 |
The UARC 146.62 Repeater System |
|
| Farnsworth
Peak Linked to the 146.62 repeater on Scott's Hill |
| |
The UARC 146.62 Repeater System |
|
| Farnsworth
Peak Linked to the 146.62 repeater on Scott's Hill |
Figure 1: A view to the north, showing the 02/62
repeater's receive
site on Farnsworth Peak -
with the "old" antenna system.
About the 146.620 Farnsworth Peak repeater
 |
Because of its location, Farnsworth Peak is one of the best transmitting and receiving sites along the Wasatch Front: Its location near the north end of the Oquirrh Mountain range west of Salt Lake City has a clear view in that direction as well as to the west and south into Utah county. Because of this, it is heavily used for transmitting radio, television, and other services. In fact, the Farnsworth peaks have most of the Salt Lake FM broadcast and television stations!
Being such a good transmitting site is not without its problems: The multitude of transmitters on a myriad of frequencies has caused the noise floor of the site to increase, thereby reducing the effective sensitivity of any receiver on-site - especially at VHF. Because of this, when the site was surveyed for amateur repeater use back in 1979 it was noted that down the mountain some distance (about 400 feet) there was enough isolation from the transmitters to adequately reduce the noise floor. When the repeater was finally built, a remote receive site was thus established.
It might have occurred to the reader that the '62 repeater is
different
from "most" repeaters that one runs across: The Transmitter and
Receiver
are in different locations! Normally, both the transmitter and
receiver
share not only the same antenna, but the same feedline, duplexer, and
cabinet.
While that could have been done in this case, the intrinsic noise level
of this site would make receive performance suffer. For this
reason,
the receiver is located located at that aforementioned site - some 400
feet north of the complex, down the
hill
some distance and this site can be seen in figure 3, 4 and 8. Also,
read "A Brief History of
the '62 Repeater", below.
 |
At the main Farnsworth complex is the repeater cabinet as seen in
Figure 5. One thing that every repeater needs (in order to be
legal, at least)
is a means of controlling it. This could be something as simple as a
switch
(if the repeater is, say, located at someone's house - or some other
location
that is staffed continuously) or a telephone line that allows remote
control.
One of the most popular ways to provide repeater control, though, is
with
control receiver. The '62
repeater
complies with this rule by having not only a switch that can be flipped
(it turns out that Farnsworth Peak is staffed all of the time anyway)
but it has a 222 MHz control receiver.
The repeater controller provides functions such as the timeout timer, the ID and its timer, audio switching and squelch control, as well as decoding the control sequences received on the control receiver. One of the unique features of the '62 repeater is the ability to remotely set the squelch of the receiver - a big improvement over having to drive up there just to turn a knob, or trying to convince one of the already-busy staff members up there to do it for you
Another device is the "Voting Controller." This box can accept
signals from up to eight receivers - all listening to the repeater's
input
frequency. By comparing the signals with each other, it can tell
which one is the "best" signal - something it does by analyzing the
amount of noise on it - and then selects it to be transmitted,
automatically. One of these receivers is, of course, the one in
Figure 4 at Farnsworth, but the other receiver is on
Scott's Hill and it arrives via a 70cm link radio (in the very top
position
in the rack.) When transmitting, the same signal being sent out
on 146.620 from Farnsworth is simulcast on 146.620 on Scott's Hill as
well.
Connected to the exciter is a gizmo called a "Disciplined Oscillator." While this sounds fancy, it's purpose in life is mainly to keep the 2-meter transmitter precisely on frequency and it does this by locking to high-stability oven-controlled oscillator. Why the need for frequency accuracy and stability? Because the transmitter on Farnsworth operates on the same frequency as the one on Scott's! By having the two frequencies set precisely with respect to each other they don't "fight" each other nearly as much, allowing one to go from the coverage area of one site's transmitter to another with minimal interference.
The audio and keying lines from the controller find their way to the exciter (a fancy word for "low power transmitter", in case you were wondering...) The exciter generates the 146.62 transmit signal and modulates the audio on it, but its ultimate output power is just a couple of watts, so there is a 100 watt power amplifier that is used to bring the power up to a more respectable level. With the antennas being used, this 100 watt level is just about right to allow the the repeater to be heard as well as it can hear.
To have too much transmit power would make the repeater an "alligator " (all mouth and no ears) while running too little power would make it an "elephant" (ears too big...) Either situation in extreme make the repeater less pleasant to use as too few people can get into an "alligator" repeater while too few people can reliably hear a repeater that is an "elephant." Clearly, if one has to choose between an "alligator" or an "elephant" one should always pick the "elephant" situation as it is more common to have people who have trouble getting into a repeater than hearing it!
