Figure 1:  Map and elevation profile showing the path between the location near Mt. Nebo in the south to Inspiration point to the north.
Click on the image for a larger version.

Better weather!

Our August 18, 2007 Expedition, although successful, left us with mixed feelings:  Even though we were able to complete the contact despite terrible optical conditions (e.g. thick haze and light pollution) we still wanted to try a "lightbeam" contact again when weather and air conditions were better.

Such an opportunity arose on September 3, 2007 when a coincidence of schedules and good air conditions occurred, so we again headed to the same places as last time:  Ron, Elaine and Gordon headed north to Inspiration Point, while I went south, again to the area around Nebo, this time with Dale, WB7FID, as Tom was out of town.   Although there had been a few clouds in the morning, it had cleared up by the evening, with no imminent threat of storms anywhere along the path.  Additionally, previous weather and wind conditions had resulted in very clear air with only a hint of haze, with each site being clearly visible from the other.

Same gear, same path - almost:

Since the last time, minor modifications had been done to the optical gear:

• The beamwidth of the optical receivers were narrowed by masking off a portion of the photodiode's active area:  A small drill bit was used to make a clean, round hole in self-adhesive copper foil that was centered over the diode.  After masking, the focusing and paraxial alignment was checked and adjusted as necessary.  The result was narrower and sharper beamwidth with no loss in sensitivity, as the holes were larger than the blur circle of the lens, plus better rejection of off-axis light, such as that from light pollution.
  
• A simple pushbutton was added to the modulator, in parallel to the "LED On/Off" switch to allow easier on/off keying of the LED using Morse:  This is useful for signaling - particularly during alignment.
• It was determined that, if the audio interface box and modulator shared the same battery, a slight amount of crosstalk occurred.  By adding shielding to the enclosure and using separate batteries, the only remaining crosstalk was that resulting from light scattered in front of the transceiver.

While, at Ron's end (Inspiration Point) they were within a few feet of the same location as before, a slightly different location was used at the Nebo end:  Last time, just to be safe, we'd moved along a fence line to better-clear the optical path from a nearby ridge.  As it turned out, the clearance with that ridge was sufficiently large that upon return, we parked at a spot much closer to the microwave radio site.  This simplified setup considerably, as we didn't need to hassle with a fence, and the gear was within a few feet of the vehicle.  This also meant that owing to geography, we were about 230 feet (70 meters) farther from Inspiration point, thereby beating our previous record - if only by a little.

Setting up:

As it turned out, Ron and company arrived at their site not too long after we arrived at ours.  We had arranged our departure times so that we would be arriving at or just after sunset, which occurred at about 7:57 PM on that day.  Before unloading too much gear, however, we decided that it would be a good idea to verify that, at our new location, we had a clear, line-of-sight path to Inspiration Point.  This was easily verified by sighting, through binoculars, the headlights from Ron's vehicle, even though it was still light enough to easily read by.  We quickly verified that the headlights were very visible - even with the naked eye.  This was a good omen, indeed - especially since, this time, the headlights appeared to be white in color rather than, during the previous attempt, a muddy brown, filtered through dense haze and visible only through the telescope.

Figure 2:  Computer simulated views of the paths.  Top:  Looking to the north, toward Inspiration Point.  Bottom:  Looking to the south, toward the Nebo Loop.
Click on an image for a larger version.

With the remaining daylight, Dale and I proceeded to set up the gear - a process more involved than the setup at Inspiration Point:  Our setup involved not only deploying the tables onto which the gear was set, but also an 8 inch reflector telescope as well as other optical gear that had been brought along for other experiments and tests.

Again using the 146.76 repeater on Lake Mountain for coordination, we continued our setup but this time, Chris, VK3AML was unavailable to join via IRLP owing to a schedule conflict.  On North end, setup was somewhat less-complicated:  Owing to the design of the optical transceiver, one need only open the front cover and flip it underneath the enclosure, at which point it becomes the elevation adjustment platform.  The only other steps required for assembly are to verify that the optical transmitter and receiver electronics are properly seated and aligned and that the electronics are plugged into the appropriate places, powered up, and functionality checked.  On the south end, however, setup is slightly more complicated:  The optical transceiver, being foldable, requires a few extra minutes of assembly - a process that involves the installation of about two dozen screws (with wing-nut heads) used to hold everything together and in precise alignment.  Not only this, we were also embroiled in the setup of the telescope which was not only to be used for sighting, but for emitting as well.

One of the final steps taken before we started to lock onto each other signal was to start our digital audio recorders.  For documentation and analysis purposes, both transmitted and received audio was recorded to digital audio recorders in a "lossless" PCM (.WAV) format with a sample rate of 32,000 sample-per-second.  It is important to note that the PCM format is chosen intentionally, as it accurately records each cycle's waveform, unlike a compressed, lossy format such as MP3 or WMA.  The sample rate of 32 ksps was chosen because it is high enough to capture suitable detail with good frequency response, but it takes less storage than, say, a file recorded at 44.1 or 48 ksps.

