LED Collimation

[kiyoukan]

Please have patience with me as this is normally not my area of expertise, i am much better in a computer and useless with out it.
I am looking to focus:
(3W PROLIGHT HIGH POWER RGB FULL COLOR LED 120 LM
Viewing Angle:120 degree, the led is self (not the board its attached to) is Id r2.95mm, OD r4mm)
Into a 0.25in beam from 12ft away.
With as few lenses as possible and cost ratio.
A bit of loss is to be expected but no more that 15% power can be lost. 20% would be absolutely pushing it.
the closest thing i could find was

http://www.luxeonstar.com/l2-3°-spot-base-module-for-luxeon-k2-leds-p-318.php

and this one:

http://www.luxeonstar.com/polymer-optics-concentrator-lens-with-holder-fits-rebel-leds-p-424.php

This one has the size i need of 6mm but would need a second lens to stop the beam from diverging.

 

[MacGyverS2000]

Why is a computer engineer working on something as potentially complicated as an optical system, then?

BTW, it's waist, not waste...

Dan - Owner

http://www.Hi-TecDesigns.com

 

[kiyoukan]

I am programing up a display system, using high speed galvos and lasers. I was able to learn enough to get them running but the leds i need to have a way of focusing them. i have a few lenses but they are in use as the lasers need them and i am unable to get them out at this point.
I am starting to go my budget due to frying a few lasers and a small programing mistake, and for the sake of my sanity i was hoping someone might help me with the final optics for the leds.
Just point me in the right direction of what i need to buy and align.
At this point I like the optic with a 6mm focus but i don't know if i need a single or double convex lens or a concave lens or gods know what else.
You say its complicated? darn i was hoping i was nearing the end of my frustration.
That image was taken from a .pdf, not my spelling but i will email them a note, thanks for pointing it out.

 

[lumenharold]

Was the RGB something that was laying in a drawer or do you need to have the capability to turn individual colors on/off or up/down? And what about the the 3W? Is that a requirement?

If all you are looking for is a less expensive collimated light source that is more robust (and less expensive when it gets burned up) than a laser diode and driver, then perhaps a lower output white or single color LED would do the trick.

Polymer Optics makes a collimator for the Cree XP-E with a divergence of 2.5 degrees or so. I think they are distributed by Future.

http://www.polymer-optics.co.uk/202 - 30mm 2.5 Deg Reflector Collimator for Cree XP-E LED.pdf

Will that work for you?

Harold
SW2010 SP1.0 OPW2010 SP1.0 Win XP Pro 2002 SP3
Dell 690, Xeon 5160 @3.00GHz, 3.25GB RAM
nVidia Quadro FX4600

http://www.lumenflow.com

 

[MacGyverS2000]

This sounds more like a hobby project than an engineering project for a company...

Dan - Owner

http://www.Hi-TecDesigns.com

 

[kiyoukan]

This project is more of a testing the water, of the technological advances.
Once completed, If i can get all output power to cost ratio under a certain limit, i have a contract to have a run of over 100 units made and more down the road.
I was not supposed to be the creator, just the programmer. However the creator had a change of luck and left the table leaving me screwed.
I was unable to find someone with all or half the knowledge to replace him, or willing to put in the time.
I could be paid for the work i had done and just leave. This would be a terrible loss, or i could somehow finish what he had started and get full rights for the software and unit.
I have been all over trying to search for what second lens i might need.
Its starting to look like know one knows how to calculate the specs out of the second lens and i would be better of spending my time buying an assortment of lenses and just trying them out or know one wants to tell me for a reason unknown to me.
However as stated before i am trying to do research before i go and waste money on things i hope to never use or see after this thing is done.
I have gotten so far and close the last thing left are the leds optics.
If i happen to be in the wrong place might please point me to another place that might be able to answer my questions?

 

[MacGyverS2000]

How clean of a beam do you expect at 12'? You said 1/4", but how clean do you expect the edge to be at that distance? If it needs to be sharp, you may need a mask, adding to the complications. You can't make an LED have the same characteristics as a laser beam just by shining it through optics, so there are shortcomings you may not yet be aware of.

If you expect any more useful help, you'll need to be specific on what you're trying to accomplish (and the budget you have to work with).