If Farnsworth Peak is anything, it is busy - at least RF-wise. It is host to, by far, the largest concentration of broadcast transmitters in the state and this means that there is a tremendous amount of energy floating around in the air. Ironically, even though it is a great site for transmitting, it is a terrible site for receiving - at least on 2 meters and it is because of this that the receive site is remotely located. (See the pictures above, and the history, below.) To keep ourselves in good graces with the landlord, as well as to protect our own equipment and to radiate a clean signal, the transmitter has several cavities on its output and it is these cavities act as very narrow filters to pass only the transmit frequency.
There is also a circulator
located
between the transmitter and the cavities. This
device
allows the transmitter's power to flow from the transmitter to
the
antenna feedline, but any power that comes back from the antenna (such
as that resulting
from an SWR or other transmitters on other frequencies being picked up
by the transmit antenna) are sent into a dummy load. This keeps the
transmitter
from every seeing a bad load - which might happen if the antenna were
to
be iced-over or damaged - which could destroy the amplifier.
More important that protecting the transmitter, it also keeps the
signals from other transmitters from getting into the
power amplifier and causing spurious signals from being created -
something that we wouldn't want to have happen!
 |
For many years the '62 transmit antenna was located on the old,
wooden tram building and at the time it was a J-Pole made from
galvanized pipe - a material that was known to be able to withstand the
rigors of the winters. Much later, the site was expanded,
requiring that the building for the now-abandoned tramway be torn down
and the '62 transmit antenna was replaced with a MaxRad whip/fiberglass
antenna and relocated to its current position. After a few years,
the whip on the MaxRad disappeared - probably fatigued and broken by
the wind (a common problem with that particular antenna,
apparently...) and it was replaced with a copper-pipe J-Pole.
This antenna served well for a number of years until, early in 2009,
it was noted by users and direct measurement that signals from
Farnsworth were weaker - by as much as 8dB in some Salt Lake Valley
locations. After determining that the power amplifier was not the
issue, it was decided that both the aging feedline and the J-Pole be
replaced. Fortunately, some dual folded-dipole antennas had
become available and one of these was installed in the fall of 2010,
restoring most of the decreased signal level that had been noted.
What had caused this lower signal level? We had, in fact, noted
that the feedline had degraded - particularly the short jumper on the
J-Pole itself. While that would explain at least some of the
change in signal level it's also possible that there were other effects
caused by new "clutter" (added masts, cable raceways, etc.) on site
that might cause some blockage of the signal in certain directions -
mostly toward the Salt Lake Valley itself. This "new" antenna -
being physically larger than the original J-Pole and thus radiating
signal from a larger area - should be somewhat less-affected by such
blockage - plus, it has about 3dB higher gain (that is, twice
the signal is radiated) as compared to the J-Pole - which won't hurt,
either!
The idea of a repeater on Farnsworth Peak began to emerge shortly after UARC's success in putting the 146.76 repeater on the air in the mid 70's. Some hams who had had experience in the television industry called Farnsworth "the best site in the state of Utah." The 9000-foot mountain sits near the north end of the Oquirrh range which is just west of the Salt Lake valley. It certainly appeared that it had a wonderful view of the valley and a chance of getting into Logan.
One sour note kept coming out, however. There were reports that the site had bad noise and intermodulation problems. So a party was assembled to go to the mountain and check it out. Permission was obtained from KSL-TV, owners of the site, to make a visit, and a spectrum analyzer was obtained. The party headed for the 9000-foot level.
They found that the reports of receiving problems, unfortunately, had been correct. As well as big spikes that the analyzer showed popping up randomly from intermodulation products, it showed a lot of "grass," a low level noise floor that would surely mask any weak signals. On the positive side, the view from the site certainly looked promising. The whole Salt Lake valley was laid out before us; most of Utah valley was easily visible; we could look straight into Grantsville; and the view to the northwest, across the Great Salt Lake toward Idaho, seemed to go on forever before disappearing in the haze.
In the months that followed, a partial solution appeared when one of the local hams offered access to nearby Kessler Peak, just north of Farnsworth. Although not quite as high as Farnsworth, this peak had no television station and probably did not suffer from the noise problem. It had its own set of disadvantages, though. Space was severely limited, so only a small antenna could be accommodated. Management concerns were such that we would be able to access the site only when the ham that worked there was available.
