When two identical antennas are mounted together (ganged) and pointed in the same direction and wired together properly, there is a theoretical possibility of a 3 dB improvement. That is, twice the signal power is delivered to the TV compared to what a single antenna would do. In practice, 2.5 dB is readily achieved, 0.5 dB being the typical loss in the combining device. But if the two antennas are pointed in different directions (towards different stations) a 3.5 dB penalty for each antenna is the likely result.
The above statements are true regardless of whether the antennas have shared or separate amplifiers. For a shared amplifier, if the antennas point in different directions, half the power each antenna takes in reflects off the combiner and is rebroadcast out the antennas. Why this doesn’t happen when they are pointed the same way is harder to explain. (For an explanation see Ganging Antennas.)
For dual amplifiers, when the antennas are pointed the same way, this signal is increased by 6 dB but the noise is increased by 3 dB, so the overall improvement is still 3 dB. (This comparison is relative to the same circuit with one amplifier powered off.) When they are pointed differently, the 3 dB noise increase causes a signal/noise ratio loss of 3 dB for both stations. Dual amplifiers will eliminate the 0.5 dB combiner loss, but only if the amplifiers are closely gain-matched.
(Ganging non-identical antennas is not recommended. They would need to produce equal voltages, and adjusting out the phase difference is not generally possible for all stations.)
Ganging a pair of Channel Master 4228 8-Bays will give you probably the best UHF antenna that a consumer can achieve with reasonable ease. The author discovered some extra problems with the 32-bay described later, a project that should be attempted only by people who love antennas.
If all your weak stations are above channel 40 then ganging a pair of Yagi/Corner-Reflectors would be smarter. That project will not be described here, except to say:
Your big decision is deciding between a side-by-side mount and a one-over-the-other mount.
The elevation view of the radiation pattern is the same as for a single 4228. But in the view from overhead, the 16-bay is 2.1 times more directional. This could be good or bad. There is no better antenna for eliminating ghosts that arrive from near the front. But even a Channel Master rotor will have a hard time hitting the correct direction. (Radio Shack rotors need not apply.) Some World War II radar antennas weren’t much different from this. Hopefully all your transmitting antennas are in the same direction, either because they are on the same tower or because the city is so far away. When a rotor is required, the one-over-the-other mount is usually wiser. Usually side-by-side 4228s are positioned so that the two screens just touch. But if they are 1-1/2 inches apart then the mast can pass in front of the screens, yielding a better weight distribution. A few tie-wraps forcing the screens to touch will reduce radiation to the rear and increase gain very slightly. Mounting the two antennas farther apart will make the main forward lobe even narrower, but the side lobes will grow in size.
In most situations, a one-over-the-other is the wiser choice for a 16-bay. The radiation pattern viewed from above is the same as for a single 4228. But in the elevation view, the 16-bay is 2.2 times more directional. This is enough to require taking the skyline elevation into account. The antenna should be tilted up to point at the skyline (the distant horizon).
Some authors will recommend that a motorized tilter be used since the angle of the incoming signal can change from day to day. It can. But high angle days are strong signal days, and the loss of a dB won’t matter. This author recommends a tilter only when a rotor must point the antenna in different directions with different skyline elevations.
The simplest mounting technique requires a single heavy angle-iron 65-70 inches long. Attaching it just below its midpoint to the top of the mast will keep the assembly from being too front-heavy.
At 8 lb., the 4228 is a heavy antenna. Putting up two of them requires a 1-˝ inch, 16-gauge metal mast. (A Radio Shack mast will bend with the breeze.) The total weight of the antennas, mast, mounting irons, etc. will exceed 25 lb. Trying to erect it yourself on a sloped roof is something akin to suicide, even without a wind. You need help. You need a large helping of good judgment. You need a rope around your waist so that you don’t fall off the roof when the whole thing tips over. Some antenna adjustments will likely be necessary, so don’t think you can put it up once and be done with it. Yagi/Corner-Reflectors weigh a lot less.
(The author has not tried a rotor mount and doesn’t know for certain if the Channel Master rotor can handle this weight or has the aiming accuracy required for a side-by-side mount. You need to discuss this with a knowledgeable sales person or installer, but such people are rare. Rotors that handle more weight can be found in HAM radio stores.)
The two antennas must be phase-matched. This means that the two signals must arrive at the combiner in phase. (a ±10° phase error is not a noticeable error, but anything larger should be avoided.) You do this by maintaining symmetry in the feed system. In other words, the wires for each antenna should be identical in type and length. The actual length is not critical.
If a ground reflection causes one antenna to be phased ahead of the other, this should be adjusted out by repositioning one antenna. This is most easily done by finding a new skyline tilt angle. Simply adjust the tilt while watching the signal strength. Different stations could require different angles, so adjust it for your weakest station.
There is a chance that you will mix up the polarities such that the two antennas subtract instead of add. Doing this will result in two forward lobes, reduced in size, with a null straight out the front. After the antenna is fully hooked up, you should rotate the antenna to check for this pattern. If so then you have to reverse the connections on one of the antennas. The antennas come with a balun that has a “China” stamp on one side. I believe this stamp is the key to getting addition on the first try.
