[Laser] 5mw laser tranceiver kit

[C. Turner]

Hello,

First, I'll remind those who reply to trim out the previous emails from 
their reply, leaving only the salient points.

* * *

The idea of a "kit" has come up more than once in discussions that I've 
had with various people.  Having worked with "kitting" in the past, I'm 
painfully aware of how complicated it can turn out to be:  I have great 
respect for those who are able to turn out kits that strike the proper 
balance at being both an affordable, comprehensive kit with useful 
instructions yet useful enough to suit the needs of those who would buy it.

While clever, the Ramsey Kit - seemingly like many of the Ramsey kits - 
is about "90% there", with the missing 10% causing a degree of 
frustration and disappointment, and unfortunately, their LCB6K falls 
somewhere in this range.  I'll admit that I've not spent the $50-ish to 
get one yet, emails that I have received from those who have have always 
been on this topic.  I've built up all of the pieces that would comprise 
the elements of the kit over the years and know that as a whole, they 
have missed the mark:  For example, with a slight amount of extra effort 
and very little additional cost they could have increased the 
sensitivity of the receiver by 10's of dB, for example and the result is 
a pair of devices that sells the entire concept short.

Now, we (in this group) have the benefit of hindsight and experience and 
it would be fair to say that a reasonable majority of "experts" in the 
field of amateur optical communications has haunted this list at one 
point or another.  To be fair, most of the articles that have appeared 
in the press in the past have missed the mark and as a consequence, 
expectations have been fairly low:  I hope that through the efforts of 
many that this is gradually being corrected.

Having said all of this, we have developed, over the years, portions of 
what could be assembled as a basis of the "electronics" portion of a kit.

There's the "Simple PWM" transmitter:
http://modulatedlight.com/optical_comms/simpler_pulse_width_modulator.html

And already-designed boards for an optical receiver described on this 
page - see the sidebar:
http://modulatedlight.com/optical_comms/optical_rx2.html

This would still leave the majority of mechanical construction to the 
user, but isn't that much of the fun?

* * *

A method of "compensating" for scintillation was suggested and, as Chris 
mentioned, a method has been implemented as seen (and heard!) here:

http://modulatedlight.org/optical_comms/optical_comm_audio_interface_device.html#scintillation_compensator

In  this system, a 4 kHz "pilot" tone is sent along with the rest of the 
audio at a level that is 25% (or 12dB down) with respect to "peak" 
(100%) modulation.  At the receive end, this tone is extracted from the 
audio and used to drive a very fast AGC and is also notch-filtered to 
prevent the user from being driven insane.

There is also an audio clip:

http://modulatedlight.com/optical_comms/scint_comp_test2.mp3

(The "original" audio clip - pilot tone removed for most of it - can be 
heard here:  
http://ka7oei.com/optical_comms/Laser_LED_scintillation_comparisons_pilot_removed.mp3 
)

This demonstrates the same audio clip with and without compensation.  
This signal was transmitted using a laser and an 8" diameter reflector 
telescope and the optical path - about 23km - crossed a strong thermal 
boundary layer that capped the Salt Lake Valley - and was about as bad 
as an experience with scintillation that we have experienced - save for 
the use of a laser pointer on that same path during that evening.  
Uncompressed .WAV recordings containing the original 4 kHz pilot were 
made and later played back for this demonstration clip, notching out the 
4 kHz tone in each case.

As can be heard, the compensation is extremely effective, minimizing the 
effects of the scintillation and its distortion on the speech and 
music.  Of course, as Chris points out, it cannot recover what was lost 
in the noise.

When designing the system, there was some concern about whether or not 
to remove energy in the area of 4 kHz from the transmit audio in the 
fear that it might confuse the AGC on the receive end.  Using recordings 
made in the field with actual scintillation (some from Laser pointers, 
even!) numerous "simulations" were run and it was determined that 
practically speaking, incident audio caused very little problem and 
whatever effects there might be were very transient in nature:  In other 
words, unless you were looking for an interaction, you'd never spot it!  
The field recordings also provided a testbed for typical worst-case 
scintillation to be expected and the AGC system was designed to 
accommodate as they were representative of the sorts of effects that one 
might encounter - and far easier and accurate to make than simulations!