We thought a hybrid arrangement could be devised. Why not put the transmitter for the new repeater and its logic on Farnsworth, and put the receiver on Kessler? We could take advantage of the lower noise level on Kessler and the greater accessibility of Farnsworth at the same time. The rough spot here was that some sort of link would have to be built to send the receiver audio from Kessler to Farnsworth.
We started procuring equipment for the new repeater, keeping in mind
that transmitter and receiver might be separated in the final
configuration.
About this time, Dirk Ostermiller, W7KCC, the club's repeater engineer,
who had spearheaded the '76 project, found that his time was
overcommitted.
He resigned and left the Farnsworth project to others to complete.
Randy Finch, K7SL, agreed to take the position, and started
assembling
the new equipment into a working repeater. Before long, a new 146.61
MHz
repeater appeared on the air from Randy's home in Magna. It came on
just
in time to be announced at the 1979 UARC Christmas Banquet. (The switch
to 146.62 came a few years later when Utah adopted 20 kHz spacing, the
plan pioneered by Washington and Oregon.)
By the time we got that far, another monkey wrench seemed to have fallen in the works. (We thought we could blame it on Murphy, but it wasn't even close to Field Day.) Other commercial installations had consumed most of the remaining space at the Kessler site, and it was no longer available to us.
After some amount of hair-tearing and carpet-pacing, the idea of a split site at Farnsworth was born. We contacted people at KSL to see if they would go along with the idea of letting us just plop our receiver outdoors somewhere away from the building (and the noise). They told us their property ended on the south just a short distance from the TV buildings, but it extended quite some distance to the north, and something north of the building might just work out.
It would have cost us too much to do Class 1 wiring to bring power
down
to the remote site. So placing the entire repeater there was not
practical.
But just a receiver could run happily on 12V at 100 ma. So a split-site
repeater seemed to be the answer.
 |
Next, we started walking down the ridge to the north, away from the buildings. The S-meter started to drop, just as we had hoped. It finally reached a zero reading about 400 feet north of KSL's northernmost building, the tram shack. We ceremoniously announced, “The receiver goes here!”
Of course, declaring it so and making it happen were two different things. Many man-hours from a large number of volunteers were required to finally place a repeater on the mountain. Steve Kleinlein, WC7G, built a power supply. Larry Jacobs, WA7ZBO, built two J-pole antennas. Russ Michaelson, N7SM, built rack-mounts for several pieces of equipment. Steve Berlin, WB7VCI, built the audio board for the controller. Don Richardson, WA7QKF, built the logic board. Mac MacDonald, WA7SVN, built an ID board. Randy built many of the remaining pieces and coordinated numerous work parties on the mountain and in his basement.
Work continued. Copper for a ground system was obtained. Burial cable that could be used to connect the transmitter site to the receiver site was located. A rack was found to use for the transmitter end of the repeater.
A space was negotiated with the KSL staff for the transmitter. It
was
in a dark corner on the upper level of the old tram building. The
brake-release
arm was conveniently located where one could sit on it to work on the
repeater.
Only a thin layer of metal separated us from the elements outside, and
this metal had a number of holes in it. We would have to provide some
rain
and snow protection for our rack. A plastic cooler cover finally met
this
need.
As bizarre as this space was in some ways, it had some features that made it just what we needed. It was at the end of the KSL buildings closest to our chosen receiver site, and it was not likely that any commercial clients would be competing with us for it.
The solution to one problem fell into our laps unexpectedly. A gentleman had donated an old commercial repeater to the club. He had planned to use it at a particularly good mountain site. Before he got his repeater on the air, however, his work took him out of the country for several years. When he returned, he found that someone had already put a repeater on at the Snowbird ski resort, so he donated his equipment to UARC. It was old tube gear and the receiver did not have a particularly good reputation, so we thanked him kindly and put the unit in storage.
But as the construction of the Farnsworth repeater proceeded we
remembered
one very useful feature of the donated repeater: it had a weatherproof
rack! Soon a worthy recipient for the transmitter and receiver strips
was found and their rack became the new home of the Farnsworth
receiver.
How time flies when you're having fun! There was snow on the ground at Farnsworth again by the time we were ready to erect a tower for the remote receiver. It took several weekend trips to get a hole dug, and about 15 hams to actually erect the tower. We had to carry up, not only the cement, sand, and gravel, but also the water. The final fifty feet of the journey was uphill and had to be done on foot.
