The idea of colocation is nothing new. We have been sharing sites among FM, TV and even AM stations for decades. The earliest FM and TV antennas were installed on AM towers, mostly because they were co-owned with the sister AM (which was the dominant medium in those days) and because the AM towers were already in place.
(click thumbnail)
CAD was used to position the new array within the old ‘on paper.’ During my career, I have had numerous occasions to colocate stations in just about every combination. But a few years ago, a different kind of opportunity presented itself as a rather elegant solution to a longstanding problem with one of our Denver AM stations. Perhaps that experience may provide inspiration for others facing siting issues.
The issue was that one of our stations was a daytimer on 800 kHz, a Mexican clear channel and home of 150 kW XEROK in Ciudad Juarez, just across the border from El Paso.
While night operation was allowed on the channel, the question was, “Why bother?” With such high interference levels from that “border blaster” at night, it was hardly worthwhile to go to all the trouble of building out and operating a multi-tower array to protect the entire Mexican border from Brownsville to Tijuana.
And so it was that as an AM major change window approached, I began looking for alternatives for the station. My search didn’t have far to go.
I found that the next channel up, 810 kHz, was a good bet for our little daytimer. The daytime allocation would support 2.2 kW from the existing site and three-tower directional array, and there was room to do something meaningful at night too — if I could find the right site.
I had four basic criteria on the new frequency: Protect KGO in San Francisco; protect WGY in Schenectady; provide interference-free community of license coverage to Brighton, Colo.; and provide interference-free night coverage to as much of the Denver metro area as possible. Piece of cake, right?
An array within an array. In reality, as AM nighttime allocations go on U.S. clear channels, that really wasn’t too tall an order. We’d need a broad null to the west for KGO, a broad null to the east for WGY and lobes both north and south for the two interference-free coverage areas. I could do that with a four-tower parallelogram or trapezoid. So the challenge became where to put this four-tower array.
In the Denver market, AM stations are generally all located on “DA Row” along the South Platte River. There are many reasons for this: good conductivity, flood plain land not otherwise useful for development and good distance from airports and flyways. So naturally I began my search in this same area … and ended it at my own doorstep, so to speak, at a 50-acre site we already owned: 560 kHz, KLZ.
Widely spaced array
It occurred to me that the two-tower 560 kHz KLZ array was a bit of an odd duck. It was a wide-spaced (197 degrees) broadside array with nearly 1,000 feet of available dirt between the two towers. How often does that happen?
I started with the original 1962 as-built survey of the KLZ site showing the property boundaries, the improvements, transmitter building, towers and guy paths, bringing that into a CAD environment. Then I did a rough directional array design to get the tower layout I would need. The tower layout was brought into the CAD drawing and moved around as a unit until I found an optimum location. All of the four new towers and their guy wires needed to be a reasonable distance from the existing KLZ towers and their guy wires while retaining the proper spacing and orientation.
The results were a pleasant surprise. There would be more than enough room for the two directional arrays to share the same site without sharing any towers (important because the height of the KLZ towers would make management of the vertical radiation difficult). The new four-tower array would fit completely within the existing KLZ array — an array within an array. Concentric colocation!
At that point, I was off to the races. I completed the directional antenna designs for both day and night, completed and filed the FCC application and then started on zoning and use permit at the county.
The new night towers would only be 200 feet high, so no FAA approval, marking or lighting would be required. That helped us at the county because galvanized, unpainted, unlit towers really do present a small visual impact, always important where neighbors are involved. This did result in electrically short towers (58 electrical degrees), but the array still more than met minimum efficiency and was no problem with the FCC.
(click thumbnail)
ATU and filter cabinets at KLZ Tower No. 2. One other thing came into play at this point. In a totally unrelated action, the county approached me about the northwest corner of our property where it wanted to construct a viaduct for a bike/jogging trail. They were seeking a one-acre easement from us for this purpose because they didn’t otherwise have the land to accommodate the approaches to the viaduct. It occurred to me that a quid pro quo might be in the offing if I played my cards right.
