Is Collimating a Small LED Beam Possible with Various Lenses?

[SMD1990]

I need to collimate a small beam of light from an LED. Getting the light from the LED focused is the easy part. Where I am having difficulty is in the collimation.

I have a varied assortment of lenses. From large, heavy, glass ones from old projectors... To cheap, little, plastic ones from toy binoculars and cameras and such...

I have been trying various things. Using the converging lens from an old slide projector, I can collimate the light emitted by an LED. However, if I try to focus that light first, I cannot seem to.

I am not seeking perfectly parallel rays. If my little beam would remain just a small spot for a few centimeters, that would be enough.

Any help, suggestions, ideas... I would highly appreciate it. I am starting to run out of ideas. In my head, it does not seem like it should be this hard.

 

[DaveC426913]

I am not seeking perfectly parallel rays. If my little beam would remain just a small spot for a few centimeters, that would be enough.

Do LEDs not do this already? They shine a pretty narrow beam. You want it narrower?

 

[Andy Resnick]

Getting the light from the LED focused is the easy part. Where I am having difficulty is in the collimation.

<snip>
Using the converging lens from an old slide projector, I can collimate the light emitted by an LED. However, if I try to focus that light first, I cannot seem to.

<snip>.

I'm confused- what are you trying to do?

 

[SMD1990]

In what world is 30 to 60 degrees narrow? Even 10...

Basically, I need a very fine beam of light. The thinner, the better. I need any divergence to be negligible until at least ten or so centimeters. I would describe it as a very short-range laser, but I feel that would be misleading.

I can collimate the light from the LED using various lenses that I have. However, if I first try to focus that light to a point, I cannot then get it collimated.

 

[Drakkith]

How about a fiber optic cable? Or do you need the light to travel through the air or something?

 

[SMD1990]

Yeah, I need it to go through the air.

I am trying to be able to project a tiny dot onto a surface whose distance can vary by several centimeters.

A laser would work, but the color choices are limited. For super cheap, there is only red. Plus, the size of the dot produced by cheap red laser pointers is quite large.

 

[Drakkith]

Ah ok. Well, I cannot help you then, sorry.

 

[DaveC426913]

In what world is 30 to 60 degrees narrow? Even 10...

Basically, I need a very fine beam of light. The thinner, the better. I need any divergence to be negligible until at least ten or so centimeters. I would describe it as a very short-range laser, but I feel that would be misleading.

OK, well, now you've given some more specifics haven't you... :uhh:

How about a fiber optic cable? Or do you need the light to travel through the air or something?

Yeah, I need it to go through the air.

What happens to the light that comes out the end of a fibre optic cable? It should be pretty well collimated to cross several centimetres of air I'd think.

 

[SMD1990]

The only fiber optic "cables" I have are these little short ones from a 1990s LEGO Technic set. I seem to recall the beam intensity leaving them as being quite weak. As in, not visible on surfaces more than a couple of centimeters away from the end of the cable. I do not know if this is normal of fiber optic cables, or if it is merely the result of my "cables" being essentially "toys".

I will give it another try, though.

 

[DaveC426913]

The only fiber optic "cables" I have are these little short ones from a 1990s LEGO Technic set. I seem to recall the beam intensity leaving them as being quite weak. As in, not visible on surfaces more than a couple of centimeters away from the end of the cable. I do not know if this is normal of fiber optic cables, or if it is merely the result of my "cables" being essentially "toys".

I will give it another try, though.

You should be able to pick up some higher quality cables cheap at a science shop or online.

 

[Andy Resnick]

Basically, I need a very fine beam of light. The thinner, the better. I need any divergence to be negligible until at least ten or so centimeters. I would describe it as a very short-range laser, but I feel that would be misleading.

If I understand you correctly, what you are trying to do is impossible- a ray of light (that is, a thin pencil beam of light) carries no energy (no intensity). This seems contradictory to laser beams, but laser beams are not collimated (as you have noticed from laser pointers).

http://www.natscience.com/Uwe/Forum.aspx/optics/1493/Slyusarev-Book

If you are trying to project a small circle onto a distant surface, there are several approaches- projecting a demagnified image of the LED, for example. Usually, this means the lens must be near the surface- like using a microscope objective in reverse, or a lithographic projection lens.

Using light out of a fiber is fine, too- the cone angle of light exiting the fiber is given by the acceptance angle, and depending on the specifics, can be a few degrees or more. Some fibers can be used with a pigtailed GRIN lens, which will reduce this divergence:

http://www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=1340
http://www.redoptronics.com/grin-lens.html

Using the exit face of the fiber as the object, you can again project a demagnified image onto a distant surface as above.

 

[ProTerran]

Well, I think that it is possible. but light passing setup will be very dim. All you need to do, is put a converging lens with small aperture and placed it at a distance equal f (focus length) from the led or from the focused image of led. Obtained dot diameter will be the same as the aperture of collimating lens, so if you need very small dot then you will have to use very bright source of light.

Of course there is more than one solution to this problem. Also you can focus image on the plate with small hole in it, and then collimate image of this hole with normal big lens. It all about blocking rays that excess given angle.