There is no point in building this antenna if you plan to use a Radio Shack amplifier. Their best amplifier will cancel out most of the advantage of the second antenna.
Having two amplifiers eliminates the combiner loss, but requires you to find amps with equal gain. The only procedure doable by consumers that I can think of for adjusting the gains requires attenuators on the outputs of the amplifiers: one fixed and one variable. But the variable attenuator might not pass D.C. I am at a loss for suggestions, other than bringing both coax lines into the house or putting a 120VAC socket at the antenna.
The problem here is finding a low-loss combiner. Supposedly, combiners and splitters are different devices. While either will do the other’s job, combiners are supposed to be lower loss. (Splitters don’t have to be low-loss since they come after the amplifier.) But the best device I have found is a VHF/UHF splitter that is quite lossy above channel 50. I will continue to search. Watch this space.
The author’s neighborhood has hot spots and cold spots, places where the signal strength is strong or weak. This is a consequence of overlapping fields. Being 40 miles from San Francisco and behind some hills, DTV reception is only possible when the antenna is positioned in a hot spot. These hot spots are 10-16 feet apart for any channel, and are in different places for different channels. If you have hot spots for any UHF channel then you will have hot spots for all UHF channels (from the same direction). The distance between hot spots is determined by the frequency, the distance to the skyline, and the geometry of the ridgeline at the skyline. That ridgeline is producing the overlapping fields.
If one 4228 is in a stronger field, then part of its signal will be retransmitted out the weaker antenna. This loss may equal the little bit of gain you had hoped for from the weaker antenna. You will likely find that two 4228s are no better than one. This retransmission is only avoided when both 4228s are in equal fields. (When dual amplifiers are used, the argument sounds different but the result is the same.) At the author’s home, the change in field strength in just 3 feet is enough to wipe out most of the hoped for 3 dB gain when the antennas are mounted side-by-side. Fortunately hot spots are not generally spherical. Rather, they tend to extend upward and forward more than they extend laterally. So the one-over-the-other configuration is much more likely to work in a neighborhood with hot spots
But there is an exception to that. If the 16-bay is close to the ground and the ground is bare extending toward the station, an efficient ground reflection is likely. This is another case of overlapping fields. But in this case, the hot spots are mainly arrayed vertically. The hot spots are likely close together vertically, but farther apart laterally. In this case, the side-by-side is the configuration more likely to work. The incoming wave is angled downward by only a couple of degrees, and so the ground reflection occurs on ground that extends perhaps hundreds of feet toward the station. If this ground is paved, dirt, water, or a grass lawn, then the reflection is efficient and will produce extremely weak cold spots. If it is covered with weeds, shrubbery, trees, or somebody’s house then the reflection is scattered too randomly to have any effect on UHF reception.
An alternate explanation: (non-essential reading)
Two signals pass through each other without interacting. But an antenna will respond to the sum of the instantaneous signal voltages.
The fact that the author’s hot spots are mostly orderly suggests that his neighborhood has two overlapping fields. That is, the signal is coming from two spots on the skyline. How far apart are those spots? This is found from the simple trig formula above.
l = 16.3 inches for channel 56
D = 11 feet for channel 56
Thus q = 7.1°
Thus the signal is coming over the horizon at two spots 7 degrees apart. A 4228 in a hot spot can pick up both of these signals. But a 16-bay side-by-side is too directional to be aimed at both. It will likely pick up twice as much of one signal but none of the other, and thus will equal the performance of a single 4228.
This seems like a completely different explanation based on a different phenomenon, but in fact the two explanations are equivalent.
If you want to explore the locations of your hot spots, a Silver Sensor on a 10-foot pole is a good method. This antenna is small enough to fit in any hot spot, and probably strong enough for a digital-lock in a hot spot for your strongest station. If possible, work at about the elevation where you plan your permanent antenna. You will need a monitor positioned there so you can see the signal strength from the receiver. The distance between the hot spots will be roughly the same for all channels.
If your hot spots are too small both vertically and laterally, then a 16-bay might be out of the question. Your option then is to put each 4228 in its own hot spot. But this only works for one channel.
You might curse your bad luck if you find you have hot and cold spots. But you would be looking at it wrong. In fact, your neighborhood is concentrating the signal for you. An antenna in a hot spot can be up to 6 dB smaller than it would need to be in a “flat” neighborhood. Now, if only the hot spots never moved. But, that is another story…
Let’s say you are 20-miles from the station but behind a big hill. Your 4228 mostly works, but you see some dropouts randomly. A 3 dB improvement will likely solve your problems.
But beyond 40 miles, weather affects UHF considerably. A 3 dB improvement will make some of the dropouts go away, but weather can always get worse. Every antenna improvement will help. But beyond 60 miles, it might not be possible to eliminate 100% of dropouts no matter how good your antenna is. Beyond 60 miles, solid reception on 9 out of 10 days is generally a good result.
This page is part of “An HDTV Primer”, which starts at www.hdtvprimer.com