Yeah, i think a S MINI sized light (16340) would be awesome.
The Olight is just too expensive for me, but i love the concept: as small as possible.
Not sure how good idea it is really, but…
Ledil Seanna A is an Fresnel + precollimator optic. It has ~150 mm diameter and costs $30 when buying 1, closer to $20 at volume.
If we ripped the lenses, removing enclosure it may get down to $20/piece. Being Ledil, it should work much better than cheap Fresnels.
Now, this is an interesting base for a big zoomie. I’ll call it Z. 
Initial sketch:
BLF GT-like shape (it’s rough, but the major dimensions are correct) on top.
Seanna without mounting screws on bottom left, Z zoomed out bottom middle, Z zoomed in bottom right.
7p 21700 battery pack. You can use any number of batteries, but the light will surely have balance issues with too few (and may still have with full set 
).
The pack should be quickly-replaceable.
With a boost driver it wouldn’t fall out of regulation too soon.
XHP35 LED. Z would be a really nice base for Luminus CFT-90, but:
- probably wouldn’t sustain it at full output without active cooling
- the LED costs much more
- the driver would cost more too
- overall it would easily double the light price to get a nicer hotspot and somewhat better performance.
Note: I draw it here with no glass lens in front of Fresnel, just like Ledil does. Bezel size is the same (quite small) too.
Overall, similar performance to GT; should be better, but not much.
Similar price?
Otherwise - a completely different light.
Nice idea, but the costs will be very high (probably >=400:money_mouth_face:.
The CFT-90 will stomp the XHP-35HI, but the energy demand will require lots batteries.
I estimated the costs to be well under $200…but I am totally green here. Could you explain why does it have to be over twice as costly as GT?
The CFT90 alone costs over $100…
Oh, you meant the variant with this LED. Yes, very expensive…
Any chance to see a reflective collar in a BLF flashlight?
I see that Waiven held a patent on it. It’s so generic that I don’t think it can be sensibly worked around.
From what I’ve read, Waiven stopped communicating with the world. So:
- there’s no way to buy collars from them
- there’s no way to license the patents from them
- they may not even exist and the patents may have changed hands
- the patent won’t expire for long (from memory 2028).
So I have 3 questions:
- Is the above assessment correct or have I missed something?
- The patent system is broken so much that it doesn’t let us find the current patent owner, does it?
- Is it possible to do without a patent license? Is there any real risk for a Chinese manufacturer? If yes, any chance of finding a manufacturer who won’t be scared off?
What is it your talking about? A reflective collar? I’ve never heard of this.
Wavien collar, search BLF Jason
Agro, ooh patents are such a difficult topic idk if it is worth burning our fingers on.
JasonWW, search on “wavien” at BLF and elsewhere.
Lemme try, I don’t know much: the idea is to surround an emitter with a mirror that sends some otherwise lost photons back to further energize the phosphor.
If anyone can find prior art for the general idea, that would be good to have publicized.
Okay, I’ve seen those before. That seems useful only for an extreme thrower. We already have the GT.
1. That is useful for any thrower regardless of size. What I find the most appealing is zoomie that fits in pants pocket, but puts 800 lm in a 500 kcd hotspot. With SST-40. (Though I have some unsolved problems with it, that’s why you haven’t seen me mentioning it here).
2. Frankly, I think that flat LED itself is a prior art (assuming it’s earlier than the patent). The flat silicon-air boundary reflects more of side-facing light back to the die, further energizing it. But even if it is indeed prior art or we find another one, patent invalidation is very hard and costly.
I have made a 3d design for someone to experiment en test with it.
So the idea is out there. 
From what I’ve seen, patent lawsuits generally fall in one of two categories: settle for money (ideally lots from a large company but a little from many small business is just as well for patent trolls) or block your competition from delivering their product (megacorp patent war).
So I’d say the risk for small scale Chinese manufacturer of flashlights is very near zero. If they were selling, say, millions of dollars of street lighting equipment, then someone might try to block import of their product. For a few flashlights sold directly to customers? Good luck trying to do anything about it.
When it come to copyrights and patents, China is like the Wild Wild West. That’s why they copy and clone pretty much everything, including iPhones.
There is a light that I’ve been dreaming of for a while:
- zoomie
- good throw, OK (or better) flood
- big-edc size (think B158 or a little bigger). 45-50 mm lens, 21700 battery.
- e-switch
I spent quite a lot of time thinking through several obstacles. Today I made the fist sketch. It uncovered other ones.
First, some numbers. A focused zoomie with a lens that has f-number of 1 collects only 20% of light. All the rest hits the flashlight body.
I wanted to improve it and I evaluated several options.