Initial lineup:

Because of our past experience, we already knew the "lay of the land" and precisely where, in the distance, that we should be pointing and with the excellent air clarity, aiming of the gear was quite easy.  Even though none of our transceivers have any sort of aiming aids (such as alignment scopes or gunsight-type hardware) it is still fairly easy to point it in the general direction by sighting the beam off the ground below and then raising the elevation.  Another helpful sighting aid is the beam itself:  Due to Rayleigh scattering, it is possible - if one's eyes have adjusted sufficiently and if it is dark enough - to see the shaft of red light going off into the distance, and it is simply a matter of pointing that shaft of light in the direction of the other end of the path.

First, we started out by having Inspiration Point roughly pointing toward us while we looked toward them with a pair of binoculars.  It only took a few seconds before we saw a brief, red flash that was easily visible with the naked eye.  After a bit more "talking in" via radio, we could see a fairly steady, red dot in the distance:  Now, it was time to switch over to the electronic aiming system.

More precise aiming:

Once we had the light from Inspiration Point "on visual," we had them modulate their end with a 1 kHz tone generated by the modulator.  Switching our end to the "audible S-Meter" mode, it only took a few seconds of moving our transceiver around before we got a "hit" - and then were able to do further tweaking of our receiver to more-precisely aim it.  Because we had already turned on our light, they could now see our end as well.

We now switched roles:  We transmitted a 1 kHz tone (while they shut their tone off) while they peaked their receiver on our signal using their audible S-meter.  After a few more back-and-forth exchanges to verify that we'd properly peaked each end, we switched to voice mode and began talking:

Initial exchange via the optical link:

• Initial 2-way exchange  (mp3, 0:47, 560 kB)  This is the first exchange, just after using alignment tones and completion of final aiming:  A brief segment of the 1 kHz alignment tone may be heard at the beginning.
  
• The audio as received at Nebo (on the south end) is in the LEFT channel.
• The audio as received at Inspiration Point (on the north end) over the 107 mile path is in the RIGHT channel.
• Into both channels - particularly into the LEFT channel - there is some crosstalk which causes the local transmitted audio to be heard in the receiver, so keep this in mind when you are listening.  You may want to mute one channel (or lift off an earphone) to hear just one end at a time.
  

Figure 3:  Waveforms showing the scintillation on a 4 kHz test tone from the Laser using the 8" telescope.  The top graph shows scintillation over a 2 second period, the middle graphs shows scintillation over a 0.2 second period, and the bottom graph shows scintillation over a 25 millisecond period.  Note that the vertical and/or horizontal scales may be different for these graphs.
Click on the image for a larger version.

Shortly after the time of this recording, we re-checked our aiming and improved signals even more and began to carry on normal conversations without any difficulty at all as demonstrated by this short clip:

• N7BDZ and KA7OEI exchange (mp3, 1:47, 1.23 Meg)  Clint, KA7OEI and Elaine, N7BDZ (and later, Gordon, K7HFV) exchange greetings across the 107+ mile optical link.  Note that the use of short duration (<30 second or 10%) music clips is considered to be acceptable fair use under current interpretations of U.S. Copyright law.  (Music:  X-Files theme by Mark Snow, DJ Dado remix)
• The audio as received at Nebo (on the south end) is in the LEFT channel.
• The audio as received at Inspiration Point (on the north end) over the 107 mile path is in the RIGHT channel.
• Into both channels - particularly into the LEFT channel - there is some crosstalk which causes the local transmitted audio to be heard in the receiver, so keep this in mind when you are listening.  You may want to mute one channel (or lift off an earphone) to hear just one end at a time.

After getting everything set up, we began to talk back and forth casually, discussing what it was that we were going to do next.

At Inspiration Point, the operation of the optical gear attracted a little bit of attention:  Because it is a popular destination for people all-terrain vehicles, riding their mountain bikes, and just to see the (ahem) inspiring view, some of the visitors were naturally curious as to what it was that was happening.  Of course, Ron, Elaine, and Gordon were happy to explain that they were "listening" to the distant red dot and that there were other, equally strange people at the other end, over 100 miles away.

Planned Experiments:

For our August 18 test, we'd planned to conduct some experiments, such as sending test tones, pictures using SSTV, and even using Lasers - but conditions, being what they were, precluded such.  This time, with the beautiful weather, we now put our minds conducting them.

Red Laser diode module - using an 8-inch telescope:

In our past tests, we compared the signal quality when a collimated red Laser emitter was used with the signal quality obtained when a similarly collimated LED was used - but using PWM techniques, as Laser cannot be conveniently modulated using linear techniques.  Using the same setup as in previous tests, operating at a wavelength of around 660 nM or so - somewhat longer than the 627 nM wavelength of the LED, but still fairly close.  I installed the laser module in the telescope and, using the audible S-meter system, was able to use the telescope's micrometer adjustments and quickly peak the signal:

• Laser with telescope (mp3, 1:17. 605 kB) Stereo audio file recorded at Inspiration Point
  
• The LEFT channel contains local audio transmitted from Inspiration Point
• The RIGHT channel contains the audio received at Inspiration point, having been transmited over the 107 mile path via the Laser in the telescope.
  