A few more weekends were devoted to burying the cable that would
connect
the two sites. It was a multi-pair cable and the design called for one
pair to carry 12-volt power down to the receiver, another to bring
audio
back up, and a third to bring squelch information up. It sounded easy
enough
to dig a trench, but most of the ground turned out to be not soil, but
stones on top of bedrock. We settled, frequently, for just piling rocks
on top of the cable.
 |
Perhaps dropping the RF gear on the road is a necessary part of the initiation of UARC repeaters. Several years earlier, the '76 repeater had been carried to its site in a truck supplied by the National Guard. Part way up the 30% grades, the whole rack fell out.
The moment of truth for the Farnsworth project finally came on a
Saturday
night in the fall of 1980 when two different teams worked at hooking up
the cables at the transmitter and receiver sites. The power and squelch
pairs had to be connected with correct polarity, but no particular
color
standard had been agreed to by the two ends. We figured we would just
hook
it up any old way, take some voltmeter readings, and then reverse
whichever
pairs were wrong. The crew at the receiver end was ready first, but
couldn't
do much until the transmitter crew fed them some power down the cable.
So they waited impatiently, making occasional wisecracks on simplex.
Finally,
the transmitter team got ready to call on '61 simplex and announce that
power was on. They never got the chance to make the call.
Suddenly, the repeater was on the air, and there was a station using it from Snowville, Utah, near the Idaho border. Magically, the right polarities had been found on the first try. Why someone in Snowville happened to be trying the 01/61 repeater pair at that moment is still mysterious. But, the repeater was on the air! It wasn't the best-sounding repeater in the world. In fact, its audio was troubled by mysterious hums, squeals, and bursts of audio from the FM broadcast stations on the site. But after another hour of cable routing, bypassing, and level adjustment, it began to sound like a usable repeater.
It's very traditional for new repeaters not to make it through their
first night. This one survived its first night, but its first day was a
different story. When things came up to daytime temperatures, it became
clear that the squelch was too loose, and the repeater was transmitting
noise for long periods. The author and Scott Bidstrup, WA7UZO, made a
trip
and cured the problem.
After that, the repeater ran happily through the winter. Usage was light, but gradually, a few people began using the machine. We thought our job was done.
The real excitement didn't start until late spring when the first
thunderstorm
came by and the repeater promptly went off the air. We went to the
mountain
and repaired some power supply components and added some surge
suppression.
The next week another storm came by and the repeater was off again. We
went to the mountain and replaced some cable-receiver components and
added
more surge protection.
Soon, it became the joke that '61 was a better weather predictor than the Weather Service. It went off the air at the first hint of a storm coming in from the west. This seemed unfair, because, by this time, we had beefed up the circuitry to the point the repeater only went off every second or third thunderstorm. Sometimes it would be on for over two weeks at a stretch. However, some of the IC sockets were wearing out because the parts in them had been changed too many times.
Our problem seemed to revolve around the fact that a 10,000-amp
lightning
stroke parallel to our cable (the one connecting the transmitter site
to
the receiver site) could easily induce enough voltage to destroy the
ICs
at both ends. We finally realized we had to convert to a new religion
and
espouse a creed the phone company had known about for years: The
way
to minimize induced voltages in a balanced pair is to make sure there
is
no ground reference. The cable pairs would have to float and have
no
ground connection.
To practice our new belief, we had to make sure everything that went from one site to the other had to couple through transformers at both ends. That, in turn, meant everything had to be a-c. That was no problem for the audio -- it was a-c already. For the power, it meant we had to wind a couple of unusual transformers and feed the line with 60 Hz. The hardest one to convert was the squelch line which carried DC proportional to the receiver's noise detector voltage.
Just a little too sure of our abilities, we decided we could do the conversion one evening after work. I breadboarded an op-amp circuit to convert the squelch circuit's DC to a pulse-width modulated signal. Randy rounded up transformers and built a more permanent model of the circuit. We enlisted Dale Jarvis, WB7FID, and headed for the mountain. We didn't leave until daylight was in the east, a scheduling feature we had neglected to tell Dale about. The famous “modification 42” was complete.
The repeater was much more reliable, now, reaching a rating of 6
MTSBF
(mean thunderstorms between failures). But it took modifications 43 and
44, involving huskier receiver and driver transistors, respectively,
before
it really settled down. Thus, by the end of 1981, the thunderstorm
problem
was pretty well behind us
 |
Go back to the UARC repeater information page...
This page last updated on 20110110