#. Mirror adds rings to the beam around the emitter image.
#. TIR lens should be similar
#. Precollimator lens increases the projected emitter size without significantly affecting throw.
It also greatly narrows the flood beam. Maybe it should be defocused together with the main lens to get rid of that effect…
It’s relatively cheap. It can recycle nearly all of the lost light.
#. Waiven collar. It makes the image of the same size, but brighter. It’s less good at improving efficiency, can recycle only about a third of collected light. The rest is wasted either by not being reflected (as far as I understand it reflects only the most promising frequency) or not re-emitted (and probably ending up as heat). It also reduces beam width, though slightly less than a precollimator.
#. 2-stage collar. A part of the collar is mounted on LED PCB and a part near the lens. It should collect slightly more light than a regular one, but the main advantage is that it has smaller effect on the flood beam. Since the near-led part doesn’t have to collect a lot of light, it can have wider opening and thus:
- block less light from the lens
- be slightly lower, allowing deeper de-focus
2-stage collar acts funny when zooming. It comes out of focus very quickly, making the beam lose a lot of intensity in the initial zooming stage and much less later.
2-stage collar seems like the best option as it allows a fairly wide beam and extreme throw distance in the same light. Though the legal challenges may make it infeasible. Legalities aside, it’s going to be costlier too. For the budget we can have either a good thrower or a good flooder. Or neither. But I see no way to do both (though maybe some precollimator defocusing would help).
When it comes to the emitter / driver I see 3 good options:
1.Dedomed SST-40 with buck-boost driver to keep it exactly at the current that we want. FET won’t work because 21700-30T will blow the LED.
Very high throw, good efficiency, the highest price.
2. XHP35 HI with a boost driver. Less throw, less price, easier to manufacture, better efficiency, better flood. But throw is much worse and there’s the dark cross in the middle of the focus beam.
3. XP-L HI. Oldie, but still a good option. Paired with FET+1 for a good price.
Now, my drawing:
For the purpose of drawing, I used a 50 mm lens. I guess a 45 mm would be better really. And I certainly made some parts too thin. And tolerances too tight. It’s a sketch, you know. Also, the drawing uses a 2-stage collar where the moving part is Fresnel style. Being Fresnel style is not important, I just liked to draw it like that.
Problems:
- I failed to account for the threads for battery insertion. Easy fix.
- There is no switch. And the only sensible way to put an e-switch that I see is to FW3A it.
- Cooling is bad. I mean really, really bad. And I see no way to fix it.
At the end, I’s also point out one potential obstacle. Collar needs to be positioned very precisely. Will various abuses that EDC light will endure make the collar lose focus over time?
Thoughts?
For zoom to flood, you can’t really beat led lenser’s optics.
Can’t use a collar with it though, but it does collect 100% of the LED light when zoomed, which is the main problem with normal zoomies.
It is half TIR and half aspheric.
A precollimator makes the beam diverge more and create a bigger spot, it does not make it narrower.
A wavien collar does not change the beam divergence, it slightly more than doubles the lux and lumen output, increasing the efficiency from ~25% to ~50%.
(assuming the lens collects all 60 degrees of light coming out of the collar)
Two stage collar is an interesting idea, that would give you basically two different lux measurements and spot sizes.
Only problem is that it would cost at least several hundred dollars to get the small collar made, and probably ~1000 for the secondary one.
Agro:
For better heat management I think you need to make the tube much thicker where it contacts the zooming part. IMHO this is the only option to make it better. Thick tube like 5-6mm walls and bigger diameter for zoom threads. The light’s thickest point still will be the head so I think it is acceptable.
Or to make it even better The tube can have 3mm holes in almost full length and we can insert 6-7 3mm thick round heat pipes in all length and those will send the heat to the tailcap part. Also at the tailcap it can be some fins to help the heatpipes.
Yes…it’s an interesting optic. But I’m yet to see a light with good throw (per size) and this type of optic.
Even if it can’t compete with collar, it still may be a great option.
The way I understand is:
Before the main lens, precollimator narrows the beam significantly. After the main lens it becomes wider.
That’s in focus mode.
In flood mode, the narrowing means that light hits only a small part of the main lens, creating a narrower beam. I actually did some basic ray tracing of this and that’s where it made the beam narrower than blocking the light directly (with collar) would.
I guess you mean a custom collar made from glass and custom coating tuned to reflect the desired frequency, right?
A regular aluminium reflector already reflects ~90% of light. May lose some more by being less smooth. But the price in volume shouldn’t be a problem.
Being tuned to the light will help recoup some of the loss. If it captures more light than the stock collar would, it’s likely to be a win.
Thanks for the suggestion. I’ll look into it.