• 0:00-0:16:  Sighting-in of the Laser using the telescope.  In the LEFT channel, one can hear the audible S-meter while the RIGHT channel contains the 1 kHz "alignment" tone being received, having been transmitted via Laser, being used to "key" the audible S-meter.  In the first few seconds, one can hear the Laser "swoop" past the receiver and then get "dialed in" to peak the signal.  The "wobble" of the S-meter's tone is due to the scintillation of the received signal.
• 0:16-0:50:  Voice commentary about the lightbeam.
• 0:50-1:17:  Music clip.  Note that the use of short duration (<30 second or 10%) music clips is considered to be acceptable fair use under current interpretations of U.S. Copyright law.  (Music:  X-Files theme by Mark Snow, DJ Dado remix)

As can be heard, scintillation is more apparent than with the LED and Fresnel Lens - the main difference being a markedly faster rate-of-change of amplitude and an increased depth of the nulls.  It is also interesting to note that during the peaking, the 1 kHz tone was, at times extremely "rough" sounding - probably from far-end illumination by the edges of the Laser's collimated beam where additional distortion due to "beam wander" was evident.  In our prior 15 mile tests, it was noted that the edge of the beam produced by the telescope was extremely sharp, going from a percieved full-brightness to being invisible over a  lateral distance of just a few feet.

Analysis:

Figure 3 shows samples of waveforms of the 4 kHz pilot carrier, showing the amount of scintillation present on the received signal.  Analysis of the resulting audio and the waveforms themselves indicate that the depth of scintillation is well over 40dB.  The scintillation depth may be greater than this, but the finite signal-to-noise ratio of the received signal limits the ability of making "deeper" measurements.

Close inspection also reveals that the rate of scintillation appears to have several overlaid periods:  The most obvious period is one of around 70 milliseconds or so, but harmonics of this period can be seen in the 1.5-2 millisecond area - but the temporal resolution of the 4 kHz tone (plus the bandwidth limits imposed by the receiver and subsequent processing) limit the observation of even faster scintillatory periods that may be present.

Another interesting and aspect is the rate-of-change of the amplitude of the received signal:  Several examples can be spotted where the audio changes by more than 20dB in under 50 milliseconds.

Standard Laser pointer:

Figure 4:  Waveforms showing the scintillation on a 4 kHz test tone from the Laser pointer.  The top graph shows scintillation over a 2 second period, the middle graph shows scintillation over a 0.2 second period, and the bottom graph shows scintillation over a 25 millsecond period.  Note that the vertical and/or horizontal scales may be different for these graphs.
Click on the image for a larger version.

Another test that was planned was the use of a simple, cheap red Laser pointer.  This Laser pointer originally used a pair of AAA batteries and was bought for just $3 and uses same type of standard red Laser diode as was used in the telescope.  As in the case of the Laser module in the telescope, Pulse-Width Modulation was used.

The Laser pointer was "nondestructively" modified by inserting a dummy battery made from a wooden dowel with connecting wires, taping down the "on" button, and using thermoset glue to attach it to a small plastic box.  To simplify aiming, the box to which the Laser pointer was attached to the spotting scope mount of the 8" telescope.  Using the telescope and eyepiece, the aiming of the Laser pointer was roughly adjusted to be pointed in the same direction as the telescope itself by looking for the red dot in the telescope's eyepiece.

This Laser is a standard pointer with an aperture size of just a couple of millimeters with no other collimation optics and because of this, aiming was somewhat touchy.  As with the Laser in the telescope, the pointer was modulated with a 1 kHz alignment tone and, using feedback from the audible S-meter from Inspiration Point, after a minute or so of sweeping, I heard a "hit" as the Laser pointer flashed past the far end's receiver.  After a bit more gentle tweaking, I was able to dial the telescope's vernier adjustments to peak the signal at the far end.

• Laser pointer (mp3, 2:20, 1.07 Meg) Stereo audio file recorded at Inspiration Point
  
• The LEFT channel contains local audio transmitted from Inspiration Point.
• The RIGHT channel contains the audio received  at Inspiration point, having been transmitted via the Laser pointer over the 107 mile path.
  
• 0:00-0:29:  Sighting-in of the Laser pointer clamped to the telescope.  In the LEFT channel, one can hear the audible S-meter while the RIGHT channel contains the 1 kHz "alignment" tone being received, having been transmitted via Laser, being used to "key" the audible S-meter.  In the first few seconds, one can hear the Laser "swoop" past the receiver and then get "dialed in" to peak the signal.  The "wobble" of the S-meter's tone is due to the scintillation of the received signal.
• 0:29-0:58:  Music clip.  Note that the use of short duration (<30 second or 10%) music clips is considered to be acceptable fair use under current interpretations of U.S. Copyright law.  (Music:  X-Files theme by Mark Snow, DJ Dado remix)
• 0:58-2:20:  Voice commentary about the communications.  (There's a bit of acoustic feedback at the beginning due to my microphone gain initially being too high.)

