Engineering is largely about tradeoffs and constraints. Sometimes, the issues are clouded and the choices are fuzzy, but occasionally, the choice has a fairly clear split.
I had one of the latter situations with my TV antenna. I'm one of those folks who rely solely on free, over-the-air (OTA) broadcasts. (The reason for this OTA life is a subject for another discussion.)
When the US switched over to digital TV a few years back, I kept using the UHF bow-tie antenna in the attic:
Its balanced output goes through a 300Ω/75Ω impedance transformer (a standard balun you can get for about a $1) and feeds about 100 feet (30 meters) of coax to the TV. Even though the antenna has the wrong electrical size for the frequencies assigned to DTV, it has been working pretty well, so I left it as is.
But one channel's signal is marginal, and I get break-up when the leaves are in full bloom or there's heavy rain -- I suspect it's due to RF absorption. The problem is that while analog TV degrades somewhat gracefully with decreasing signal, it's quite different with digital TV: You get "all or (almost) nothing at all," to paraphrase the old song. The decoded image "falls off a cliff" and freezes, breaks up, or goes blank when the received signal strength falls below some threshold.
So, I have two choices: either replace the bow-tie antenna with an antenna with a preamp, or put up a bigger, better passive antenna. Since this is an in-the-attic installation with plenty of room, I am not constrained by appearance or space, so I can take that out of my mental equation.
What are the parameters to consider? There are four: received signal strength at the antenna, SNR at the antenna, signal strength at the TV, and SNR at the TV. They differ at the ends due to loss in the coax run, plus added cable noise.
Choice No. 1: Adding an antenna with a built-in preamplifier should help. It won't improve received SNR at the antenna itself, but it will help deliver a signal with much better SNR to the TV. Cost is around $30-50, and from what I can tell, they are omnidirectional or at least bidirectional. That's good, because the signals I want to get are from two major cities, and I am midway between them.
There's also a small AC/DC "wall wart" that would plug into the attic AC outlet to power the preamp. Here I am a little concerned, as those basic, inexpensive adapters don't have the high MTTF (mean time to failure), and I suspect I would only get a year or so of use in a hot attic. Admittedly, replacing one is not a big deal; it’s just another thing to do.
For Choice No. 2, I could build a bigger antenna with some additional gain. There are plenty of plans available online for these antennas, and they all seem to use the same basic design (see here for a typical one). The SNR of the captured signal should increase, but the losses in the coax might still be a problem. Some designs add a screen or solid reflector spaced behind to add additional directional gain, which may actually be a bad thing here as I am looking for both front and backside gain. Cost is next-to-nothing, and reliability is not an issue at all. As for a power supply -- none needed, of course.
So, that's the choice I face, and the pros and cons are fairly clear. The preamp version adds gain at the head end, and should reduce the effects of the coax run. The passive version is much lower cost, and extremely reliable, but may not improve the far-end signal strength and SNR numbers enough.
Since I don't have the instrumentation or inclination to make measurements, any analysis or modeling I do would be a wild guess. When faced with such a dilemma, I'll choose the simpler option first and try that passive antenna. Sometimes, despite our affinity for "electronics," a passive solution can be the best, or at least worth a first try. If that doesn't work, well, hey, it will have been fun at no cost, and with no risk. (See: No pain, and some gain: how "more metal" saved my RF link.)
Have you ever had a similar A versus B choice? How did you decide what to do?