Over-the-air reception with an antenna can be a maddening affair. TV signal strength can vary dramatically with small position changes, especially indoors. Depending on your location and willingness to get up on the roof, it may take a great deal of tinkering to get an antenna system set up right.

My own TV antenna system is the product of a great deal of fiddling. The motivation is the start of baseball season in March. Oakland A's games are broadcast on channel 36 in the Bay Area, but channel 36 is not broadcast from the area's main TV tower. To bring in the additional channel, I needed to find out where to point the antenna, join it in to my existing antenna, and ensure that a good enough signal was available throughout the whole system. It required extensive experimentation, and it took quite a long time because of a demanding travel schedule.


To call my previous antenna a "system" is to give far too much credit to the single antenna I put in place as part of last year's TiVo project. I pointed it at Sutro Tower, the main TV broadcasting antenna in San Francisco. The single antenna served me well as long as it was connected to one receiver, but its shortcomings became apparent when baseball season started and I wanted to receive additional channels.

The first step in setting up a system is to find out how many antennas are necessary and where they need to point. In the United States, the best planning resource is antennaweb.org, a project of the Consumer Electronics Association. From a street address, AntennaWeb will use terrain maps and geographical information to show the direction to television transmitters, along with the distance. The calculations are based on relative altitudes and the terrain, and they will account for the differences in propagation between analog and digital stations, as well as the station's licensed transmission power.

After looking up the location, AntennaWeb will print out a table that has each station's call sign, channel, compass orientation, and distance. You can also get a street-level map that shows the direction to the transmitters in relation to local streets. Figure 1 is similar to the street-level map shown by AntennaWeb for the location of San Francisco City Hall.

figure 1

Figure 1: Station map for San Francisco City Hall

For further information, you can use the use the FCC's TV query form to learn more about the transmitters. TV stations are licensed broadcasters, and must be granted the right to use a frequency. As part of that process, a great deal of information about the transmitter is made part of the public record, including the location of the transmitter, its power, and the direction its energy goes in.

Right away, I could tell I was in for a challenge. Table 1 shows the effective radiated power for the major stations in the Bay Area. The last two lines are for stations located on either Monument Peak or Mt. Allison, both of which are approximately 35 miles from my home.

Table 1: Effective radiated power for local stations
Channel Analog power (kW) Digital power (kW)
KTVU (2) 100 1,000
KRON (4) 100 1,000
KPIX (5) 100 1,000
KGO (7) 316 561
KQED (9) 316 777
KNTV (11)* 316 103.1
KBWB (20) 3,470 383
KCSM (43) n/a 536
KBHK (44) 5,000 400
KICU (36) 4,070 251**
KTEH (54) 661 290

*KNTV recently moved to a new transmitter; the values shown are from the construction permit for the new transmitter.
**The construction permit issued by the FCC was for 1,000 kW.

In addition to the close-in stations on Sutro Tower, I wanted to build an antenna that supported the digital transmissions of both channel 36 and channel 54. Luckily, both stations have digital transmissions that are close to each other: KTEH-DT uses channel 50, and KICU-DT uses channel 52. Both transmitters are close to each other, so a single long-range antenna should pick up both stations. The plan is quite simple: keep an antenna pointed at Sutro Tower, and join it to an antenna that points at Monument Peak.

The map printed out by AntennaWeb is most useful for pointing an outdoor antenna. With no obstructions, the directions indicated by the street map are relatively accurate. Indoor digital reception can be somewhat more difficult, and it often makes sense to point at a strong reflection. Directly at the source is not always the strongest signal indoors. I chose to keep my second antenna indoors because it is easier to tweak slightly, and I did not want to secure something to the roof. The penalty is that the house causes at least a 1 dB loss of signal, and probably more.

Antenna System "Glue"

With a plan in hand, I also need to get some spare parts for the whole system to tie it all together. Depending on your situation, some or all of these may be important.