A pilot frequency of 4 kHz was chosen because of the (intentionally) 
limited bandwidth of our optical receivers and the fact that a 4 kHz 
tone can be easily synthesized using DDS techniques:  The fact that our 
modulators include a PIC-based tone generator system made this a 
"no-brainer."  A simple notch filter at 4 kHz doesn't materially affect 
the way the audio "sounds" as can be observed from listening to the 
clip.  Being that the pilot tone is consuming 25% of the audio power, 
its presence degrades the peak power only minimally.  The way the 
modulator we use is designed, it preserves the AGC's margin when 
achieving 100% modulation of the LED.

Interestingly, we've had relatively little need for the scintillation 
compensator as the problem of severe scintillation manifests itself to 
such a degree ONLY with the use of lasers.  Since we don't often use 
lasers, the scintillation hasn't been too bad.

Another observation is that if there is a noticeable amount of dust or 
haze in the air, scintillation can go nearly to zero.  While the path 
attenuation will tend to skyrocket, we've observed that moderate amounts 
of optical dispersion tends to more-rapidly break down coherence along 
the path.

This effect has been seen on several of the occasions that we have 
spanned one of our "standard" 107 mile (173km) paths.  Having done this 
path with crystal clear air, extremely hazy air (that is, the opposite 
end of the path could not be seen with the naked eye) and somewhere 
in-between, we've noted that the worst case is the "crystal clear" air 
path.  In the other two instances, the scintillation was quite low - 
even when we've used just Laser Pointers for 2-way communications on 
that same path.  (Again, the scintillation on the LED-based systems - 
which we use to coordinate or Laser activities - was comparatively low 
in any case.)

Now, were I designing a kit, would I include a scintillation 
compensator?  No.

* * *

We've done some testing with diversity transmission and detection as well.

As Chris says, listening to multiple optical receiver in "stereo" is a 
rather interesting experience and can even be disorienting as the 
amplitude (but not phase) bounces between the two.  While I've used this 
scheme before, it's not normally done as it's usually a bit of a hassle 
to set up (or even bring along!) a second optical receiver as one is 
usually in a bit of a race to get things going - especially if one (or 
both) parties are stuck on a mountain in the middle of the night!  Even 
if these two receivers are NEXT to each other, there is a noticeable 
difference in the fading between the two.

I've also use dual transmitters, making use of a "Y" cable that allows 
two transmitters' LEDs to be put in series (since my modulators modulate 
the current through the LED, they don't care if there are two in 
series...)  It, too, made a notable difference in the scintillation.

This brings up another point:  Increasing far-field flux.  Since one 
simply cannot use a larger LED to increase the amount of "signal" being 
received at the far end (a larger LED simply means a larger "spot" - of 
which no more than before is "hitting" the receiver) the easiest way to 
increase far-field flux (aside from using a same-sized LED that runs 
more current and thus emits more photons per unit area) is to run 
multiple transmitters.

In addition to increasing the far-field signal by 6dB (assuming that 
both are equal in output as well as properly aimed and focused) this 
also decreases scintillation by virtue of enlarging the effective 
aperture of the transmitter.  The obvious hassle is that one needs two 
emitter systems and the practical difficulties that this involves (e.g. 
having to build two devices that emit and, possibly, the increased power 
consumption.)

Having observed the differences that multiple receivers and/or 
transmitters make in the audio recovery brings to mind the importance in 
using as large an aperture - for both transmit and receive - as 
practical.  Having used these devices in the field I'm also aware that 
going too far overboard with the size of the lenses can risk trading 
practicality with performance so I'm not likely to attempt the use of a 
"giant" Fresnel lens anytime soon!


As always, comments welcome!

73,

Clint
KA7OEI