NCJ November/December 2005 31
With the advent of using wire verticalsaround our towers to make a phased ar-ray system, it has become apparent thatthe tower can affect the array’s perfor-mance, usually for the worse. Many timesI am asked about detuning a tower andhow to do it. This short article is the resultof trying one commonly cited way that didnot work, and doing research on how thebroadcast industry does it, and how wefinally, successfully, detuned a tower witha 160-meter 4-Square array.
The tower in Figure 1 belongs to W5IZ,and it is a 200 foot tower with an 80-meterbeam on top and four cantileveredPhillystran guys that hold up a 160-meter4-Square system. Each vertical is onequarter-wavelength tall and is fed atground level with buried radials. The 4-Square system is an Array Solutions op-timized box that uses what has becomeknown as the Lahlum method of tuning.(The namesake is Robye Lahlum,W1MK—see his two articles “Phase Ad-justment Technique for a 4-ElementSquare Phased Vertical Array” and “PhaseCorrection for a Quadrature Hybrid-FedAntenna Array” in the May/June 2005 is-sue of NCJ). The object of this method isto increase the front-to-back ratio and gainof the array by using optimized phasesettings of the elements.
We installed the system and noticed theF/B of the array was down from the pre-dicted pattern, which should have beenon the order of 30+ dB. We were only see-ing about 12 dB of front to back using asignal source that was carefully placed togive us a good reference to adjust the ar-ray. By measuring the currents in the towerlegs we could see there was significantamounts of current as viewed on an os-cilloscope (that had a current probearound a leg of the tower) when we trans-mitted on the array. This was not predictedin the NEC model of the tower we made.What to do? We needed to detune thetower.
Method 1Doing some Internet research, we found
that one solution was to create a looparound the tower at some midpoint. Then,we could drop a wire down below it at 40-60 feet and put a capacitor in series totune the circuit to resonance. We tried thisat various heights and we spent a lot oftime without success. The tower was justnot going to be detuned using this methodno matter if we resonated the “loop” byincreasing or decreasing the currents init. We still had significant currents in thetower and this disturbed the pattern.
How to Detune a Tower By Jay Terleski, WXØB
Figure 1—The 200 foot tower at W5IZ,with an 80-meter Yagi on top and a 4-Square system around it.
Figure 2—A typical detuning skirt.
Method 2I knew the broadcast industry often
grappled with this problem, so I starteddoing more research about AM broadcastdetuning systems. This is a serious issuein the broadcast industry. An AM broad-cast station has to maintain a patternbased on FCC regulations. If the patternis changed by a newly erected cell tower,for example, the cell tower owner is re-quired to detune his tower so as not toaffect the pattern of the AM broadcaster.There are several companies that make“tower skirts” to accomplish this.
As you drive around you may notice thatsome cell towers have these skirts aroundthem. This is a sure sign that they are lo-cated near AM broadcast stations—andthat their presence has disrupted the sta-tions’ patterns.
I called my friend Goose Steingass,W8AV, who is a broadcast engineer, andhe agreed the best way to detune a towerwas to skirt it. With the W5IZ tower herecommended to skirt the whole thing tomake it “disappear,” versus just detuningone quarter-wavelength of it.
Figure 2 shows a schematic diagramof a typical detuning skirt. It usually has 3or more detuning wires running the full-length of a tower. The top wires are elec-trically attached to the tower. The lowerend of the skirt is insulated from the tower,tied together in a halo around the tower
Figure 3—W5IZ’s tower with thedetuning skirt (you have to look closelyto see the wires around the tower).
32 November/December 2005 NCJ
and attached to a detuning LC networkto adjust the skirt to be nonresonant atthe desired frequency. This is a muchmore rigorous application to build versusthe Method 1 approach, but that’s whatwe had to do. We built a 3-wire skirt usingwater pipe standoffs at the top, midwaydown (which were insulated), and at thebottom. See Figures 3 and 4.
At the bottom of the tower the threewires were tied together and held downfrom the top with turnbuckles and an in-sulator to electrically isolate them from thewater pipe standoffs. I built an LC network(see Figure 5) just out of trial and errorwith #12 AWG magnet wire. We broughtdown one wire from the skirt and tied it toone end of the coil. I scrapped insulationon top of the magnet wire to allow me toadjust a jumper from the coil. I also had ahandful of RF capacitors.
For measurement we used a 500 MHzoscilloscope and placed a current probe
Figure 4—The insulated bottom of the skirt. Note the halowires that tie the skirt wires together.
Figure 5—The LC network used to used to detune the tower.It’s made of a coil of enamel wire and an RF capacitor.
in the wire that was feeding the coil fromthe skirt. The trick is to minimize the cur-rent in the skirt, or the voltage read onthe scope.
The current probe can be made simplywith just some hook-up wire. Place fiveor more turns around your network wire,or just enough turns to easily produce 1-2 volts on your scope when transmittinginto the array with 10 W or more. The‘scope I used easily picked up the trans-mit signal. As a reference, we adjustedthe scope down to the millivolt scale andwe could see AM broadcast signals mov-ing the trace. Al, W5IZ, lives way out inthe country and has no nearby AM broad-casters, so these signals were in therange of 2-3 millivolts. We knew we prob-ably could not make the 160-meter sig-nal that weak, but we could try.
Adjusting the coil and capacitance val-ues I found a sweet spot for a tap point onthe coil that minimized the 160-meter sig-
nal being transmitted. I was further ableto reduce the signal by changing differentvalues of capacitors. I ended up using a100pF cap soldered to the coil to form a T-network. Al and I were amazed that yes,indeed, the signal on the skirt was downat the same level as the AM broadcast sta-tions. Now to see if it worked!
We set up the measurement systemand, lo and behold, the F/B had jumpedup 18 dB or more. I readjusted the4-Square system and we saw 30 + dB ofF/B.
The moral of the story is towers cansignificantly adversely affect the verticalphased arrays we hams build aroundthem. The good news is we can fix theproblem by carefully building industry-standard skirts. But you get what you putinto it, and trying to cut corners just maynot pay off in the long run.