Cable. Without cable, you don't move signal around. There are several grades of cable. RG-59 is the standard quality, but has quite high loss, especially over long runs. Putting a large high-gain antenna at the end of a long cable run and then pushing that signal through RG-59 is pointless. Use RG-6 or better. The cable is thicker and slightly harder to work with, but the performance is much better. It may also help to buy high-quality cable. In the course of one weekend's experiments, I picked up some RG-6 cable at a local RadioShack. That turned out to be a regrettable decision because the connectors are hard to tighten or unscrew, which makes modification of the system more time-consuming and frustrating than it needs to be, especially if you are reaching around to the back of any equipment.

Preamplifier. Antennas at the fringes of reception areas are going to be picking up weak signals. In such cases, one of the best investments you can make is in a high-quality preamplifier. Rather than send a weak signal from the antenna down a long cable and try to amplify a very weak signal at the other end, a preamplifier makes the weak signal stronger at the antenna, and sends the strengthened signal down the cable. One of the reasons this is effective is because very weak signals are lost in the noise, so amplifying the weak signal and the noise just results in a very noisy signal. With a preamplifier, the signal is strong enough at the antenna that it can be amplified again relatively cleanly.

High-quality preamplifiers mount on the antenna mast. Power is supplied over the antenna cable from an AC transformer at the other end. Putting a preamplifier in place will require a number of short cables to "splice in" both the preamplifier and its power supply.

One of the key specifications in a preamplifier is its noise figure. Preamplifiers that introduce noise are not effective because they do not improve the signal relative the background noise. Low-noise mast-mounted preamplifiers are a common antenna accessory, and the best models are made by the same companies.

  • The Channel Master 7775 is a UHF-only preamplifier with a 2.0 dB noise figure.
  • The Channel Master 7777 is a VHF/UHF combination that has the same 2.0 dB noise figure.
  • You can also use Winegard preamplifiers, such as the AP-4800, but they have a 2.9 dB noise figure and roughly the same gain.

Of the widely available preamplifiers, I chose the Channel Master 7775 because I only need amplification for UHF, and it has the lowest noise figure.

Joiners. Joining two antennas is easy, but doing it correctly is hard. A two-way splitter can also be used as a two-input combiner. However, the energy on each of the inputs will be split between the other outputs. Channel Master makes a passive device called a JoinTenna that has two inputs: one for a specific channel, and one for the rest of the channels. To add a single channel from a second antenna, hook it up to the specific channel input, while using the main antenna input for everything else. I am fortunate that the JoinTenna works in a way that is ideal for my situation. Most of the stations come from Sutro Tower, and I only need to add a single channel from farther away. JoinTennas can be hard to find, so you may have to be persistent. (I could not find them online, but eventually found Schad Electronics in San Jose, which carries a full line of Channel Master equipment.)

Choosing Antennas

With knowledge of the distance and direction of each signal, you can pick out the appropriate antenna. Most digital TV stations are broadcasting on UHF, which simplifies the antenna design. UHF-only antennas are smaller, and are often the "panel" type that can be much more compact. Although a panel is taller, it requires less space than a long antenna boom. It is also easier to rotate a panel than a boom in limited space.

All of the stations I want to receive are in the UHF bands, with the exception of KNTV, the Bay Area NBC affiliate. I am, however, quite lucky in that I have a nearly unobstructed line of sight from my front door to the transmitters at Mt. San Bruno. The signal is so strong that no antenna is necessary. A 75-ohm shielded coaxial cable plugged into my HD receiver can get a perfect signal from the tower. Even using a UHF antenna, I am able to receive the signal easily. (VHF HDTV broadcasting is relatively uncommon, but broadcasters value it because the much lower frequency require less electrical power to run the transmitters.)

