[Laser] Multimode optical TRX (was: led txr)

[Clint Turner]

The recent discussion on using various modes has been interesting:  If 
the path is more line-of-sight-limited rather than atmosphere-limited - 
that is, there is excess link budget - then it isn't necessary to use 
the most sensitive-possible gear and forgo a couple 10's of dB in favor 
of a modulation scheme other than straight AM:  To be sure, all optical 
communications in which we are likely to to engage is going to be AM of 
some sort - be it direct "baseband" audio or the modulation of a 
subcarrier of some sort.

Early on in testing optical gear, I tested with a number of modes using 
some subcarrier methods - one of which simply involved using a 
singly-balanced mixer (e.g. a couple of CMOS inverter gates) and an 
audio source to produce double sideband.  For receiving, I used a "bat 
listener" detector that I used - essentially a superheterodyne (yes, it 
has an IF) that converts frequencies 10-120 kHz down to audio (I've also 
used it as a VLF receiver.)

I've also used a laptop PC.  Since about any PC has a 48 kHz sample 
rate, subcarriers up to 20-ish kHz can be accommodated and the same DSB 
modulator above works as well.  For receiving, there are a number of 
DSP-type programs around such as Winrad or SDRadio by I2PHD and PowerSDR 
- just to name a few.

These can be used to generate signals as well.  PowerSDR - even if you 
don't have one of the Flex Radios - can be coaxed into generating the 
"IF" signals for transmitting (in the 10kHz range) but about the only 
"Swiss Army Knife" program that I've run across is Spectrum Lab by 
DL4YHF:  This can be made to produce and demodulate AM, FM, SSB, DSB or 
even operate in digital modes (PSK31, FSK, etc.) - but it's not 
particularly "new-user" friendly.  It, too, can use just a standard 
sound card and one can even "cascade" demodulators and filters to do all 
sorts of things like noise reduction, hum removal, notch filtering, etc.

In testing, all of this seems to work just fine and just requires a pair 
of cheap laptops - one for each end:  Some inexpensive netbook computers 
(Acer Aspire Ones) work fine when properly interfaced.  In many cases, 
cheap laptops (such as these Acers) tend to have somewhat inaccurate 
sample rates and fairly high intrinsic noise levels, but Spectrum Lab 
can be configured to compensate for this and the noise floor issue is 
easily mitigated by having adjustable gain in front of everything:  That 
and a simple high-pass filter (above a couple of kHz) would prevent the 
sound card from being blasted by low-frequency noise from the receiver.

It's actually the interface that is a bit tricky:  Switching between TX 
and RX is slightly awkward in the program - but scrips/presets can 
simplify this as the Spectrum Lab  was designed with this in mind.  One 
thing that I *didn't* do was built a T/R interface where the sound 
card's in/out would be swapped back-and-forth during TX/RX switching - 
using a USB Serial interface's handshake lines for "keying" so going 
between TX and RX required moving cables back and forth.

Of course, this can all be done with the "base model" of optical TX/RX 
system:  My already-built transmitter will easily accommodate 10-15 kHz 
modulation (and some minor component changes would take that up to 30 
kHz - all that is required for a cheap 48kHz ksps-capable computer) and 
the Version 3 receiver, as has already been demonstrated, has usable 
response into the 10's of kHz - although I could foresee a slight 
modification the the existing design to further-improve its performance.

Why haven't we done this in the field?  There's the hassle of lugging 
laptops along - especially when it's necessary to carry all of our gear 
several kilometers up rugged mountains!:  The existing gear is 
plug-n-play and we haven't really run across any situations where 
baseband AM wasn't more than adequate for the task - even when we are 
shooting across a sea of lights.  When we *have* used laptops, we tend 
to spend a while just fiddling to get things up-and-running -  but lots 
of practice beforehand would certainly make that easier!

If anyone is interested in how a program like Spectrum Lab would be used 
for this, let me know.

73,

Clint