Help needed with a LED project

[EmanuelM]

Being new to this forum I hope I place the question in the right place.

I am currently involved in the design of a power LED searchlight. The requirement is 4 lux at 200 meters with a 5-6 deg. beam (+/- 2.5-3 deg.)

While LED and LED systems is what I do for the last 13 years, I never attempted to work on something with such a narrow beam at this intensity.

First I would like to know if somebody has any experience with such optics.

Second, it will be interesting to hear opinions about the single vs. multiple sources to match this intensity.

There are plenty of COB (chip on board) arrays available providing more than enough lumens, however the optical designers hate to work with them as they are far from being the ideal "point source".

From the other point, working with multiple sources each with it's own optics is also considered by some non practical, they claim the total power from several light sources all pointed to the same direction will be much less than the same power emited from a single source !? Is this true?

The smallest emitter I managed to get is 11.5mm diameter with an output of 1330 lumen at 17W. Why is it so hard to design a narrow beam optics?

Any input will be highly appreciated.

 

[MacGyverS2000]

working with multiple sources each with it's own optics is also considered by some non practical, they claim the total power from several light sources all pointed to the same direction will be much less than the same power emited from a single source !? Is this true?

While it may have practicality issues, basic physics says it doesn't matter how many sources offer the power, only that the power adds up to the same total amount.

The practical issue, particularly for such narrow fields of view, is getting the different sources to overlap appropriately. Think of two point sources 2' apart... you can make them both focus well together at 200', but at 5' the same focal direction will not match up the beams at all. If the point sources are only an inch or so apart, this will be somewhat irrelevant over the field of view you're looking at matching.

So, creating the optics itself for such a small angle isn't a problem for s single point source. Doing it for multiple points says you have to watch out for beam separation.

What you can do, if you have the room, is provide separate optics for each point source that send the beam to a combiner optic... form the standpoint of an outside observer, you're now dealing with a single point source. this can actually be a single piece made from cast acrylic/glass if designed correctly, and you can carefully control the placement of each point source in relation to each other and the optic.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[IRstuff]

Fundamentally, it's not possible for multiple sources to behave like a single, more powerful source. Any set of discrete sources sharing the same optic will behave like a source with a larger angular extent. The physical distance between the individual sources translates into an angular spread.

The best you can hope for is to have multiple sources converge onto a single area at a single distance from the optic.

TTFN

FAQ731-376

 

[EmanuelM]

I think that at this distance (200 meters not feet) at such a narrow beam there won't be such a separation. Let's not forget that practically the beams are not 100% sharp.

Pointing all to the desired distance, say 200m makes sense. This is of course a very small adjustment but interesting idea to explore.

There are two other dilemas speaking of the design:

What is preferable - reflector or TIR collimator?
I tested lots of reflectors plastic and metal small and large for different projects - all designed for LED by the way - the results are quite dissapointing both efficency and beam sharpness. The TIR method seems to work better with LED but sill having a 15% average loss.

A thing I am considering is testing a combination of RBG leds instead a common white - I always had the feeling that making the optics for a white led is different then the result with any other color type.

It is like collimating a fluorescent vs. filament. The way white leds are built considerably increases the light emitting surface, as the die is not the emitter, the entire phosphor substrate is. With a color LED the source is much smaller (the die itself).

Is this my feeling or there is an explanation?

 

[MacGyverS2000]

Yes, but with an RGB LED you now have three point sources separated some distance from each other. Again, on your scale it's most likely irrelevant, but if a 3mm diameter emitter of white causes you concern, three separate colored points within a 3mm diameter circle are going to be even more hairy to deal with.

Sounds like you need to more carefully specify what you're willing to give up performance-wise for certain gains...

Dan - Owner

http://www.Hi-TecDesigns.com

 

[EmanuelM]

You are right - At close distances the colors won't mix, there will be 3 clear RGB spots. At distance they might mix well. Anyway, balancing RGB will also be a task and will almost for sure need decreasing some intensity in a device where each lumen should be squeezed well.

I was just describing one of the dilemas I am facing.

About the projector, it is supposed to be a hand-held device as small as possible, preferable 4-5" diameter.
I can use any type of led commercial or custom made.

