[Laser] Newbie

[C. Turner]

Hi Keith,

I've been somewhat involved with amateur optical communications for a 
while and, at the risk of starting a discussion, run down a list of a 
few things.

Lasers.  As everyone knows, Lasers are cool!  Nowadays, with cheap laser 
pointers available at the dollar store it's easier than ever to buy 
something that will output a reasonably-collimated light (albeit of very 
small diameter) that can be easily modulated with PWM techniques.

Advantages:
   - Lasers are cool!
   -Very cheap.  If you pay $10 for a low-power red laser pointer, 
you've paid too much!
   - Easy to use - from an electrical standpoint.  It's very easy to use 
PWM to modulate them at audio frequencies.
   - Visible.  It's much easier to point a beam that you can see!
   - Relatively safe.  Cheap laser pointers are pretty wimpy, so you 
really can't hurt anyone with them - although they can still be 
distracting and you can get into big trouble if you do something 
stupid!  (In some localities, their use in a manner to communicate may 
actually be illegal - check your local listings!)
   - Modulation bandwidth.  Some laser pointer modules can be modulated 
at MHz rates.  (This isn't true of all modules - it depends on how they 
and their drive circuits are built.)  This potentially allows high data 
rates/video bandwidth to be modulated.
   - Lasers are cool!

Disadvantages:
   - Small beam diameter.  The small beam cross-section makes it 
vulnerable to potentially murderous scintillation  problems and/or 
difficulty with beam-blockage by small particles such as bugs or water 
droplets.
   - Very difficult to aim.  Most people on their first foray into laser 
communications greatly underestimate this particular aspect:   Trying to 
use a tripod aim a laser pointer really doesn't work very well!  It is 
recommended that one build a very simple fixture (that can be attached 
to a tripod) that allows precise Az/El adjustment to save yourself hours 
of frustration and the removal clumps of hair.  It is particularly 
recommended that this be combined with an electronic aiming aid.
   - Coherent light.  The use of "coherent" light from a laser can 
dramatically increase the degree of scintillation on the transmitted 
beam.  The use of coherent light also limits the sorts of optics one can 
use to collimate the beam:  The use of "diffraction-limited" optics 
(e.g. extremely accurate and precise lenses) is required to avoid 
disrupting the beam from a laser.  If one wishes to obtain a large beam 
diameter, you'll need a piece of glass that large, along with the 
weight, expense, fragility and awkwardness that goes along with it.  
After traveling some distance through the atmosphere (a couple of km - 
at most - will do it!) the beam loses its coherent properties, anyway.

***

LEDs.  It is no secret that I'm partial to LEDs when it comes to optical 
communications so I have a definite bias on this point.

Advantages:
   - Intrinsically safe.  Because it is necessary to collimate the beam 
to a fairly large diameter, the power density of the beam of systems 
based on high-power LEDs is lower than that of even a cheap laser pointer.
   - Visible.  Unless you pick an infrared LED, you can see these, too.
   - Inherently collimated to a larger diameter.  Because you need to 
involve larger optics to effectively collimate light produced by an LED, 
you end up with a larger-diameter beam.  Since you'll probably be using 
a fairly large lens for your receiver, anyway, this isn't too much of a 
hardship, really.
   - Non-coherent light.  While LEDs are monochromatic, they do not 
produce coherent light.  This allows relaxation of the lens 
requirements, permitting the use of large, inexpensive optics such as 
molded plastic Fresnel lenses.  This reduces scintillation while the 
practicality of a large transmit aperture also reduces few other 
potential sources of degradation (e.g. through reduction of "Local 
coherence", adding the property of "aperture averaging", etc.)
   - Relatively easy to modulate.  LEDs can be easily modulated by 
varying current or with PWM techniques.
   - Much easier to aim.  With the wider divergence, it's much easier to 
set up a communications system.

