The Call of the Open Sidewalk

Here are the other articles in this series:

• Background
• Going Digital
• Forgotten Dreams
• More Than The Sum
• Gathering Information
• Design

Here is what the antenna looks like:

Doesn't that look great? Pure function...

The support structure is a 10 foot chain link fence top rail bought at the local home improvement store. It is attached to the roof with a tripod sold as an antenna support. The top antenna points at channel 13. The middle antenna points at channel 7. The bottom antenna (low gain UHF band) receives channels 27, 35, 40 and 51 and points between the transmitters. The cables are held to the mast with multiple wraps of black vinyl electrical tape, which seems to be the local custom.

Here is the place where the signals are combined:

Here is the place where I ground the antenna:

The requirements for grounding things like TV antennas are different from place to place. If you can't manage to ground it exactly as required, don't just give up. Ground it as best you can. You should ground the outside conductor of the cable where it comes into the building as well. You can buy a thing called a grounding block to make this convenient.

As the digital TV conversion is finally complete in my area I can provide results:

Channel	Strength	Quality
7	90%	100%
13	87%	100%
27	86%	100%
35	80%	94%
40	94%	80%-100%
51	63%	75%

Strength is proportional to the level of the signal. Quality is a measure of how certain the demodulator can be that a particular bit is a 0 or a 1 (higher is better). Channels 7 and 13, each with a dedicated high gain antenna, are as would be expected quite good. The rest of the channels are fairly good with the exception of channel 51. I have no real explanation for this. The transmitter is less than 3 km away. The quality of channel 40 is fluctuating. That is probably due to either noise or a reflection that is changing faster than the receiver can compensate.

posted at: 17:18 | path: /tv | permanent link to this entry | Comments (2)

Creating a usable over the air (OTA) TV antenna system is one of those things that usually involves a significant amount of experimentation ... but you have to start somewhere.

A good place to start is with the worst signal you want to receive. Here is the important bit from my TV Fool report (as discussed in a previous post):

The farthest transmitter in our list of desired channels (7 13 27 35 40 51) is channel 13 at 41km (25.2 miles). Applying a realism filter to the claims of the people that make TV antennas, we come up with a requirement for some sort of medium gain antenna. How about we specify an antenna that does both our bands of interest and just point it at the channel 13 transmitter? This, by the way, is all we have to do in most situations. Deal with the worst and hope for the rest.

We would then have the close 27 40 and 51 directly behind the transmitter. Good. Most antennas work reasonably well to the rear. We unfortunately have 7 and 35 that are not so close and they are pretty much at right angles to the antenna. Most antennas are quite good at rejecting signals to the side. The solution here is to rotate the antenna so it points between 13 and 7/35. Lower gain antennas are better at receiving stations over a wider angle so we might be better off downgrading to a low gain antenna ... an unhappy but common compromise.

If I was in a more rural environment I might be done. I am instead in what might be termed a high density suburban area. My neighbours all have cable and do not have to bother with interference caused by their electronics. I am not blameless with respect to interference either. Low gain antennas are also good at receiving noise over a wider angle. I used to use a single low gain antenna in the analog days and had to live with all the received noise.

We go on ...

Before we unleash the dogs of complexity we might want to think about our available assets. I just happen to have a couple of high gain recovered antennas that are optimized for channel 12 and were expected to work reasonably well on channel 7.

Channel 7 is on our list. So that just works. Channel 13 is only a few percent different from channel 12 in terms of radio frequency. It would have a lot of gain when used on channel 13. Unfortunately the 13 and 7 transmitters are are at right angles so a single high gain antenna is the last thing we want here.

We have two antennas available. I didn't pay anything for the antennas. