[Laser] Newbie

[C. Turner]

I might chime in again with respect to what Chris says.

I've experimented with using a Fresnel to collimate a beam from a laser 
diode in the past - and it proved to be a demonstration of exactly what 
"diffraction limited" optics contribute to the efficiency of a system.

While you can do it (that is, use a Fresnel) it doesn't work very well 
and you lose most of the the laser's light - for several reasons.

For starters, take a look at this picture:

http://ka7oei.com/optical_comms/LED_Laser_spot.jpg  (Ignore the color 
balance...  I never bothered to "fix" it...)

(This picture is on the page:  
http://ka7oei.com/Coherent_versus_noncoherent_test.html and is Figure 1. )

What you see are two light sources:  The one on the left is an LED and 
the one on the right is from a semiconductor laser.  The reasons for the 
picture (as explained on the linking web page) was to illuminate the 
aperture of an 8" reflector telescope with both an LED and a laser.  In 
order to do that, I had to radically alter the optics of the laser diode 
assembly.

What's important to note is the shape of the laser light source.

In checking with some online literature, it would appear that the 
beamwidth of a "typical" laser diode (whatever that is...) is in the 
area of 25 degrees or so (in the vertical axis of the diode junction - 
the so-called "fast" axis) while it can be in the area of 5-10 degrees 
in the horizontal axis of the diode junction (the so-called "slow" axis) 
- but from the picture, you can see that it's more like a 4-5 to 1 ratio 
- especially since in the picture above, the beam's width is truncated 
somewhat by the assembly, so it is actually "wider" (or "taller" from 
the perspective of the picture) than pictures.

With the optics in a cheap laser pointer (or an expensive one, for that 
matter) one can actually still see that the spot that the laser pointer 
produces isn't really a round dot, but is, in fact, somewhat elongated - 
but the extent to which it is elongated is somewhat obfuscated in practice.

So, what do you do with a wide "fan" beam like that - and how do you 
illuminate something with it?

In the case of the telescope experiment, I had to throw away most of the 
beam, using only a portion in the center.  I estimate that I lost more 
than 80% of the laser diode's energy simply from it NOT reaching the 
mirror.  Since the purpose of the exercise wasn't to maximize the 
radiated power, I wasn't too concerned.

***

How about a Fresnel lens then?  Unless you have a really short and wide 
Fresnel (that is, a lens that resembled a "strip" you are going to have 
the same problem in that you are going to have to throw away most of the 
energy in order to illuminate the "back" of the Fresnel.

Then there's the problem of "matching" the laser's pattern to 
"illuminate" the Fresnel - or any other lens.  Ideally, you would want 
the the light coming out of the laser to be a square shape that, at the 
focal length of the Fresnel, just happens to "light up" the entire 
backside of the Fresnel - that is, the beamwidth of the laser 
corresponding to the subtended angle of the width/height of the lens as 
seen from the perspective of the laser diode at the focus.  Clearly, 
with a laser diode you don't even get a "round" beam and with a "fan" 
beam you'll have to sacrifice some energy in that you'd have to arrange 
it so that in the "slow" axis (the narrow dimension of the beam) ended 
up being as wide/tall as the Fresnel at the distance of the focus.

That could be arranged by picking a Fresnel with a corresponding 
focusing distance such that the laser's light spread out to be the size 
of the Fresnel at that distance, or you could add yet another lens to 
adjust the size of the beam coming out of the laser diode so that it 
better-fit the lens that you had.  This "secondary" lens, too, would 
best-be diffraction-limited as well, but since it probably need not be a 
very large lens, it shouldn't be too difficult or expensive to obtain 
such a lens with 1/4 wavelength (or better!) accuracy.

***

Coming back to my experiment with using a Fresnel lens and a laser, two 
things happened.

- The first, already mentioned, was the fact that I had to throw away 
most of the light from the laser to illuminate the back of the Fresnel lens.

 -The second, to which Chris alluded, was the fact that while some of 
the light through the Fresnel lens was thrown forward - as you would 
expect it would be - a significant portion was simply scattered about, 
resulting in a rather cool-looking diffraction pattern all over the 
wall.  (I have no way to know what percentage of the light was lost in 
this way, though.)

Not only that, but the "collimated" beam emanating from the Fresnel was 
most definitely non-uniform - that is, it was quite "blotchy" and 
"speckly" with bright and dark spots across its cross-section.  (I 
*really* meant to take a picture of the result - and I should probably 
re-create the experiment someday to do so...)

***

How does one collimate a laser to a large diameter in an economical 
manner?  While there are a number of ways to do this if you have a very 
large budget or serendipitously happened on some cast-off optics from 
some project that, itself, had a large budget, neither of these 
scenarios is likely to happen or be replicated in the experimenter's 
community.  Aside from collimating to a "reasonable" size (a few cm 
diameter - a size for which diffraction-limited optics are still 
affordable and practical) the other practicality is to do what has been 
done in the past:  Use several disparate laser sources, separated from 
each other to "synthesize" a large aperture.  Strictly speaking, this 
method isn't coherent either - unless you've beam-split one laser or 
have happened to lock them all together:  Nether scenario being likely 
on an experimenter's budget!

***

As Chris pointed out, all of this isn't really a problem on the 
"receive" end because the "coherence" is already been demolished by its 
traveling through the atmosphere.

Anyway, it's all fun!

73,

Clint
KA7OEI