Disadvantages:
   - You need to build more stuff.  You can buy a laser pointer that 
produces a ready-to-use beam, but you'll have to build an LED-based 
transmitter from scratch!  (Again, this may not necessarily be a 
disadvantage to some!)
   - To get reasonable performance, physically large apertures are 
required.  Because of simple geometry, a fairly large lens is required 
to decrease the divergence of an LED source in order to maintain 
far-field flux.  This has, in part, to do with the fact that the source 
size of a laser emitter's aperture is on the order of 10's of microns at 
most while an LED's source can be millimeters.  However, since large 
apertures are desired to minimize atmospheric disturbances, this isn't 
really much of a disadvantage.
   - High power LEDs are a bit more difficult to modulate at high 
frequencies.  This increases the challenge in obtaining the frequency 
response required to obtain high-rate data or video.
   - More power is required.  It is not practical to get reduce the 
divergence of an LED transmitter down to that of even a cheap laser 
pointer (for the reasons mentioned above) so the far-field flux of an 
LED-based system (as compared to a laser-based one) is lower for a given 
amount of total optical power.  To overcome this, high-power LEDs are 
typically used to "brute force" the far-field flux to similar levels.  
What this amounts to is higher battery drain and larger heat sinks - but 
it's still at a manageable level.
   - It's not coherent.  The fact that LED emissions aren't "coherent" 
(e.g. "Rule 1.12" - refer to "laser" list QTH.net archives from November 
2009) has caused some heartburn/dispute when it comes to such a "LED 
QSO" counting in contests which has, in turn, lead to a healthy and 
spirited discussion that will likely come to some sort of resolution in 
the coming months/year.  If you simply wish to "achieve" and do some 
interesting things, this aspect is of little consequence.

***
***

Detection methods:

Photodiode

As for detection,  my personal preference is the Silicon PIN photodiode.

There are plenty of cheap devices and the arguable favorite is the 
venerable BPW34.  This device can be obtained for less than $1 U.S. and 
it has quite reasonable characteristics.  For voice-bandwidth use (e.g. 
up to 3 kHz bandwidth) it is as good as you'll ever need for all but the 
most demanding situations:  The intrinsic noise of this photodiode is 
lower than that of any practical analog front-end amplifier that you'll 
ever use (unless you resort things like Peltier/Cryogenic cooling - but 
that gets really complicated really fast - just ask the "CCD/Telescope" 
folks!) and it is very easy to use.  The active area of the BPW34 is 
large enough that it is quite forgiving in terms of focusing and alignment.

Note:  For frequencies that are lower-still (<300 Hz) the photodiode's 
intrinsic noise does become more critical, but these aspects are 
well-covered in the various discussions of the "K3PGP" circuit found on 
the web which detail the optimal photodiodes, amplifiers and 
construction techniques required to obtain the best performance.

The BPW34's main fault is that it's a bit on the "large" side, being 
several mm square - a fact that directly translates to higher 
capacitance which, in turn, reduces obtainable sensitivity at the higher 
audio frequencies.  Ironically, smaller devices - while available - are 
often more expensive, but they may be found - just make sure that they 
aren't "too" small - that is, smaller than the "blur circle" of the 
"receive" optics that you use.  Making them too small can also 
complicate pointing as "beamwidth" may get to be too narrow.

With a properly-constructed photodiode receiver, you should be able to 
easily "hear" a raised noise floor when you point at any star that you 
can readily see in the night sky:  Any of the "naked-eye" outer planets 
(Mars, Jupiter, Saturn - that is, those that are visible in the middle 
of the night) and the brighter stars (Sirius, Vega) should pretty-much 
saturate the receiver!  With such a receiver you should also be able 
"copy" voice from a distant, modulated light that is simply too dim to 
be visible to the naked eye.

***

Phototransistor:

Don't!

Phototransistors are pretty terrible if you wish to obtain reasonable 
performance at low light levels as they produce a fair amount of 
intrinsic noise of their own.  It has been reported that the "dark 
currents" of phototransistors are largely dominated by the leakage of 
the transistor, rather than the photo-active portion.  It is unfortunate 
that a few devices available to the would-be optical experimenter (such 
as the Ramsey "Laser Communicator") use a phototransistor.

Another practical consideration for the phototransistor is that its 
photoactive area is quite small in area, making it a bit more of a 
challenge to match it to a lens.

***

CdS Cells and "Solar Cells."

I'm lumping these two disparate devices into the same category in that 
they are both generally unsuitable.  CdS cells aren't extremely 
sensitive, either - and they are quite slow in responding - but they do 
have the advantage of being optimally-sensitive in the "yellow-green" 
portion while silicon devices (such as photodiodes/phototransistor/solar 
cells) favor red and, especially, near IR.

In some ways, photodiodes could be considered to be tiny solar cells 
(they both produce current/voltage when illuminated) but unlike 
photodiodes, solar cells are not at all optimized in the way that would 
matter for medium-rate communications (voice frequencies) because they 
tend to be quite noisy (not a problem when sun is shining on them) and 
being fairly large, they have very high capacitance - which would wreck 
their ability to "demodulate" light at such frequencies.

***

Photomultipler Tubes (PMTs):

Having been around for >60 years, PMTs are a "mature" technology - but 
still unparalleled when it comes to sensitivity, being able to detect 
single photons which is why they are still widely used - most 
prominently in the study of high-energy physics (think "Neutrino 
detectors").

