Luxeons & Reflectors - Focusability??

I just began running "Luxeon reflector trials" too. I'm using a 5W blue HD driven by a 6AA battery pack. Using 5 moderatly used alkaline batteries (7.5v) yeilds 650mA, and 6 cells (9v) yeilds 1.1A The test reflector is from a D cell Mag.

What I've found, it that I can acheive the same super-bright hotspot with a tightly focused beam.

The problem is that I cannot retain the same quality of beam when I attempt to open the beam towards flood. I get "black holes" and other undesirable beam artifacts.. Has this been noticed by others using reflectors?? I'm a bit disappointed that I cannot get as nice of a flood beam using a reflector as I can using a collimator. Am I doing somthing wrong here or is this what I should expect?

On another note, I really like the beam I got when using a reflector in conjunction with a collimator. Now I just have to find a way to mount the collimator (without blocking light) while the reflector is in place..

Hey Slick,

That is a good idea, using the Collimator AND reflector. Hmmm, which to put in front of which though? Have you tested them both ways? It may present a chicken before the egg question.

Originally posted by Slick:

The problem is that I cannot retain the same quality of beam when I attempt to open the beam towards flood. I get "black holes" and other undesirable beam artifacts.. Has this been noticed by others using reflectors?? I'm a bit disappointed that I cannot get as nice of a flood beam using a reflector as I can using a collimator. Am I doing somthing wrong here or is this what I should expect?

Click to expand...

<font size="2" face="Verdana, Arial">It has been my experience also that a reflector can produce a good tight spot, or a good wide spot, but not both. If you want a good wide spot only, you might want to try a "MR" style lamp reflector that is designed to produce a wider beam. These are a PITA to work with though. These were recently discussed in the following thread. You can go to a lamp Manufacturer site like GE, Philips, etc and search on Halogen MR and check out the beam angles for the various models.
http://www.candlepowerforums.com/cgi-bin/ultimatebb.cgi?ubb=get_topic;f=14;t=000736

here's an idea...side-emitting optics

seriously though, might prove kind of interesting, and it might even work in some applications.

electro-lumens, could you have these made?

Slick,

This is jumping the gun here a bit but I designed a simple "drop in" reflector for flood applications. It has straight, angled, walls so it just gets the light from the LED out of the front. The exterior dimenshions are such that it should replace any collimator in an application or flashlight where the collimator can be removed. I just received the production pieces and it is likely that Dat2zip's sandwich shop will be selling these next week sometime. Unfortunately, I couldn't afford to have a bizillion of them made so they will be more expensive than the collimators.

A picture of an early prototype can be seen at:

reflector

In the dase of a RB 5W, I found I needed to turn the reflector and apply some fine scotch brite type material in it to give it a satin finish. Otherwise the reflection was a bit too true and some artifacts were present

BTW, the black hole is always lurking and waiting to rear its ugly head in the parabolic reflectors. You are not alone.

The only way a parabolic reflector could be really adjustable would be to have the parabolic surface made out of a flexible membrane who's shape was dictated by some adjustable frame work. Otherwise, you have a certain lattitued in un focasability before black holes or other unacceptable artifacts show up. A curved mirror in front of the LED bouncing light back at the parabolic reflector might have some adjustability but there are a lot of flashlights out there and we haven't seen anything earth shattering yet. I would bet there are some elegant reflector/ lens combinations that could be designed but the cost of efficient optics probably precludes this from becoming a reality.

- Don

Charlie - My test emitter is mounted facing upwards. I put the reflector on first, then just sat the collimator over the emitter. It seemed brighter AND smoother than either one by itself. I'll be experimenting with this idea more..

Doug - Your comments pretty much confirm my observation that reflectors have an area of "optimal focus" and every other setting is a compromise.. Thanks for the link, I need to experiment with other types of reflectors. I have a set of diamond cutoff wheels for my Dremel that I could use on one of these reflectors. I actually save one from a film project bulb, just gotta find it.

Roth - SE's weren't available when I made my last order but ARE in my current order... Others here have gotten outstanding results using HD's (with reflectors), so using the HD's can't be ruled out just yet. I DO expect the SE to disperse light differently, so they should work better.

Don - Thanks for the interesting ideas. I like the ideas of lens/reflector combos. I also wonder if there is any advantage to be had by "metal finishing" the exterior surfaces of a collimator? To me that would seem help keep light from escaping anywhere but the front. I too have a 5W RB to build and am looking forwards to it.

****************

To sum it up, I guess that I had too high of expectations for focusability while retaining a flawless beam...

Now I'm off to drop a big acrylic ball in the middle of the reflector to see how that does. I'm thing that maybe using the reflector as a "mount" for the sphere will give me the ability to adjust it for the smooth focusability I want..

EDIT - the acrylic sphere didn't produce good results..

Hi Guys,

For starters I think any reflector after a collimator would have no effect. There is insignificant side spill from the collimator to angle out and hit the reflector. The stock optics are depending on TIR (total internal reflection) to send most of the light out the front end. The fact that you can see light from the side lets you know that the T is not total. As to what happens to that light once it comes out the side of the collimator, if the collomator was painted white or silver, this light might be reflected back and add somewhere to the projected beam. Otherwise, it will likely get absorbed in the non reflective surface of the interior of the flashlight. In my first 5W mod for Wayne Y., I found I needed to lift the collimator a bit to unfocus the dreaded grill image of the die. There was obviously significant light coming out the side of the emitter below the bottom lip of the collimator. This light for sure would have been wasted if I hadn't made a small reflector/ cup that bridged the void between the dhoulder of the LED and overlapped the bottom of the collimator. In this way, light was at least directed at the collimator. Whether any of this light hit the collimator below the critical angle which would allow the light to go into the collimator, I don't know. It seemed like it was worth a shot. It would have been nice to have gotten the integrating sphere from Ebay. Perhaps I could have been able to do befor and after tests to see if additional light was in fact getting out................