As we filed the use permit application for the new towers, I noted the easement request in the narrative and stated that we would be willing to simply deed over that one-acre corner of our parcel. The staff jumped on that and in no time we had a deal!
The remainder of the project went like clockwork. The FCC granted the application very quickly. Excavations were made for the tower base piers and guy anchors, steel was stacked, trenches were opened for the transmission and sample lines, and a complete new ground system was installed for both stations. KLZ stayed on the air during all this work, but sometimes at reduced power.
Next came a new Kintronic phasing and coupling system for both the new array and the old. This included pass/reject filters for all six towers and detuning components for the two KLZ towers. The new 810 towers were sufficiently short that they would be no factor at 560 kHz if we would simply float them on that frequency.
However, the 450-foot KLZ towers are 131 degrees tall on 810 kHz and required careful detuning. We did this by exciting one of the 810 towers with 100 watts. We placed a field intensity meter on a tripod situated roughly halfway between the driven tower and each of the 560 towers in turn, with the meter oriented toward the 560 tower and perpendicular to the driven tower. We were able to walk the re-radiation down to a point where I could hear on the FIM a co-channel station in South Dakota beating with the remnants of the 100 watts from the driven tower a few hundred feet away.
Arc Measurements to Determine Radiation
Finding out the inverse distance field (IDF) of a potential re-radiator may seem like a daunting task but it is accomplished easily using a technique called arc measurements.
Using a topo map, start by drawing a line between the station’s tower and the re-radiating object. Draw an arc centered on the midpoint of that line with a radius equal to half the distance between tower and re-radiator. Mark likely accessible points around that arc starting as close as possible to the re-radiator and extending around to about three-fourths of the distance to the station’s tower. Enter these as waypoints in your DGPS/WAAS-equipped GPS.
Now take the FIM out into the field and locate the closest point to the re-radiator with the GPS as possible. With the FIM on a tripod, orient it so that the electrostatic break slot in the bottom (the top when opened) of the lid/antenna is oriented toward the station’s tower (use it like a gunsight). Do not rotate the FIM for maximum signal! Read the field strength and log it along with the distance. Repeat at as many of the other points around the arc as you can access.
Back at the shop, plot distance vs. field strength on a piece of log-log paper and graphically analyze the plot to determine the inverse distance field. It’s as easy as that! Tune-up went very quickly and we had the new pattern nailed and proofed within days of first RF excitation. We had some work to do on KLZ because the new 560-pass/810-reject filters created some bandwidth issues on 560. We had employed slope correction in some of the networks, but that ended up working against us, producing a “knot” in the impedance plot instead of the smooth horseshoe we were seeking.
Still, this was relatively easy to deal with and in short order we had program test authority and were operating both stations from the concentric arrays. No second- or third-order IM products were observed from this operation, a testament to the effectiveness of the filters and the broadband design of the Nautel transmitter power amplifiers.
Night coverage was everything we hoped for. The community of license receives a nice, fat lobe as does most all of the Denver metro area (with the exception of the far western suburbs — that KGO protection again).
Interestingly, the three-tower inline day array turned out to be a lot harder to tune up on the new frequency than the four-tower trapezoid night array; we had some very localized re-radiation that gave us a lot of grief at the tune points. But we got it done and soon had a much improved day signal for the station as well.
In the end, we spent very little money to turn a sleepy 1 kW daytimer into a full-market 24-hour facility without buying any dirt or developing a site de novo. We now have a very efficient operation that makes better use of facilities that we already owned, keeping the neighbors and politicians happy in the process.
Without a doubt, this was a unique situation. I can’t imagine there are that many two-tower, wide-spaced arrays out there on very low frequencies that might lend themselves to concentric colocation. But if the time comes when you have to find a new home for that AM, don’t discount existing sites because the tower line is all wrong. Perhaps partial tower sharing is an option. As sites get scarcer we’re all going to have to get more creative.
Cris Alexander is director of engineering at Crawford Broadcasting and a past recipient of SBE’s Broadcast Engineer of the Year award.