As can be heard, scintillation is even more severe than that observed with the larger aperture telescope, yet the intelligibility is still reasonably good - mostly owing to the redundant nature of human speech and the fact that the scintillatory periods were, on average, far shorter than syllables:  This is an example of the ear and brain doing a good job of "filling in" the gaps.

Analysis:

Figure 4 shows samples of waveforms of the 4 kHz pilot carrier, showing the amount of scintillation present on the received signal.  Analysis of the resulting audio and the waveforms themselves indicate that the depth of scintillation is also well over 40dB, and again, the scintillation depth may be greater than this, but the finite signal-to-noise ratio of the received signal limits the ability of making "deeper" measurements.

Close inspection also reveals that the rate of scintillation appears to have several overlaid periods:  The most obvious period is one of around 18 milliseconds or so, with harmonics apparent in the 1.2-1.5 millisecond area.  Again, the temporal resolution of the 4 kHz tone (plus the bandwidth limits imposed by the receiver and subsequent processing) limit the observation of even faster scintillatory periods that may be present.

It can also be seen that the maximum rate-of-change of the amplitude has increased:  Several examples can be spotted where the audio changes by more than 20dB in under 20 milliseconds - somewhat faster than that observed with the Laser collimated through the telescope.

It should be noted that such an increase in scintillation is not unexpected, for it is well-known that an increase in aperture the emitting and/or receiving aperture will cause an effect known as "aperture averaging" - that is, if scintillation quashes the luminous flux at one point across the area of the aperture, it is statistically more likely that as the aperture is made larger, some other portion of it will still be intercepting some of the signal.

It is also worth mentioning that the Laser pointer - with its inexpensive, plastic lens, does not offer nearly the minimization of divergence that would be found in a higher-quality, collimated laser source using wavelength-accurate optics.  With the cheap Laser pointer, the beam need only travel a small fraction of the total path distance before it has diverged to a size larger than that of the receive aperture - a property that artificially increases the aperture size of the Laser pointer.  For the most part, the worsening of the Laser pointer's scintillation as compared to the scintillation of the beam from the telescope is that which is incurred over the first portion of the overall path.

Ironically, this also indicates that if a Laser with higher-quality optics used, the results would be even worse as the self-divergence of the beam require a longer portion of the path and be more-affected by air turbulence.  It should be noted that such a narrow divergence would be self-limited by the atmosphere, anyway:  A rule-of-thumb of 1 milliradian-per-kilometer is stated in some of the literature - a value that is widely variable, depending on many atmospheric parameters.

Notice that, unlike the case with the Laser in the telescope, there is no severe distortion present at the apparent edge of the beam - probably due to the fact that, while the Laser pointer's beam is still quite narrow, the edge-falloff of the pointer is much more gradual than that of the telescope.

High-powered LED using an acrylic Fresnel Lens:

Figure 5:  Waveforms showing the scintillation on a 4 kHz test tone from the high-powered LED using an inexpensive Fresnel lens.  The top graph shows scintillation over a 2 second period, the middle graph shows scintillation over a 0.2 second period, and the bottom graph shows scintillation over a 25 millsecond period.  Note that the vertical and/or horizontal scales may be different for these graphs.
Click on the image for a larger version.

It should be mentioned at this point that testing was done using a standard LED in the 8" telescope.  Unfortunately, the far-field luminous flux output of the LED was much lower than that of either of the Lasers and insufficient signal level was obtained to be able to make useful measurements of scintillation.

The following tests were conducted using a 3-watt red LED (peak wavelength of about 627 nM) with an acrylic Fresnel Lens.  To be certain, some reduction of scintillation was observed simply because of the larger aperture, but the bulk of the reduction was achieved through the use of a noncoherent light source as indicated through previous experimentation.

It should be noted that the previous recordings on this page (e.g. those prior to those exhibited in the above "planned experiments") were obtained using a pair of similar optical transceivers, both using high-powered LEDs and plastic Fresnel lenses.  What follows is an audio clip that contains the same music segment as the above clips, plus some general ragchewing between Ron and Dale over the link.  As can be heard, the communications was easy, "armchair" copy:

• High-powered LED with Fresnel Lens (mp3, 1:58, 927 kB) Stereo audio file recorded at Inspiration Point
  
• The LEFT channel contains local audio transmitted from Inspiration Point.
• The RIGHT channel contains the audio received at Inspiration Point, having been transmitted over 107 miles via the high-powered LED and Fresnel lens.
  
• 0:00-0:08:   Left channel:  440 Hz tone with the tone from the audible S-meter in the background.  Right channel:  1 kHz alignment tone from Nebo, across 107 mile optical link.
• 0:08-0:35:  Music clip.  The brief interruption in the audio channels was due to re-seating of the audio connectors.  Note that the use of short duration (<30 second or 10%) music clips is considered to be acceptable fair use under current interpretations of U.S. Copyright law.  (Music:  X-Files theme by Mark Snow, DJ Dado remix)
• 0:59-1:58:  Ron and Dale, ragchewing over the link.