I started by considering the four main options for long-range UHF reception:

  • The Channel Master 4228, which is rated for 60-mile reception by Channel Master
  • The Channel Master 4221, which is a half-size panel rated for 45-mile reception (the 4228 is two 4221s coupled together)
  • The Winegard PR8800, which competes with the 4228
  • The Winegard PR4400, a half-size version of the PR8800

With such a long-range reception, I wanted to get the biggest antenna possible. Ken Nist's HDTV Primer has a chart of gain plots calculated from mathematical models of the antennas. The Winegard 8800 and Channel Master 4228 are similar antennas, but the Winegard is better at the low end of the UHF spectrum, and the Channel Master's strength is at the high end of the spectrum. My target channel is 52, so I decided to try the Channel Master first.

Installing and Experimenting With Antennas

It was my initial intent to install the new antenna for channel 36 indoors. Indoor installations are safer because there is no risk of severe electric shock from overhead power lines, and much less risk of severe falls. Antenna grounding is also not necessary in indoor installations. Mounting an antenna indoors is not without challenges. Severe multipath interference can create "hot spots" for signals throughout the building, and the hot spots for the signal may not always match up with aesthetically pleasing mounting locations.

I started off by using the 4228 on the floor of the living room, wired directly to my HD receiver. Surprisingly, it worked quite well and was able to deliver a stable picture without any special effort. However, the cable linking the antenna and the receiver was only 15 feet long. When I tried joining the 4228 to the existing set-top antenna, the signal was just shy of being able to lock on. Adding a preamplifier to the antenna compensated for the loss through the joiner, but when I moved the assembly to the attic, the antenna was just too tall to be able to rotate it towards the signal.

Through experimentation, I discovered that a Silver Sensor pointing in exactly the right direction would bring in the signal, so I took it up to the attic and connected it to the preamplifier. I could get a signal lock, but it was just shy of delivering a quality picture. After consulting the HDTV Primer antenna graph, I decided to try the AntennasDirect DB-2, which has several dB higher gain on channel 52 than the Silver Sensor. Even with the preamplifier, the receiver was unable to lock on.

At this point, I had exhausted all of the common antennas, but I was tantalizingly close. All I needed was a few more dB of antenna gain. Most mainstream antennas constructed with a central boom would likely offer the gain, but would be too large to fit in my confined space. Many all-band antennas, or even UHF boom antennas, are several feet long.

In desperation, I turned to a "last resort" antenna. Most mainstream antennas are designed to work over a relatively wide range of frequencies, but a few specialized antennas are designed for a single channel or a relatively tight channel range. Most of those are designed for use with the lower VHF frequencies.

The only narrow-band UHF antenna I know of is the Blonder-Tongue BTY-10-U, shown in Figure 2. There are six models, each of which is specially designed for a sub-band of the UHF range and has slightly different dimensions. Even though I was turning to the BTY-10-U out of desperation, it had an additional advantage in that the narrow band design allows it to be smaller. Even the largest of them is over four feet long, with a foot-long element at the widest point. The Blonder-Tongue antenna generally sells for $150 new, but I was lucky enough to find one on eBay.

figure 2

Figure 2: Blonder-Tongue BTY-10-U

When the BTY-10-U arrived, I tried using it to lock on to the signal from channel 52. To start, I tried locking in channel 52 with a preamp from the living room floor. Once I was able to lock on to a signal, I moved it to the attic. Rather than set up a mast mount straightaway, I used empty cardboard boxes to prop it up. The signal received was somewhat sensitive to antenna position. To get a slight upward angle, I propped up the front of the boom. Interestingly, the antenna is not aimed at the station. The strongest signal must be an internal reflection from the material inside the roof.

figure 3

Figure 3: BTY-10-U installed

Indoor Antenna Adjustment

With the long-range antenna set up, I decided to revisit the indoor antenna that I use for most of the other channels. Due to a strange polarization effect, the Silver Sensor antenna that I was using as the main antenna was set up at a crazy angle. The angle was an attempt to balance the vertical position with the best reception for channel 2-1 and the horizontal position with the best reception for all the other channels.

