Light Pipes in Series

[Rafferty]

I'm designing two light pipes in series with each other (see attached sketch). Both will be cylindrical rods of approximately .3" diameter with a few extra features (not shown) for mechanical assembly. Light pipe A will be embedded in a base housing. Its entrance face will sit .020" above a surface-mount LED; its exit face will protrude above the housing top panel. Light pipe B will be embedded in a removable component and its entrance face will sit .030" above the exit face of light pipe A. Both light pipes will be injection molded out of polycarbonate and will have diffuse end faces.

I realize that a single, continuous light pipe would be much better than two light pipes in series. But the requirements of the application dictate the arrangement described above. Is this arrangement feasible?

Any relevant feedback would be appreciated.

 

[CorBlimeyLimey]

If "the requirements of the application dictate the arrangement described above" then you have no other choice.

Yes it is feasible, but without knowing the requirements of the final output it is impossible to say whether it will be acceptable for its intended function.

 

[Rafferty]

Thanks for the response, CorBlimeyLimey. I'm going to build a mockup to test the concept. I'm also going to make the entrance to light pipe A concave so as to better capture the light from the LED.

 

[IRstuff]

One, usually, when refering to requirements, is refering to things like transmission, beam quality, etc.

TTFN

FAQ731-376

 

[Timelord]

rafferty,

My first thought would be to see if there is any material you could make an index matching gasket from. Perhaps a clear silicon rubber. Index matching goes a long way to cutting the losses you will see from an interuption in in the pipe. this technique is used to mate optical fibers for high powers where the loss would lead to self destruction, except we use a liquid (easier to tailor the refractive index in a liguid). If not a gasket, try to arrange the ends of the light pipe so they come into tight contact when the removeable part is in place instead of a .030" gap.

Timelord

 

[lumenharold]

You'll have a fair amount of loss with the couplings. You also didn't mention what the sides of the lightpipe surface will be so i assumed the same finish as the ends.

Standard 120 degree LED, polycarbonate light pipes, etc. You are also going to get light spraying other places than where you planned. Modeled the ends flat.

A test bench mock-up would be easy, so why not? Specs would be a good idea to know what you are designing to.

Harold
SW2008 SP3.0 OPW2008 SP0.1 Win XP Pro 2002 SP2
Dell 690, Xeon 5160 @3.00GHz, 3.25GB RAM
nVidia Quadro FX4600
www.lumenflow.com

 

[Rafferty]

Timelord,

You bring up some interesting points. I could conceivably bond a clear silicone rubber pad to the top of light pipe A. The pad could be a few thousandths of an inch thicker than the .030" gap. However, the pad might not be durable enough to withstand a couple thousand cycles of mate/unmate with the removable part. I'll keep the idea in mind if mock-up testing doesn't go well.

I'm a bit leary of bringing the tips of the two light pipes into contact. The light pipes are too delicate to support the weight of the removable part, plus the handling forces caused by the user.

Thanks for the ideas!

Rafferty

 

[Rafferty]

Lumenharold,

Thanks for the ray tracing image!

I've made a few design changes that should correct some of the problems that you've highlighted. The gap between the two light pipes has been reduced to .020" and it will occur within a hole in a white, plastic housing. The entrance end of the bottom light pipe is now concave and has been brought closer to the PCB to better gather light from the LED.

To clarify: The sides of the light pipes will be as smooth as possible, hopefully a near-mirror finish. The entrance face of light pipe A will be smooth and concave, while the other light pipe entrance will be smooth and flat. The exit ends of both light pipes will be flat, with diffuse textures.


I appreciate everybody's helpful comments. I hope to get this built and tested next week. I'll let you know how it turns out.