Tri-Lobe reflector concept for CREE LEDs

In 2004 I assembled my first concept reflector, the idea of the concept came the year before, and I built one up. It suffered a bit from precise alignment of the CREE LEDs on the post
(I know it definitely looks hobbied, but it shows the concept)

In 2005, I asked Don to machine a three faced post to improve alignment over the ground down bolt I'd used back in 2004.
(Thanks to McGizmo for making the Triad post for me!)

So, what does this thing look like?




Lit up:



Obviously the tri lobes of the reflectors are a bit out of alignment in these pictures.


Anyhow, here are the results, the camera can not even come close to showing the vividness of the color rendering.

A white LED is on the left and the concept tri-lobe or triad is on the right.
Some folks will recognize these pictures from March 2006:


Another view of the enhancement, notice how the reds really kick in on the RGB compared to the white LED:

:thumbsup:

I am glad that Newbie shared this! I made a side mention in another thread that this idea was not new and mentioned Newbie as having such a solution but it was not for me to be more specific. When he contcted me about possibly making a triangular faced pedestal, I knew immediately what he was after but not that it would host R,G,B. I had previously made some rough samples of the side shooter for some small Nichia SMD's which had a viewing angle of 120 degrees and the concept of a tri segmented reflector was an obvious solutition for those LED's (3x120=360) but the thought of making such a package was overwhelming. There was also to be a fair amount of loss due to rear directed light that would not make it out of the reflector.

When I started messing with the narrow viewing angle of the XR-E, I realized it was much better suited to this side shooting scheme and in fact the non lambertian distribution required some new approaches if one were to have access to managing the light output.

Credit where credit is due and to my very limited knowledge, Newbie was the first to construct a 3 shooter. As far as I know, it was a concept that he came upon himself and if there was previous art or examples that influenced him, only he can answer that. I knew where he was coming from when he contacted me as I had also considered such a solution but I had not tried to make such a beast. As I recall, he was hoping to use a single continuous large reflector initially and not segmented sections set as a side shooter focus specific to each image source; something I mentioned to him from my side shooter experience. Perhaps we both influenced each other in the spirit of discovery.

At any rate, I can claim independent thought concluding in the 3 shooter solution but I won't claim to be first or the origin of such an idea.

What came to me as an unexpected surprise was the beam I got from the 3 shooter with the XR-E's. I assumed with a relative long focal length that there would be much less divergence in beam angle than the reality of the case turned out to be. I anticipated a tighter and more concentrated beam. It is clear to me now that a longer focal length resulting in an even larger diameter optic would do a better job of collimating the light and the results might be quite impressive where lux is the goal. With this first sample I did, a lux measurement of 10,000 is nothing to be embarassed about but more impressive is the size of the beam that has this intensity! And of note, unlike with a lens solution, the side shooter does have viable spill light in addition to that which is redirected by the reflector.




Compaired to a stock SF L6:



I suspect of more interest to all of us is whether this idea will find itself in any production examples on any scale. :shrug:

Don said:

I suspect of more interest to all of us is whether this idea will find itself in any production examples on any scale.

Click to expand...

That would be your part of the game, Mr. McGizmo. Get working on it, and we will buy it.
10.000 lux in a spot that size is like a free triple-cheeseburger every day.

bernie

McGizmo, your prototype looks magitudes better than mine. Very nice!

Can you tell I'm not a machinist?

For myself, the concept didn't come from anyone else. It was an evolution of the partial parabolic curve that is commonly utilized in some projectors to reduce optical path size. In that case, they coat a small piece of a parabolic curve, and optically direct the light to hit that piece by design, before it hits the LCOS or whatever, imaging device, which then goes into a lens stage for projecting on a screen.

I'd been pondering, how to combine RGB. I figured, heck, why not just use a half reflector, back around January of 2004, and it worked great with one LED. So it proved the concept. So I set out, hacking up MagLite reflectors into a tri-lobe/triad or whatever you want to call it, and grinding faces on a steel bolt. It worked better than expected, but not really nice, until you made the aluminum triangle bolt for me last fall.

Getting the mechanical and optical locations right in the first place (with the three-faced "bolt" you made me), made helped a lot (thanks!).

Personally, I just have a tremendous amount fun tinkering, if you haven't noticed by now. I do design stuff for a living, professionally, but I enjoy the flashlight hobby to unwind.

So, when can I buy one of your nicely finished creations shown here?

When several good ideas come to together in true synergy, that's a beautiful thing.

Dig that SECRET color.

Dig that crazy pedestal sink.

This was worth logging on for.