Dan - Owner

http://www.Hi-TecDesigns.com

 

[kiyoukan]

A pic i drew up to help me, thought it could help other understand what i am thinking.
As this is only being used for a display the edges are not expected to be crisp. if a dot at 12' size of 1-2.5" then had a bleed or splash of about 1" maybe even 1.5" on top of that it would be fine.
Other than that i am unsure how to get more specific.
I have changed my orginal goal of .25 as i think that is unrealistic to the cost.
after talking it out 2.5" is the new max for 12'

 

[lumenharold]

I went to the Prolight site and I'm a little unsure of which LED you are using. There is a 3W p/n PG1M-1LFP-3 but it is three individual LEDs in a 24mm circle; there is also a .3W PPGN-TFFE (note the order of magnitude difference in output). It would be beneficial to know which LED you are using for prototyping and if that is what you will be using long term in order to make a rough model.

Harold
SW2010 SP1.0 OPW2010 SP1.0 Win XP Pro 2002 SP3
Dell 690, Xeon 5160 @3.00GHz, 3.25GB RAM
nVidia Quadro FX4600

http://www.lumenflow.com

 

[kiyoukan]

PG1N-3LFE

 

[lumenharold]

Ah, missed that one.

I played around with this a little as time would allow and there are a few things that really get in the way of my resolving your question. The most important would be as Dan asked, what are you trying to accomplish? I know you want a tight spot at 12', but knowing more about the application would help.

The other issues I ran into that could be addressed with more time are that the LED you selected does not have downloadable IES files available (meaning I have to model it which I did), the optic does not have a file availble (I'd have to approximate that which I did), the ProLight LED is phyically different enough from the Luxeon that the optic will not work as it is drawn (close but not quite), and the LED actually has a half angle of 140 degrees and an included angle of 160 degrees which will make a difference as compared to your expected 120 degrees.

I was having fun but I just don't have that much spare time. If this is really a significant project and opportunity for you I would suggest you track down a lighting/optical designer and spend the money to do it right. The Photonics Spectra Guide should have a ton of references for a start.

Best of Luck.

Harold
SW2010 SP1.0 OPW2010 SP1.0 Win XP Pro 2002 SP3
Dell 690, Xeon 5160 @3.00GHz, 3.25GB RAM
nVidia Quadro FX4600

http://www.lumenflow.com

 

[MacGyverS2000]

Harold,

If I'm ever back in Indiana territory, I'll be sure to take a side-trip to MI to say 'hi'... I'd be curious to see your lab up close and personal.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[lumenharold]

That would be great Dan. It is very "humble" to say the least but a lot of great toys. Bonus: there is a micro-brewery upstairs. Booyah!

If you come back make it in the summer. Florida winters are nothing like West Michigan. We are right in the lake effect snow path (previous two winters were 100" plus of snow).

Harold
SW2010 SP1.0 OPW2010 SP1.0 Win XP Pro 2002 SP3
Dell 690, Xeon 5160 @3.00GHz, 3.25GB RAM
nVidia Quadro FX4600

http://www.lumenflow.com

 

[kiyoukan]

Wow interesting stuff well i found someone who has some lenses from Edmund optics who is willing to mail them to me only for the postage then send back when i am done so i am going to see what i can accomplish.
The one nice thing about Michigan is that there are a ton of machine shops to go to with tons of machines not being used, bad for them but great for me.

 

[lumenharold]

Which side of the state are you on? Or which side of the bridge?

Harold
SW2010 SP1.0 OPW2010 SP1.0 Win XP Pro 2002 SP3
Dell 690, Xeon 5160 @3.00GHz, 3.25GB RAM
nVidia Quadro FX4600

http://www.lumenflow.com

 

[IRstuff]

Note that 2.4 inches at 12 ft is just under 1 degree. The POL lens, when matched to its Cree LED is actually only nominally 5 degrees at FWHM. Interestingly, the Intensity Distribution graph in the POL datasheet is lopsided; it's 50% at 2.5° on one side, but ~63% at 2.5° on the other side. The POL optics are actually Cassegrains, and not a simple optic. You probably need to talk to POL about a custom optic.

I didn't see anything comparable on the Edmund Optics website.

TTFN

FAQ731-376

 

[kiyoukan]

i am on the south east side, about 20 min out from Detroit.
Oh and i am sorry i am making a lighting system, my programing was for advance dmx controlling and ilda support along with a few custom built features. if all goes well this will be a all in one lighting unit it will be able to to use lasers, leds, lumina wheels diffraction gratings. it is meant as an entry level all in one stage unit.