Although widely available on the surplus market, relatively few of them 
work well in the "red" portion of the visible spectrum as it is the 
tendency for most known photo-active materials suitable for use in a 
photomultiplier to function at shorter wavelengths.  There are some 
devices (so-called "multi-alkali" and GaAs) that have reasonable-to-good 
sensitivity at the "red" end but these are a bit harder to find on the 
surplus market.

Photomultipliers are also somewhat fragile.  Being a vacuum tube device, 
they are physically fragile, but more importantly they can be "wrecked" 
by exposure to room-intensity light while powered up (or even powered 
down) - at least temporarily (for minutes, hours or days) while some 
devices (especially GaAs ones) can be destroyed if so-exposed with 
voltages applied.  They also require a fairly high voltage (1-2kV 
typical) to operate.

Having said this, they are, in fact, relatively easy to use:  You apply 
power to one end and pull the signal from the other - but you must be 
REALLY careful with them!

***

Avalanche Photodiodes (APDs):

These are a type of photodiode that have intrinsic self-amplification.  
Almost exclusively used at higher frequencies (high-rate video and data) 
their utility at "medium rate" frequencies (e.g. voice) is rather 
limited.  If one wishes to obtain good sensitivity at high bandwidths, 
about the only two options available are Photomultipliers and Avalanche 
Photodiodes.

Not surprisingly, very limited information has been published concerning 
the use of APDs at "voice" rates, but I can relate a bit of my own 
experience:   If you run them at "full gain" (e.g. rated voltage with 
maximum self-amplification) they are terrible - FAR worse than a simple 
PIN photodiode!

Why?  In short, their self-generated noise goes up at a much faster rate 
than their own gain amplification.  For high-speed circuits (e.g. 
high-speed data and video) this isn't much of a problem as the 
fundamental noise floors of these "front ends" is naturally quite high, 
but for low/medium rate communications (voice and, especially very 
low-frequency - below 300 Hz) they can be a disaster as they will simply 
"roar" with noise!  For really low" frequencies (under 300 Hz) PIN 
photodiodes (in a "K3PGP"-type circuit) and photomultipers would 
probably be the smart choice!

In my "voice-bandwidth" experimentation with various APDs, they will 
only out-perform a simple PIN photodiode when they are run at rather low 
gain (e.g. low self-amplification) and even at that, the only really 
important aspect - the Signal/Noise improvement (as compared with a 
standard PIN photodiode receiver) appear to be rather modest - only 6-10dB.

APDs tend to be quite expensive, too - costing as much (and more!) as 
you'd really want to pay!  You can get them surplus in the form of 
"Laser Link" receivers (to wirelessly connect your camcorder to your TV) 
but these devices have quite small active areas, somewhat complicating 
their use with, say, Fresnel lenses.  There are also other types of APDs 
as well that may offer better performance, but these are rather 
specialized and I've not found any of them on the surplus market (e.g. 
that I'd willing to pay for!) and have not evaluated them.

Having said this, despite these difficulties and rather high cost, I'm 
constructed an experimental APD speech-frequency receiver that I'll be 
evaluating in the field later this year.

***

So, what have we done with Lasers/LEDs?

As a matter of routine, we can easily span the Salt Lake valley 
(20-30km) using LED or Laser pointers, with our gear tending to 
"saturate" due to high signal levels.  We have also (somewhat routinely) 
spanned a 172km distance with both LEDs and laser pointers, using 2-way 
voice communications - although the LEDs worked quite a bit better.  
Finally, we did manage a distance of about 278km - one-way voice and 
2-way MCW - under less-than-optimal (hazy!) conditions using LEDs:  We 
were unable to span the path with lasers...  We have also done 2-way MCW 
(and 1-way voice) using "mountain-bounce" (e.g. both of us pointing at 
the same mountain) spanning the Salt Lake Valley (and its humming mass 
of city lights!) over a 20+km distance.  In the future, we do hope to do 
more...

***

This response turned out to be more long-winded that I expected (but I'm 
not really surprised...) - but more information can be found at this 
site and its links:

http://www.modulatedlight.org

73,

Clint
KA7OEI


>
> Hello all - 
> I was wondering who is on this list? I am new to the idea of laser communications, although I have been involved both professionally and as a ham in microwave communications.
>
> I have found a few web-sites with information on amateur activity but would like to know what equipment people are using - is it mainly the solid-state lasers from pointers, or what? Also what detectors are the most common?
> ?Regards, 
> Keith VA3QF 
> CAMBRIDGE, Ontario 
> EN93tj 
>
>