Having cooling limitations, the best will be to settle at 30-36W providing at the best 3000-3600lm and if absolutely necessary, can somehow get close (with some limitations) to 50W.

I measured an old projector with a 100W filament lamp, quite a large one, 7-8" diameter. It makes the 6 deg beam with 4 lux at 200 meters. The bulb was measured in an integrating sphere and the total flux measured was <2000lm.

It seems like the problem is finding an open mind optical designer. As described in my first post, so far those I meet are only raising the problems expected, none of them attempting to find a solution. It seems LED is scaring them.

 

[MacGyverS2000]

It seems like the problem is finding an open mind optical designer. As described in my first post, so far those I meet are only raising the problems expected, none of them attempting to find a solution. It seems LED is scaring them.

Are you paying them yet? If not, it's in their best interest to point out the large number of failure points to this project. For the right money, I'd take the project on, but I'd also be upfront with you about realistic expectations.

To point out another issue... I hope you plan on having a lot of forced cooling. 30-50W in a hand-held device is no simple matter, and LEDs cook quickly!

Dan - Owner

http://www.Hi-TecDesigns.com

 

[EmanuelM]

I am aware about the heat issues - as I mentioned, LED systems is what I do. I prepared a concept model of this searchlight working with an array of 36x1W Rebels. The heat is reasonable and the housing is still unfinished as I need to finalize the optics first. When finished cooling ribs will be added and I guess that it will do fine.

So what would be the realistic expectations as you see them?

 

[IRstuff]

Perhaps your optical designers are shying away because what you are asking for is not physically realizable in an imaging optical design.

Your 6x6 array has a maximum physical extent of ~27 mm. To get a 6° spread from this extended target requires a focal length of 257 mm. Assuming a 2-inch clear aperture, the optic would collect no more than 11° of the emitted 120° FWHM cone of light. That means that only about 1% of the emitted light will actually get projected through the optic.

TTFN

FAQ731-376

 

[EmanuelM]

IRstuff,
If the design was impossible nobody will be shy and will say this upfront. One thing I learned through the years is never make sugestions or show my way of thinking as the designer for some reason gets "locked" in this direction and will not try to think of other options. I do offer or explain all the possibilities available.

Now, the 36 led array (spread on a round aluminum pcb and not exactly 6x6) were mentioned to give macgyvers2000 an answer about the heat question.

They were used in another project, with TIR collimators in one case and tiny reflectors in another (both types available commercially)- and I mean one for each LED.

Speaking of that, the beam produced was 22deg with the reflectors and 18deg with the TIR - up to 50 meters it looks great.

There was no intention to try and collimate this led array.

By the way, Lamina Ceramics offers some large reflectors for their Titan 7 emitter module - one of them for 9 deg and another for 11 deg. I made some testing with both and it was dissapointing.

But the point is, they do exactly what you were talking about - a single reflector designed to concentrate light from 7 sources spread on a 1.25" (31mm) diameter PCB with a 6" (150mm) reflector.

check this link:

http://products.lsgc.com/product/titantm_series_led_lighting_system/

 

[lumenharold]

I'm sort of in a quandry about how to address this within the eng-tips policies. I think this is do-able. At least the quick modeling I did says it is but the issue I have is that I used one our technologies. I can't post the design (I would assume the red flag would go here?). Basically I used a Luminus Devices CSM-360 W45S which is nominally 3650 lumens at 4500K, our LED coupler to change the hi- NA LED output to a lower NA, and then took a swag at a CPC. The result was 4 lux at 200 meters with the angle at about 8 degrees. Further futzing would bring it down tighter. The roll-off was not very sharp either but then I didn't want to spend all day with this. The CSM-360 has an emitter area of 36mm^2 (6 X 6) and has a minimum output rating of 3600 lumens.

There are other LED couplers on the market and plenty of CPC's out there. The Luminus Devices LED might be a little expensive and a current hog for what you are doing but then you didn't say what the limits were in those regards.

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

 

[EmanuelM]

Hi lumenharold,

What a coincidence - I was just looking the other day at the CSM-360, just got the data sheet from a friend. It has a 16mm dome and impressive lumen output.