Now that I was no longer using the DB-2 as my long-range antenna, I decided to try it as the main antenna. It has significantly higher gain than the Silver Sensor, and can be mounted in a location where it picks up all channels well. With the Silver Sensor, I had a weekly ritual of fiddling with it to make sure that I had it in the right position. The DB-2 was easier to set up, and is a bit more tolerant of changes in position. Figure 4 shows the position of the DB-2, as well as its major drawback. It does not come with a stand, so you need to make one. My "stand" is composed of two stacks of books to get an inclination angle towards the transmitter. (I live in a valley below Sutro Tower.) To hold the correct incline, I use the antenna's mast clamp as a support, with a bit of foam to keep it from slipping across the surface of the book.

Although the antenna is somewhat directional and pointed away from the new transmitter for channel 11 on Mt. San Bruno, it still comes in loud and clear. I have direct line of sight to the new transmitter. The signal is so strong that my HD receiver can lock on from just a 6-foot length of coaxial cable without an antenna. I was quite pleased with this because I was not looking forward to joining a third antenna into the setup to receive channel 11.

figure 4

Figure 4: AntennasDirect DB-2 and "stand"

Wiring It All Together

I had initially hoped when using the antenna joiner that I could receive both channels 50 (54-1) and 52 (36-1). Channel Master JoinTennas are channel-specific, so I tried using JoinTennas for channels 49, 50, and 52. Although the JoinTenna uses a band-pass filter to let the "alternate" channel through, the filter is not perfect and will allow some of the signal from adjacent stations through. In the course of experimentation, I found that the signal from Monument Peak was so weak that I needed to use a JoinTenna for channel 52. As much as I tried to receive channel 54-1, I could not get a strong enough signal to do so.

In the testing phase, I was using a single HD receiver, the Hughes HTL-HD. When I went to connect the new antenna system to the MythTV system, I needed to split the signal three ways. Two outputs fed the MythTV tuners, and one the existing HTL-HD. My first attempt at connecting everything was to buy a single 4-way splitter. Splitters work by dividing the signal. A perfect splitter will cut the signal in half, and send half each way. In practice, there is a slight amount of additional loss, and most two-way splitters are rated for a 3.5 dB loss. Passive 4-way splitters impose a 7.5 dB loss. When I connected the two antennas to a 4-way splitter, none of the receivers could lock on to any channel.

To split the signal while ensuring that each component could be used for recording, I replaced the passive splitter with a Channel Master 3044 distribution amplifier. Unlike a preamplifier, distribution amplifiers can introduce a bit more noise because they work with signals that are already strong. Once I replaced the splitter with a distribution amplifier, I could recover high-definition signals on all three outputs.

Even with the distribution amplifier, there is a noticeable difference in quality between the Hughes receiver and the pcHDTV tuner cards. The Hughes receiver is able to lock up on weaker signals, and shows a visual glitch about once every month, while the pcHDTV cards will have glitches every week or two. The Hughes receiver has a much better capability to deal with the weak signal on channel 36-1.

I connected the Hughes receiver only to the TiVo because it was much more reliable when was hooked into the TiVo after hitting the distribution amplifier, and the signal is too weak to be reliably recorded by the pcHDTV boards in the MythTV system. Figure 5 shows the end wiring diagram, with both antennas, the antenna joiner, and related amplifiers. It also notes the other change I recently had to make because a TiVo software update disabled serial control and forced me to use an IR emitter instead.

figure 5

Figure 5: How it all fits together

Final Thoughts

Although there is still some fiddling left to do, the antenna system is finally complete. There are occasional picture glitches on channel 36, but they are small and short-lived. Given the distance I live from the tower, an occasional hiccup is worth the perfect digital picture the rest of the time. Although I could almost certainly get a better picture by doing a roof mount, I am perfectly happy to forego the climbing on the roof and put up with the occasional glitch.

Return to digitalmedia.oreilly.com