Analysis:

Referring to Figure 5, one can see a dramatic difference in both the amplitude and rate of the scintillation.  These particular waveforms are a selected showing of the worst-case scintillation observed over a period of several minutes, with a worst-case scintillation depth of about 22dB.  Perhaps most striking is the dramatically slowed rate "dA/dT" - that is, the change in amplitude versus change in time:  The major scintillatory periods are, perhaps, 125 milliseconds in length, roughly twice the rate noticed with the telescope-collimated beam, and it is unusual to see a change of more than 15dB occurring in under 100 milliseconds.  In the "zoomed-in" 25 milliseconds portion, one can see a rather weak scintillatory period of around 9 milliseconds, but with an amplitude variation of only, perhaps, 7 dB over that time.

A few more audio clips:

Waving a green Laser pointer about:

In these sorts of outings, there is the irresistible urge to shine whatever lights one has at each other - and I happened to have a low powered green Laser pointer on hand and because the eye is many times more sensitive to green than red, I had no doubt that it would be seen.  What is interesting, however, is the sound that was heard at the receive end when the Laser swept across the receiver:

• Listening for the green Laser pointer (mp3, 1:11, 557 kB)  Stereo audio file recorded at Inspiration Point
  
• The LEFT channel contains local audio transmitted from Inspiration Point.
• The RIGHT channel contains the audio received at Inspiration point.
  
• 0:00-1:11:  This clip contains some banter back and forth between the two sites.  During this clip, however, the receiver picked up the sound as the light from the green Laser was intercepted by the receiver.  Instead of hearing a "click" or white noise, however, an odd, insect-like "whine" was heard - most notably around time index 0:50.  This particular Laser pointer is known to produce a continuous beam as noted by having "listened" to it on previous occasions.  Why widely varying pitches of "whine" were heard on this occasion is unknown.
• Note that much of the time, my voice is way off-microphone, which is why it sounds like it is the background:  Dale occasionally relays what I said to make sure that Ron had heard it.
  

A transition from Laser, back to LED:

As it turned out, I had been transmitting to Inspiration point for nearly 20 minutes while I sent music clips, test tones, and even SSTV images - and they sent similar things back.  After this, we got back to our normal back-and-forth ragchewing, so I switched from the Laser pointer back over to the high-powered LED and Fresnel lens, as can be heard in this clip:

• Laser pointer to LED cut over (mp3, 0:29, 228 kB)  Stereo audio file recorded at Inspiration Point
  
• The LEFT channel contains local audio transmitted from Inspiration Point.
• The RIGHT channel contains the audio received signal at Inspiration point over the 107 mile path..
  
• At about 8-10 seconds into the recording, there are a series of clicks and thumps as the Laser is switched off and the LED is turned back on.  As can be heard, the audio prior to this point is transmitted via Laser pointer over the 107 mile path and after that, we hear Dale and Ron talk back and forth, with Dale being received via the optical link.  Gordon can also be heard in the background, off microphone.

Figure 6:  Montages of SSTV images received via the 107+ mile optical link.  Top:  Images received at Nebo from Inspiration point - at both 200 milliamps and, later, at the full 1100 milliamps of LED current.
Bottom:  Images received at Inspiration Point via the optical links, using both LED and Laser pointer.
Click on an image for a larger version.

Switching to full power:

One of the experiments that we did was to measure our LED current and then begin to reduce it, noting the current, until we could no longer understand what each other was saying.  At about 2 hours and 10 minutes into our testing, we decided to do this just before we concluded our tests - and at this time Ron noticed something that he'd not expected:

• Going to full power (mp3, 1:14, 583 kB)  The audio as received at Nebo (on the south end) is in the LEFT channel while the audio as received at Inspiration Point (on the north end) is in the RIGHT channel.
• Into both channels - particularly into the LEFT channel - there is some crosstalk which causes the local transmitted audio to be heard in the receiver, so keep this in mind when you are listening.  You may want to mute one channel (or lift off an earphone) to hear just one end at a time.
  
• At the start of this recording, Ron had just connected a voltmeter to the current monitor points on the modulator, read the current, and noticed that up to this point, it had been set to 200 milliamps instead of the more-normal 1100 milliamps:  We hear the exchange as this is discovered...
• For most of this clip, the audio gain for the audio received at Nebo (in the LEFT channel) was set at a fairly low level to accomodate the change when the LED current at Inspiration Point was increased.
  

After this, we continued with our "LED limbo dance" (e.g. "How low can we go?") and passed random words back and forth, gradually reducing the current.  As it turns out, we both began to have severe difficulty in understanding each other as we lowered the current below 60 milliamps - a reduction of roughly 25dB in signal-noise ratio from our "normal" operating level.

Sending slow-scan television (SSTV) via the optical link:

Another experiment that we decided to try was to send slow-scan television (SSTV) signals over the optical link.  Because SSTV images are frequency-modulated at audio frequencies, there was little doubt that this would work, but we wanted to try it just the same.

While we could have taken pictures on-site and then used a laptop computer to send them, we decided that we were going to have our hands full, anyway, just doing the planned activities, so I prepared, beforehand, audio files containing the SSTV pictures.  These were generated using the MMSSTV program, recorded to a digital audio file, converted to MP3 format, and then loaded onto digital audio players.  I noted that the MP3 encoding caused some visible degradation to the SSTV pictures - a sort of weak "solarization" type of noise, but I figured that the likely amount of degradation over the optical link would be greater than this.

In the case of the SSTV files to be used by those at Inspiration Point, I simply emailed the MP3 files to Gordon for later playback.  To simulate some air of authenticity, I used pictures taken during the 8/25 expedition in addition to simple computer-generated graphics.

During the evening, the SSTV images were transmitted in both directions and recorded:  While I managed to see some of the images at the time of original transmission by placing an open microphone near the speaker, it wasn't until later that the recordings were played back and images displayed and captured.  These images are shown, as-received, with no additional noise reduction applied in Figure 6.  As can be seen, the images transmitted from Inspiration Point with a 200 milliamp LED current are noisier than those transmitted at full power - but this isn't unexpected.

It may also be noted that the SSTV images transmitted via Laser pointer look pretty much the same as those transmitted via LED, despite the extreme amount of scintillation from the former:  Because SSTV is based on FM rather than AM, it is, for the most part, resistant to amplitude variations - plus, the SSTV decoder has a bit of a "flywheel" that allows it to "fill in" very brief periods where signal is absent.

A number of image formats were used:  The small black-and-white images use the 12-second monochrome standard while the small color graphic-only images used the 24 second Robot format.  The "real" pictures are, as was mentioned, taken from the August 18 expedition, with the small versions having been transmitted using the 36 second Robot format and the large images with the PD120 (120 second) format.  I noticed, after the fact, that a complete 120 second image had not been transmitted via the Laser pointer.

The noise present in the images is mostly a result of static crashes - some of them from very distant, unseen thunderstorms, while others were from strobes from passing airplanes, warning strobes on towers, and the occasional "pop" of noise from an unknown source.  For whatever reason, the amount of extraneous noise (pops, crashes, etc.) seemed to diminish over the course of the evening:  Because the images sent at 200 milliamps were transmitted early on, they would have been "cleaner" had they been retransmitted near the end of the testing, possibly looking more like those transmitted using the full 1100 milliamps.

Trying out the "Cheap" transceiver:

The second of my optical transceivers is one that had been constructed quickly and cheaply for the sole purpose of rapidly assembling another unit to take out in the field:  After all, what good is just one unit if there isn't someone else with another one to talk to?  From this need was born an inexpensive and quickly-built optical transceiver constructed out of foam-core poster board, using page magnifier type Fresnel lenses.  Its performance has been measured as being notably inferior to that of the "good" transceivers, but we wanted to know how well it would work, so I set it up.

• 2-way exchange using the "cheap" enclosure (mp3, 1:14, 583 kB)  The audio as received at Nebo (on the south end) is in the LEFT channel while the audio as received at Inspiration Point (on the north end) is in the RIGHT channel.
• Into both channels - particularly into the LEFT channel - there is some crosstalk which causes the local transmitted audio to be heard in the receiver, so keep this in mind when you are listening.  You may want to mute one channel (or lift off an earphone) to hear just one end at a time.
• Note:  At the time of recording, the audio level being fed to the recorder at Nebo (the audio in the LEFT channel) was extremely low, and much of the noise being heard was from the audio recorder itself:  Over 50dB of amplification was added to the digitized recording in order to make its level usable.
  

As can be heard, the signal-noise ratio is lower, but communication is still quite possible.  It might also be noted that the level of the voice is noticeably lower than that of the tone - an indication that the speech compressor in the modulator (I was using the PWM circuit for this particular exchange) was not working to its full potential.

Figure 7:  Pictures from the south end of the path at Nebo.
Top Left:  Clint, doing final assembly of the optical transceiver.  Top Right:  Looking at the distant LED through the telescope.  Center Left:  The optical gear, set up on the table and operating.  Center Right:  Dale and Clint, in front of the optical transceiver.  Bottom Left:  A view toward Inspiration Point from Nebo, showing the distant LED in the background on the left, the lights of Provo on the right, and a nearby fencepost in the foreground.  Bottom Right:  The distant dot, as viewed through the 8" telescope.  (The nearby ridge may be seen in this monochrome photo.  (Photos by Dale, Clint, and the camera's built-in timer.)
Click on an image for a larger version.

Lessons learned

Testing the gear:

One of the main reasons that we wanted to re-do this path was to verify that the gear was working properly, giving us hope that under better air conditions, we would be able to span even farther distances than this 107 mile path:  While we were encouraged that we were able to complete a contact using this same gear under deplorable seeing conditions on August 18th, we still wanted to know how well this would work under more normal, clear air conditions.  We were gratified to find out that this did, in fact, work nicely.

Failure of the scintillation compensator:

One notable failure was that of the Scintillation Compensator built into the audio interface box.  When it was switched in, an effect was noted at both ends - one that Ron described as being similar to "squelch clamping" - that would occur in which audio peaks were muted with a loud click:  Despite having tested this circuit in the field on two previous occasions, it seemed that this evening's combination of audio level, audio content, and scintillation was causing something else to happen.

Fortunately, the digital audio recordings contained what was needed to diagnose and fix the problem after the fact.  Because these recordings contain "un-processed" audio, it was simply a matter of playing them back into the audio interface box to recreate that night's conditions and problems.  By doing so, it was discovered that when the AGC in the scintillation compensator was near its maximum gain, the DC offset in the variable gain amplifer's op amp - combined with leakthrough of the gain control voltage - would cause the op amp to smash into one of the supply rails.  As it turned out, a simple resistor value change was all that was required to speed up a time constant and completely solve the problem.

Adding more microphone and headphone spigots:

As per Ron's suggestion, I copied his idea of making an "octopus" box that had two microphone inputs (with a selector switch) and four headphone output jacks:  This allows several bystanders to all wear headphones - something recommended to avoid feedback between the transmitter and receiver - and have the capability of quickly and easily switching between two microphones - which, in our cases, were built into the headphones.  I also added a second "speaker output" jack to the audio interface into which one could plug headphone without muting the speaker - just in case others were nearby who wanted to hear, or if you wanted to connect an external speaker that one could place somewhat distant from the transmitter to avoid feedback - a useful feature if one is wandering around nearby, trying to reconfigure gear, but unable to be close enough to wear headphones.

A few final comments on the audio:

Using identical audio recorders made it much easier to synchronize the disparate audio tracks:  Both devices had sample rates that were extremely close to each other, minimizing the drift over time.  For synchronizing, I used the "Audacity" program - an open-source software package that is available for many operating systems that has a number of very useful features.

When synchronizing audio, I noticed that most of the time, a "pop" or crash would be heard at one end of the path but not the other.  Interestingly, however, there were a number of instances were strong single "pops" were very audible at both ends of the path.  The interesting thing about these single "pop" noises is that they did not appear to be due to lightning:  Experience gained on August 18th shows that lightning strikes always seemed to have multiple pulses and were not single-impulse events.  While it is certainly possible that these common "pop" noises are from a particular strong portion of a lightning strike - most of which consists of multiple discharges - they seemed to be solitary in nature.

In addition to the pops and clicks, there was the expected "hum" from city lights, mostly consisting of 120 Hz and 360 Hz, both being results of modulation of lights on both sides of the sinusoid with three-phase AC power.  Also related was a steady "hiss" which is the strong thermal noise component of urban lighting.  A final test was to point the optical receiver skywards, at which point the hum and some of the hiss went away, the result being that the "zero-signal" level as measured across the bandwidth of 0-4 kHz decreased by about 6 dB.

---

Additional details:

I'd like to thank those that helped, including:

- Dale, WB7FID who was with me at the Nebo end.
- Ron, K7RJ, at the far end.
- Elaine, N7BDZ, Ron's much better half.
- Gordon, K7HFV - also at the far end.

And, of course, Chris, VK3AML and Mike, VK7MJ, and the others in VK-land.

Figure 8:  Pictures from the north end of the path at Inspiration Point.
Top Left:  Gordon and Ron, connecting all of the many cables.  Top Right:  Looking at the distant LED to the south.  Center Left:  Gordon and Ron operating, with the distant LED between them and the city lights in the lower foreground.  Center Right:  The "business" end of the LED transmitter.  Bottom Left:  Ron and Gordon, in conversation with those at the other red dot.  Bottom Right:  Elaine in conversation over the optical link.  (Photos by Elaine and Ron)
Click on an image for a larger version.

At the south end of the QSO:

Present:  Clint, KA7OEI with Dale, WB7FID.

Location:  Along the Mt. Nebo Scenic Loop Road that goes between Payson and Birdseye, Utah.  This location is about 525 feet southwest of the one used during the August 18th, 2007 expedition.

WGS84 coordinates:  39°, 51' 16.9" North,  111°, 42' 14.7" West, Altitude was 9406' (2867 meters) according to GPS.

Grid square:  DM49du

At the north end of the QSO:

Present:  Ron, K7RJ with his wife Elaine, N7BDZ, and Gordon, K7HFV

Location:  A place called "Inspiration Point" that is slightly north and west of Willard Peak, which is north of the city of North Ogden, Utah - the same place as last time

WGS84 coordinates:  41°, 23' 26.6" North, 111°, 59' 9.6" West.  I don't have Ron's GPS reading for the altitude, but according to the USGS topographical maps, the altitude is almost exactly 9400 feet (2866 meters).

Grid square:  DN41aj

Distance:

The calculated distance (as a crow flies) using the Haversine method is 107.09 mi. (172.34km) using the RadioMobile program version 8.0.5.  This is about 230 feet (70 meters) farther than the August 18th expedition.

Other path statistics:

• South-to-North azimuth:  352.1° (true)
• Elevation angle at each end:  Approximately  -0.77°.  Because our altitudes were  pretty much the same, this is downward angle is due to Earth curvature.
• North-to-South azimuth:  172.0° (true)  This is slightly different than 180 degrees from the reciprocal bearing due to rounding off.
  
• Maximum difference in elevation along path:  Approximately 5150 ft. (1570 meters)

Equipment common to both sides of the QSO:

• The LED was amplitude modulated with a current-linear modulator with a resting current of 1.1 amps.  Details of the modulator are here:  LED_linear_modulator.html
  
• The transmit LED in both cases was a Red Luxeon III emitter module (Lumileds M/N:  LXHL-PD09) epoxied to a heat sink.
• The optical receivers were my "version 3" design, described here:  optical_rx1.html#ka7oei_rx_ver3 with both receivers using BPW34 photodiodes.
• Audio interface units, incorporating audio amplifiers, audio recorder interface, audible S-meter, and a few other features were used - details are here:  optical_comm_audio_interface_device.html
  
• Both transceivers have separate and identical TX and RX lenses mounted side-by-side.
• Digital audio recordings were done on both ends of the path using an Insignia NS-DV4G portable audio player.  This device can play not only audio files (such as MP3, WMA, OGG, and WAV) but it will also play videos, show still pictures, receive (and record) FM radio broadcasts and, most importantly on this occasion do an excellent job of recording audio from a stereo line-in connection.  Although capable of recording compressed audio in WMA format, we used the (uncompressed) WAV format to yield a 2-channel, 16 bit audio with a 32,000 sample/second rate.  At this sample rate, well over 5 hours of recording is available on the 4 gig storage space - with the caveat that one must not allow any single audio file to exceed 2 gig of file length - a precaution exercised by occasionally stopping and restarting the recording to a new file.
  

Optical transceiver used on the North-to-South link:

• This enclosure is described in detail here:  Optical_enclosure_first_version.html
• Lens size:  Unmounted, the Fresnel Lenses are 250mm x 318mm and have a focal length of 318mm.  The mounting frames vignette the lenses by about 10mm in each dimension, so the available lens area is about 240mm x 308mm.  Each lens is protected by a sheet of Plexiglas and the front surface has been coated with a protective polymer to prevent scratching and moisture accumulation.
• For optimal far-field optical flux density, a glass PCX (Plano-ConveX) lens is used in front of the LED to appropriately illuminate the Fresnel, the LED-Lens distance being set empirically for best output.

Optical transceiver used on the South-to-North link:

• The enclosure is described here:  Optical_enclosure_foldable_version.html
  
• Lens size:  Unmounted, the Fresnel lenses are 404mm x 430mm and have a focal length of 229 mm.  The mounting frames vignette the lens by about 10mm in each dimension, so the available lens area is about 394mm x 420mm.  Each lens is protected by a sheet of Plexiglas and the front surface has been coated with a protective polymer to prevent scratching and moisture accumulation.
• For optimal far-field optical flux density, an optical acrylic DCX (Double-ConveX) lens was reground to an aspherical shape to provide optimal illumination of the Fresnel.  This turned out to be necessary owing to the very short focal length of the lens that made it difficult to efficiently illuminate the lens.  After adjustment, this LED/Lens combination produces about 25% higher far-field flux than the other assembly, with an almost identical half-power beamwidth.
• For modulation of the Laser and the LED used in the telescope - and for the "Cheap" cardboard enclosure, the Pulse-Width modulator was used.
  

Notes about the audio clips on this page:

• The audio clips on this page have been edited to remove "dead" time and irrelevant bits of dialog.  This editing has been done solely to make them more "listenable" and to keep the file sizes manageable.
• In the audio clips, amplitude and gain adjustments have been made to improve listenability.  At the time of the actual event, the volume control was used to similar effect for the benefit of the local listeners.
• Except as noted, no noise reduction or audio filtering has been done, other than some lowpass filtering that was done during the MP3 encoding process.
• For the audio clips transmitted via Laser, the audio level was reduced to prevent clipping during the occasional bright peaks.
• With the LED, the average audio level could be higher, owing to the lower amount of scintillation - a fact that brought up the background noise to a higher level.
  

Return to the KA7OEI Optical communications Index page.

If you have questions or comments concerning the contents of this page, feel free to contact me using the information at this URL.
Keywords:  Lightbeam communications, light beam, lightbeam, laser beam, modulated light, optical communications, through-the-air optical communications, FSO communications, Free-Space Optical communications, LED communications, laser communications, LED, laser, light-emitting diode, lens, fresnel, fresnel lens, photodiode, photomultiplier, PMT, phototransistor, laser tube, laser diode, high power LED, luxeon, cree, phlatlight, lumileds, modulator, detector