- Don

I have found the best way to get a collimated beam - or any angle of beam you want - (on paper anyway) is first to study the datasheets (as Doug S has in another thread). The info needed for the ideal curvatures for any reflector you could need is indeed in the very output pattern signatures (ie. batwing/lambertian/SE) given by Lumileds.

Taking this info,you will need to perform some really simple trig calculations to determine the angle of the reflecting plane relative to the angle of incidence from the lightsource and the desired beam exit angle. It *is* just slightly tedious to do so, but due to the distribution of the light coming out luxeons like the common batwing and the increasingly interesting SE, its actually more rewarding if you're after a fast result and effective result. The lambertians are a slightly different animal because of their very even light distribution over all viewing angles, as their "lambertian" name implies. You would need a pretty deep and possibly large reflector to work ideally with these guys. I think the SEs and batwings seem to be underperforming because they are generally misunderstood.

The angles of the reflecting plane (or more correctly - planes - because of the compounding effect required by any reflector-based app)is one thing, but the size of the plane counts too and is determined trigonometrically by the plane's distance from the lightsource *and* the distribution pattern. I know it sounds obvious now, but from what I've been seeing, most ppl just chuck a luxeon into a curved shiny surface and hope for the best when it really isn't difficult to calculate this (but I *am* admittedly cheating since I use a CAD/CAM type package). The size of the planes will affect the eventual size of the reflector, its curvature and its effective beam pattern, of course. Since physical space is limited in most apps, you sometimes must pick a range of reflective planes to prioritise over others so that these capture the bulk of the LEDs output at their respective angles of exit from the lightsouce and redirect them forwards. For eg, there is no point putting a large reflecting plane - curved or otherwise -
in the "shadow" of the LEDs output pattern ie. where light output is, say, less than 20% of the total output. You'll wanna put optimised planes right smack where you see all those flux peaks.

I know this seems to be gibberish, but this idea can be seen very easily in a single picture, which I am working on now and will post later.

I am pretty sure professional optical designers would interpret lumileds datasheet info in this similar way too since it is the ONLY practical way to predict how your reflector design will work with any given luxeon, at a given diameter, and depth.

IMHO, its time to end the guesswork, guys!

I have two questions. I was thinking of mounting a 1 watt side emitter in a Brinkmann 3AA. What kind of results can I expect using the Brinkmann reflector, that is what will the beam look like, and where can I buy a 1 watt white side emitter?

Ted T, been there and done that, and Sonic will attest to its crappiness as he was over at my place when I put it together. The SE simply needs a bigger reflector (think Cyclops). No other solution, such as the MiniMag reflector (which was slightly deeper cupped so we tried that as well) or Lumileds' own collimator, worked very well. The only other possible solution might be what Slick is trying out, with a combination of two beam focusing components.

well hotfoot, it sounds like you have the software...make us a reflector that fits a dcell mag and throws extremely far

also, while the brinkmann is up...how do reflectors fare as opposed to the optics...and what is the best led for throw when used with the reflector?

Thanks Mr. Bulk, What emitter with or without optics should I use to get a fairly tight beam, best color, best over all result? Where should I buy it?

Originally posted by Rothrandir:
well hotfoot, it sounds like you have the software...make us a reflector that fits a dcell mag and throws extremely far

also, while the brinkmann is up...how do reflectors fare as opposed to the optics...and what is the best led for throw when used with the reflector?

Click to expand...

<font size="2" face="Verdana, Arial">Rothrandir,

While an almost D-Mag sized reflector (but not for flashlights) is indeed something I am working on, it is the commercial intellectual property of the company I work for - and thus, cannot be shared without lurking competitors benefiting.

I am, however, working with some other noted CPF modders, to produce an NX05 alternative.

What's funny is that highly collimated optics in PMMA (and I'm not talking about the NX05s) have been available for some time now from Fraen Optics - the same folks who designed the NX05. When I posted their URL here at the CPF some months ago, nobody paid any attention. Those seeking a 6-degree beam FWHM can just buy these NX05-compatible optics directly from Fraen. The only reason I'm working on a custom design for my company and the CPFers I was talking about is because the part we want is *not* available from anyone off-the-shelf and we need out parts in solid machined Al.

Nonetheless, when the products are ready, the CPF will definitely hear about them

Originally posted by hotfoot:
What's funny is that highly collimated optics in PMMA (and I'm not talking about the NX05s) have been available for some time now from Fraen Optics - the same folks who designed the NX05. When I posted their URL here at the CPF some months ago, nobody paid any attention.

Click to expand...

<font size="2" face="Verdana, Arial">Hey!, I remember noticing. I looked at the pictures at that site with great interest. I wish some of their other designs were available "off the shelf".

i remember that thread. they had some pretty cool pictures on that site.

hotfoot, i am very excited now...why couldn't you wait until they were ready

also, when you have them made, would it be possible to get rid of "the evil tab of annoyance"?

also, could you please tell us more about these highly collimated optics from fraen? and how much they cost and where can we purchase them?