McGizmo said:

And of note, unlike with a lens solution, the side shooter does have viable spill light in addition to that which is redirected by the reflector.

Click to expand...

And, unlike a lens solution, the side shooter is wasting tons of photons on spill when the user is trying to spot something at range.

A lens solution can be focused--without a donut hole--to give maximum throw or a wide flood as the user requires. Focus for distance and no light will be wasted on spill when spotting at long range; focus for flood and there's no extremely bright hotspot to glare back in the user's face when working at close range.

NewBie said:

Can you tell I'm not a machinist?

Click to expand...

:laughing: LOL!! Newbie, my first thought as I was skimming the thread was, "My God!! It's the first report of a Cree explosion on CPF!" I was all set to see a thread that would parallel your CR123A testing/venting one.

No one remembers this?

http://candlepowerforums.com/vb/showthread.php?t=40228

Don, what is the diameter of the hydra-reflector? A custom torch would be great, but for those with low budgets, a reflector/heatsink assembly that we could put in maglite heads or something would be awesome as well!

I'd just die for a quad one that would fit a C sized maglite head!!!!!

Don, also, how well does that dual one work? That one is realitivly easy to build, and the thermal relief seems as it would be pretty good. The size is proabably more manageable for smaller torches.

evan, I forgot about that one until you reminded me.

evan9162 said:

No one remembers this?

http://candlepowerforums.com/vb/showthread.php?t=40228

Click to expand...

Absolutely.

hmm.... i see the tri cree heat sink and i was thinking of there is enough heat transfer from the cree to the heatsink and beyond?

that will be a great place for a heatpipe to be in....

wow....i wonder how much technology will be in that light...
1)led
2)drivers
3)optics
4)heatpipe
5)mechanical switches
6)HA(hard anodise)
7)li ion

and some more...who wants to add on to it?

chester, just use a copper heatsink and solder the crees directly on there. I'm sure there would be no problem of transfering heat then. Just got to make sure the heatsink touches the torch well. Add on to the list? 8) user programable interface (low/high/strobe lol)

evan9162,
I don't remember that thread and I am embarassed to admit it! Please forgive me. For all I know, and it seems likely in retrospect, your thread may have inspired my initial sideshooter exploration. It's either that or I elected not to post of my similar experiments for some reason. I would guess it was the former case.

enlightenment,
I would love to see a proper optic assembly designed with multiple elements and allowing for an adjustment of the focal length instead of just defocusing. I have bugged PK for a few years to get something like this going. If you could do a nice job of beam adjustment and with a reasonable range in beam angle then you could get away from a combo spot spill reflector package. To date, aside from a couple Panasonic lights using a Nichia and adjustable optic that left much to be desired, I have not seen anything that would come close to replacing a reflector for me personally.

Talk is cheap. :nana:

evan9162, Newbie, McGizmo. Interesting concept. :twothumbs

Don, beautiful execution of the the reflector and pedestal.

McGizmo said:

I would love to see a proper optic assembly designed with multiple elements and allowing for an adjustment of the focal length instead of just defocusing. I have bugged PK for a few years to get something like this going. If you could do a nice job of beam adjustment and with a reasonable range in beam angle then you could get away from a combo spot spill reflector package.

Click to expand...

(achromats assumed)

While I have enough of an idea about optics to know this is a good start, I'm afraid real optical engineering is out of my depth of field. If anyone happens to have something like ZEMAX at their disposal, though, it shouldn't take too long to hack up something workable.

enLIGHTenment, that looks like it's part of my G-lens

Props to evan9162 and Newbie!

chesterqw said:

that will be a great place for a heatpipe to be in....

Click to expand...

This is the only time I have thought that a heatpipe might have a place in a flashlight. With three or four LEDs with their backs together, you could drive them very hard by sticking a heatpipe in the middle and then running it out into a heatsink-body.

I love the idea of the focused output vs defocusing, Don.. like the megaray i suppose?

I had a similar concept to remove the donut-hole in a LuxV solution.. the post is here That was from aug. '05.. an independent concept from the unknown solution that Newbie had going for different reasons far earlier it seems.

-awr

I vaguely remember ye olde post with a concept sketch showing 3 side facing emitters and a 4th one facing forward. Don't know how to search for it.

Why just 3 emitters if using white LEDs. 4 LEDs on a square tube. 6 LEDs on a hex tube.
Or to simplify things just 2 LEDs back to back.....ummm.....Terralux already did that - Maxstar. Wonder if they will switch to XREs. :sleepy:

here is a similar design i came up with apparently at the end of 2004 for making a PR replacement for big flashlights.

a picture of the 12W version is here: