A high power, high-sensitivity remote repeater covering deep river gorges in Utah

From the late 1950s until about 2021, there was a (mostly) annual event held in southeastern Utah that was unique to the topology:  The Friendship Cruise.

The origins are approximately thus:  In the late 1950s, an airboat owner - probably from the town of Green River, Utah - decided to go down the Green River, through the confluence of the Green and Colorado rivers, and back up to the town of Moab.  Somehow, that ballooned into a flotilla in later years - with as many as 700 boats - in the 60s and 70s.  By the mid 90s, interest in this unique event seemed to have waned and by about 2012, it seems to have petered out.

Communications is important:

Figure 1: A high-Q 80 meter magnetic loop on one of the rescue boats Click on the image for a larger version

From the beginning it was realized that there was a need for the boats and support crews to be able to communicate with each other - but the initial attempts using CB and/or public safety VHF radios were unsuccessful, reaching only a few miles up and down the river - not too surprising considering that most of the course runs through winding, deep gorges.  In later years, cell phones - and even satellite phones - were tried, but due to the remoteness and narrowness of the gorges they were of extremely limited use.

At some point, probably in the mid 1960s, amateur radio operators involved, successfully closing the link using the 80 meter amateur band.  This tactic worked owing to the nature of 80 meters:  During the daytime, coverage is via skywave over a radius of about 200 miles (300km) and this high angle of radiation allowed coverage into and out of the deep canyons.  Furthermore, small antennas on this band were well-suited for radiation of RF energy at these steep angles.

For (literally!) decades, this system worked well, providing coverage not only anywhere on the river, but also to the nearby population centers (e.g. Salt Lake City) where other amateur radio operatyors could monitor and relay traffic as necessary and summon assistance via land line (telephone) if needed.  Because the boats were typically on the river only during the day, this seemed to be a good fit for the extant propagation.

While it worked well, it was subject to the vagaries of solar activity:  An unfortunately-timed solar flare would wipe out communications for hours at a time, and powering and installing a 100 watt class HF transceiver and antenna was rather awkward.  Occasionally, there was need to communicate after dark, and this was made difficult by the fact that 80 meters will go "long" - often requiring stations much farther away (e.g. in California or Nebraska) to relay to stations just a few 10s of miles away on the river!  Finally, it was a bit fatiguing to the radio operators and boat owners to have to listen to HF static all day long!

Enter VHF communications:

Figure 2: General coverage map of the course showing coverage of various sites. Click on the image for a larger version

While VHF communications had been tried early on - and had been available in the intervening years - the biggest problem was that these signals could not make their way along the river for more than a few miles between twists and bends in the deep river gorges.  While useful for short-range communications, it simply wasn't suitable for direct boat-to-boat communications along the vast majority of the river's course.

By the time that the 1990s had come along, there was renewed interest in seeing of we could make use of VHF, on the boats, on the river.  The twist was that instead of direct communications between boats, we would try to relay signals from far above, on the plateaus, and a few experiments were tried.  It 1996, I was on a boat on the river and took notes on what sites covered and where, trying nearby mountaintop repeaters and temporary stations set up at places near-ish the river courses themselves - the resulting map being presented in Figure 2.

Using the color-coded legend across the top and the markings on the map itself, one can see what sites covered where.  Included in this was the coverage from the 147.14 repeater near-ish Green River, Utah, the 146.76 repeater near Moab, and several other sites atop the plateaus surrounding the river.  As can be see, coverage was spotty and inconsistent over much of the route - with the exception of a site referred to as "Canyonlands Overlook"(abbreviated "Cyn Ovlk") which commanded a good view of the Colorado River side of the river course.  Clearly missing was reasonable coverage in the depths of the gorges along the lower parts of the Green River side - which started, more or less, where the coverage of the "Spring Canyon"(abbreviated "Spring Cyn") stopped.

Figure 3: The two TacTecs used for 2 meter reception, the voting controller (blue box) and the FT-470 used as the UHF link radio. Click on the image for a larger version.

As it happened, there were amateur radio operators camping at a site called Panorama Point when I was on the lower Green River and because we were using the Utah ARES simplex frequency, they just happened to hear the simplex activity on the river.  At that moment, I happened to be in areas that were not well-covered by any of the other sites and while their signals weren't extremely strong, it made me wonder what could be accomplished should I wield both gain antennas on the receiver and high power and gain antennas on the transmitter of a 2 meter repeater.

The birth of a repeater:

During the next year I put together a system that I'd hoped would make the most of the situation.  Because of the remoteness of the site, it had to be relatively lightweight - and I also wanted to avoid the use of any duplexers (large cavity filters) that would add bulk and - more importantly - losses to the system.  Taking advantage of a weekend to visit Panorama Point the next spring we determined that we could split the transmit and receive portions by about 0.56 miles (0.9km) apart, placing the receive antennas behind some local geographical features.  The back-of-the-envelope calculations indicated that this amount of separation - and the rejection off the backs and sides of the beam antennas - would likely be sufficient to keep the receiver out of the transmitter.  The receive site - surrounded by three sides by vertical cliffs - also provided a commanding view of the terrain as can be seen in Figure 5, below.

Figure 4: GaAsFET preamplifier mounted right at the receive antenna to minimize losses. Click on the image for a larger version.

In addition to site separation and gain antennas, I decided to go overboard, adding mast-mounted GaAsFET preamplifiers, right at each antenna, and implementing a voting receiver scheme - something made much easier with the acquisition of two, identical RCA TacTec "high band" VHF transceivers.  These receivers were modified - clipping the power lead to the transmitter and adding a 3.5mm stereo plug to each radio to bring out both discriminator audio and the detector voltage from the squelch circuit.

A relatively simple PIC-based repeater controller was constructed, using a simple comparator to determine which receiver had the "best" signal, based on the detector voltage from the squelch circuit, and also using another set of comparators and onboard potentiometers to set the COS (squelch) setting for the receivers.  As it turned out, the front-panel squelch control adjusted the gain in front of the squelch detectors, allowing each receiver to be "calibrated" from that control, allowing easy fine-tuning in the field.

To link the receiver site to the transmitter site, UHF was used and I modified my old Yaesu FT-470 handie-talkie to this task.  The mysterious rubber plug on the side of this radio was replaced with a 3.5mm jack, providing a direct connection to the modulation line of the UHF VCO while using the top panel 2.5mm external microphone jack for transmitter keying.  As it turns out, not only did this transmitter provide linking to the nearby transmitter site, but its UHF beam was pointed across the way, to another 2 meter repeater at Canyonland's Overlook that provided coverage on the Colorado River - providing what amounted to a linked repeater system.

Figure 5: The remote RX site, surrounded on 3 sides by sheer cliffs.  The mast has two 2 meter and one UHF link beam antenna.  The solar panels are just visible along the far right edge. Click on the image for a larger version.

One of my goals was to minimally process the audio, causing as little "coloration" as possible, and to this end I took the receivers' discriminator audio from the voter and put it directly into the modulator of the UHF link radio, completely avoiding the need for de-emphasis and pre-emphasis.  This worked pretty well - but I noticed during the first time that it was used that when weak signals were present on the input, the noise and hiss from weak signals would sometimes cause "squelch clamping" on the receivers being used by us and others owing to the fact that such noise was being passed along the link without alteration, causing the addition of a 3.5 kHz low-pass filter to remedy this the next time it would be used.

The receive site itself was solar-powered, using lead-acid batteries to provide the power when insufficient sun was available (e.g. heavy clouds, night).  In later years, the PIC controller was modified to not only read the battery voltage, but to connect/disconnect the solar panels from the battery bank in a "bang-bang" type charger but also to report the battery voltage when it identified.  In this way, we could keep an "eye" on things without having to walk out to the receive site.

The two 2 meter and the 70cm link antennas were mounted on a single mast, the VHF antennas pointed in different directions to take advantage of the slight difference in physical location and in the hopes of providing diversity for  the weak signals from the depth of the canyons - which were all reflections and refractions.  As it turns out, despite the close proximity of the antennas, this worked quite well:  At the site, one could monitor the speakers on the receivers and watch the voting LED and see and hear that this simple, compact arrangement was, in fact, very effective in reducing the number of weak-signal drop-outs caused by the myriad multipath.

In testing on the work bench, the measured 12dB SINAD sensitivity of each of the receivers was on the order of 0.9 microvolts - far and away better than a typical receiver.  Later, I did the math (and wrote about it - see the link at the bottom of this article) and determined that it was likely that the absolute sensitivity of this receiver was limited by the thermal noise of the Earth itself and that it could not, in fact, be made any more sensitive.  This notion would appear to be borne out by a careful listening to the repeater in the presence of weak signals:  Very weak signals - near the receive system's noise floor - sounded quite different than what one might hear on a typical FM receive system near it's noise floor.  Instead of a "popcorn" type noise, signals seemed to gradually disappear into an aural cloud of steam.

The transmitter site:

Figure 6: The transmit site.  The tall (30 foot) mast and 2 meter transmit antenna is visible in the background with the UHF link antenna and the VHF "backup" TX antenna in the foreground. Click on the image for a larger version.

With so much effort having gone into maximizing receiver performance, I decided to do the same on the transmit site in the years that this system was used.  For the first year, the transmitter was modest:  A Kenwood TM-733, on low power, driving a 50 watt RF amplifier into a vertical - but the next year I decided to erect a taller mast and place atop it a 5 element beam.  To boost my RF output power, I scavenged a pair of 110 watt RF amplifiers from some ancient Motorola Mocom 70 mobile radios (with some DC fans) and combined them with a Wilkinson Power divider, yielding a bit over 200 watts of RF and about 1500 watts of ERP (Effective Radiated Power) - all without causing any measurable desensitization of the receiver system.  After a few days, one of these amplifiers failed, but the remaining 110 watt amplifier happily chugged along.

The next year I acquired a 300 watt Vocomm amplifier and was able to use it for the remainder of the times that the Friendship Cruise was held.  Requiring 50 watts of drive, I still had to use the 50 watt amplifier, driven by 5 watts from the TM-733 to attain the full RF output.  When keyed down, the entire transmitter system drew about 60 amps at 12 volts, requiring frequent topping-off by a generator and DC power supply that were brought along.

With that much transmit power, the antenna was held aloft by a 30 foot (9 meter) mast to keep it well clear of people - and to help clear any local terrain and its effects.  As can be seen in Figure 6, there was a second mast with the UHF link antenna and a "back-up" 2 meter antenna.  When we arrived at the site, the first order of business was usually to set up the receive site, but once back at camp, we used a radio in cross-band mode and the two antennas on the short mast to get it on the air, providing "reasonable" transmit coverage.  Because of the effort required to set up the tall mast, battery bank and power amplifier, we often waited until the next morning to complete the setup, bringing our radiated transmit power up to its full glory!

"Listening" on the link frequency, this transmitter not only relayed my own, nearby receive site, but also the "other" repeater at Canyonland's Overlook. 

How well did it work?

The Panorama Point repeater itself worked better than we could have hoped:  It was "reachable" nearly everywhere on either the Green or Colorado River - although some sections of the upper Green and Colorado had somewhat weaker signals.  Unexpectedly, it also provided coverage into the town of Moab, as far north as Price, Utah and even down near Hite, Utah - both well outside of its expected coverage range.

With about 2.5kW of ERP, one would expect that this repeater would have been an "alligator"(all mouth, no ears) but this was not the case:  When its signals were weak - as in a deep river gorge - using a 50 watt mobile radio, the transmitter and receiver seemed to be more or less evenly matched, and despite running this much power, we did not experience any detectable "desense" where the strong transmit signal would overload the receiver.  At least part of this was attributed to the receivers themselves:  The RCA TacTec receivers used only modest amounts of RF gain and a passive diode-ring mixer.  I have little doubt that if we had used more "modern" receivers, we would have experienced overloading and would have had to place notch cavities, tuned for the transmit frequency, between the GaAsFET preamps and the receivers.

As a system, the Panorama Point and Canyonlands Overlook repeaters completely replaced the need for HF gear on the boats in the last decade or so that the Friendship Cruise was held, providing nearly seamless coverage from start to finish.

* * *

Additional (related) articles:

• "The Most Sensitive Repeater On Earth?"(link)  - an article with more technical details about the link margins associated with the receive system.
• "Clint's Friendship Cruise Page"(link) - pages describing the system, a bit of history, and other details.

This page stolen from ka7oei.blogspot.com

[End]