how the collar does seem to fit just fine in a Z1 host

Please report back with your results. I tried my collar in two different lights, one a Z1, and both times my “after” lux was about 30% lower than the “before” measurement. Obviously I aligned it incorrectly since I have no idea what I’m doing lol. But the led looked well centered to me before I held the collar in place with hot glue. The beam looked noticeably dimmer after adding the collar, but otherwise similar in profile.

That doesn’t sound promising. I dont have a driver yet for this nor do I think my ability to focus LEDs is any better than yours so hopefully one of the resident experts can chime in with some suggestions for you/us.

Without owning a Z1 I think this is what happens:

The collar needs to match the lens to get an improvement. The Z1 lens has a short focal length and is quite close to the led, therefore the cone of light coming out of the collar does not fill out the whole surface of the Z1 lens, which in fact then behaves like a lens with smaller diameter and thus less throw: even if the collar is focussed well the gain is more than compensated by the effectively smaller lens.

Agreed Djozz.

I think Marinebeam probably have a well matched lens (diameter and focal length) which will not be easily bettered, at best equalled with very similar parts.

To progress beyond just making a clone, I think you will need a larger diameter, longer focal length lens, an altogether bigger torch. And do some fine tuning of the collar with shims.

The maths doesn’t seem difficult, at least to get an approximate idea of what might work.

I think the Collar will only help, if the flashlight is focussed.
Defocussed it will decrese the angle of outcoming light, as Djozz wrote.

But since it’s a zoomie cant he just adjust the length until it reads the highest lux and fix it at that point or at least make a mark on the light? or am I oversimplifying it? The lens really isnt much smaller, only a few mm if that.