UARC, the Utah Amateur Radio Club, is proposing to implement a linked repeater system to increase coverage of the 146.62 repeater.  (Click here to read more about this repeater.) The existing repeater is on Farnsworth Peak, about 18 miles southwest of downtown Salt Lake City.  Currently, this repeater covers down south to Nephi, to the west, well past Wendover, Nevada, and to the north, well into southern Idaho.  To the east, however, coverage is severely limited.

Farnsworth Peak is in the Oquirrh (pronounced "Oh-Kerr") mountain range - a range of mountains with peaks mostly in the 9000 foot elevation range.  The Wasatch mountain range, along the east side of the Salt Lake valley has many peaks that are over 11000 feet which severely limits coverage to the east for a repeater located on the Oquirrh range.

There are already several repeaters in the Wasatch range. One of these is the 147.18 repeater, which has been atop the Snowbird Ski Resort tram (at 11000 feet) for over twenty years.  This repeater has proven its worth in its ability to cover the high-valley communities to the east reliably over the years.

When the opportunity arose for UARC to be able to obtain permission to place a repeater in the Wasatch range as well, we jumped at it.  After considerable red tape, UARC has finally gotten the go-ahead to begin work.

The site (on Scott's hill, which is west and north of Guardsman Pass) consists of a building that has been abandoned for 20-plus years that is nestled among several other buildings (pictured on the left.) The first order of business is to secure the building, perform repairs and aesthetic improvements.  (Click here to see more pictures of this site.) This work has already begun! The old wooden generator annex and the poles have been removed, the entry on the south side has been blocked up, and many of the through-holes in the building have been patched. The building still needs more work though.

Only preliminary planning has been done to date and, as the certainty of being able to secure this site was, until recently, in great doubt, we have not made plans to install very much equipment on-site this year.

But we have done enough preliminary planning to know have a general idea of how the system will work:

One of the advantages of FM (Frequency Modulation) over other modes is that, in the presence of two signals of different strengths, the stronger of the two will override the other without much deterioration of the stronger.  This is quite unlike AM where even a very weak signal will result in a heterodyne (a tone) that may be annoying.  Signals that are of roughly equal strength, however, will destructively combine, resulting in garbage, and neither signal can be copied.

Why, then, would we purposely put two repeaters on the same frequency?

First, I'll point out that the coverage areas of the existing Farnsworth '62 repeater and Scott's hill do not overlap very much:  Farnsworth cannot cover east into the Wasatch range very well, and Scotts, being behind the front range of mountains, does not have a view of the Salt Lake and Utah valleys.  On this basis alone one could operate two repeaters on the same frequency - with the Scott's Hill repeater having antennas that direct eastward - and get away with it.  There are certain places where coverage overlaps (up the canyons adjacent to the Salt Lake valley, in the Cache valley, and a few other points, primarily to the north.)

In these overlap areas, then, how do we prevent destructive interference?

The destructive interference between two FM signals (often called doubling) is a result of the two signals randomly adding and canceling each other out as they are modulated.  There is also a heterodyne that results between the two carriers not being on exactly the same frequency at any single instant, causing further distortion and noise in an FM receiver.

So, if we can get rid of the effect of the signals randomly adding to each other and heterodyning against one-another, then we can avoid the problem.

The way to do this is to design the system so that the two transmitters are on exactly the same frequency and are modulated in exactly the same way.  If this is done, the receiver, when it sees both signals of equal strength, won't be able to tell them apart and will consider them to be the same signal (but coming from different directions...) and happily demodulate them.

Multipath distortion results from reflections of the same signal arriving at different times to the same antenna.  Sometimes they will add up and reinforce each other and at other times they will cancel each other out.  This is the effect that you see when you experience picket-fencing while driving (the signal chops in and out, the rate depending on your speed and direction) or when you find that one spot where you can't hear the repeater, but moving a few inches on either side allows you to hear it fine. (That spot is usually the one in the chair that you wanted to sit in to talk, by the way.)  If the second transmitter is just like the main transmitter, then isn't that going to behave just like a source of multipath?

Well, since the overlap areas are pretty small, there are only a few places where this is likely to happen. Also, there is a technique that can be used to reduce the problem in areas where overlap occurs..  This technique involves shifting one transmitter slightly off frequency (between 10 and 20 hertz) from the other so that a stationary receiver will never be stuck in a spot where the signal is canceled out completely.  The person listening in that spot may hear what sounds like a bit of picket-fencing even when you aren't moving, but you will still be able to easily copy the signal (even though it may be a bit choppy...)

This picket-fencing is not to be as severe as you might first think:   First of all, you would have to be in an area where the signals are almost exactly the same strength:  you would have to work really hard to find something like that, as it would have to be in a canyon somewhere...  Secondly, as you move along, you will find that you will likely hear either one transmitter, or the other:  In normal picket-fencing conditions, there is only one transmitter to hear at all, so you are more likely to hear the signal drop into the noise as you picket-fence. With two transmitters, you'll probably hear the other transmitter rather than just noise. This is only true in the overlap area, of course: Outside this overlap area things will be no different than any other repeater.

This technique is applied in the commercial world and has proven to work pretty well.  The fact that there is little overlap between the two transmitter sites (and what overlap does exist is in the mountains anyway...) makes for an almost ideal implementation of this technique.

As we explained, your receiver will hear either one or the other transmitter as you move around.  These two repeaters also share the same input frequency.  As you move around, one site will sometimes hear you better than the other.  It would make sense, then, to design the system so that the receiver that hears the signal best is the one that feeds both transmitters.  This is what the voting receiver system does.   If you start to get noisy into one site but you still have a good signal into the other, the system will automatically switch to the receiver with the best signal.

This has the effect of greatly reducing multipath in the weaker-signal areas for signals going into the repeater from those overlap areas.  Even in the non-overlap areas, it is possible that, out of some fluke of propagation, you just happen to be putting in enough signal to the other site that, instead of completely dropping out, you can put at least some signal into the other site.

Some interesting implications...

Running a voting/synchronous repeater system has some interesting implications that deserve to be mentioned:

• Improved coverage.  If you think about it, the coverage of a two synchronous, voting repeaters is better than the total of two individual repeaters on different frequencies that are linked.  Why is this?  For the synchronous, voting system, you will simply hear one or the other transmitters as you move into the overlap area and as you transmit, one or the other receiver is constantly monitoring your signal and choosing the best one.  If you had individual repeaters on a linked system, you would need to switch constantly between the audible repeaters to see which one was doing best at that instant - even if it only did better for a fraction of a second.  That fraction of a second may result in a lost word or sentence, requiring a repeat, and it increases the tedium of trying to make sense out of weak, choppy signal.  That doesn't even take into account the hassle factor involved in constantly switching between different repeaters. Don't forget the the voting receiver is working for you in receive as well.
• Spectrum savings.  Repeater channels are at a premium.  Using the synchronous/voting scheme you can expand the coverage of a repeater system without occupying yet another repeater channel.  The linking between the two repeaters can be done on a UHF or microwave band where spectrum isn't so tight. (That's what that spectrum is for!)
• Demonstration of a technology.  It is important that we amateur radio operators improve our methods and develop technology and techniques.  Assembling this system will not only advance our state-of-the-art, but lead the way for others to implement similar schemes to improve our ability to provide communications and do that which is required of us as amateurs.  (Hint:  Read the first part of FCC part 97 - the Basis and Purpose part in particular...)

If you are interested in helping with this project, please contact the Utah Amateur Radio Club (UARC) Scott's Hill project coordinator, Bruce Bergen, KI7OM, at (801) 943-1365 or send him email at bbergen@xmission.com.