 

[chrisbaird]

An application that needs small spot sizes over different distances is laser scanning/lidar. They use lasers instead of LEDs because they give the best spot size (and are coherent). So your best bet is a laser. You can get a presentation laser pointers inexpensively these days at an office supply store (or as free give-aways at conferences). But, as you mentioned, they are usually red.

 

[Claude Bile]

Any light source will naturally diverge in space. The ability to collimate a source depends on its divergence (sometimes called the M^2). LEDs typically have poor M^2 (high tendency to spread in space), so it is not surprising to see that you are having problems collimating such a source.

Practically speaking, collimating an LED over cm is difficult. One solution is to pass it through an optical fibre. The numerical aperture of the fibre will reduce the divergence (as well as the intensity).

Claude.

 

[sophiecentaur]

The application is still not clear. If you want to have a spot of light at a certain distance then it can be focused. A single lens would produce a magnified image of the original spot (and less intense, of course. A larger aperture lens of longer focal length would help (we're down to 'f value' - as in cameras, here).
If you want a narrow, parallel, beam then a light pipe (i.e. a glass rod fatter than an optical fibre) might let more light through. I doubt that you'd do better than a laser pen though.
I suspect that we are all suggesting solutions that may be out of your range of possibilities / cash, though.
BTW, did you consider a large lamp in a box with a small hole in the side? Or even an old slide projector pointed through a blackened, narrow tube might do. You could have any colour you like, too.

 

[SMD1990]

The application is still not clear. If you want to have a spot of light at a certain distance then it can be focused. A single lens would produce a magnified image of the original spot (and less intense, of course. A larger aperture lens of longer focal length would help (we're down to 'f value' - as in cameras, here).
If you want a narrow, parallel, beam then a light pipe (i.e. a glass rod fatter than an optical fibre) might let more light through. I doubt that you'd do better than a laser pen though.
I suspect that we are all suggesting solutions that may be out of your range of possibilities / cash, though.
BTW, did you consider a large lamp in a box with a small hole in the side? Or even an old slide projector pointed through a blackened, narrow tube might do. You could have any colour you like, too.

I am currently try to find an optical cable with which I can experiment. I would hate to spend $17 for such a cable from Best Buy, just to find less than satisfactory results. Surely, someone I know has one I can borrow...

In addition to that LEGO Technic optical "fiber", I have also tried some fishing line. Actually, it worked about as well as the LEGO part. Since both are bound to be comprised of some sort of plastic, not silica, I suspect my results will be better with a proper optical cable.

I was hoping to maximize the efficiency. That is why I was wanting to focus the light from the LED, then collimate it. I wanted all of the light output, not just a narrow degree of it. That is why I have not tried much with small aperature sizes. The things you suggested in your last paragraph seem like a waste of much light.

 

[sophiecentaur]

I think you need to be realistic and go for a solution that works. Only then can you start to get 'fussy' and start to get 'elegant' about it. If you don't have the technology or the money to produce a hi tech solution then you may just have to come to terms with the reality.
If the whole purpose of the exercise is to produce the tiny spot then, of course, you could / should be picky but, if you just need this spot fro some other purpose then be pragmatic, for a start, at least.
The fact is that the sort of optics you are after is non-trivial. A laser diode /pointer would get you most of the way there but that's not good enough? Well, why do you think they're not any better than they are? - Probably because it's another step up in complexity and cost. If you come to terms with the fact that your high efficiency is too high a goal, then you could consider, perhaps, a Quartz Halogen bulb as a source and a simple tube collimator.

As far as your cheapo optical fibre goes - it probably is actually no more than some nylon monofilament - why not, if it does the job they want it to?

 

[SMD1990]

Okay... An update. I found a fiber optic cable that was negligibly cheaper than the one at Best Buy. More importantly, its packaging was such that I could easily open it, test it... Then return it without question, if unsuccessful.

Well, my initial results have been a bit of a letdown. The divergence out the cable is quite large. Smaller than that of the original/source LED. Still, maybe fifteen to twenty degrees. Far off from producing a tiny dot (the smaller, the better; no more than 0.1 millimeters, preferably) over a range of few centimeters.

It performs better than the LEGO fiber optic or the fishing line. A lot better. (With them, the output cannot be seen beyond maybe a centimeter or so.) With the optical audio cable, the range is a lot further. Though, not evenly remotely collimated.

Anyway, I was wondering if maybe there was something wrong with this cheap cable. Would more expensive ones perform closer to how I need (and hence why you guys have suggested fiber optic cable, in the first place)?

I am not done trying with it. I was just hoping for any additional input. It is quite late now. So, I will continue in the morning, after checking to see if you guys can provide me with any further guidance.

 

[sophiecentaur]

It would be necessary to have a very flat, polished end on your fibre, if you want minimal spread.

 

[ProTerran]

What you want to achive is physically impossible - I recommend you to get familiar with diffraction effect. As i said before, collimating light source like LED to small point can be only done by blocking rays - it's very inefficient.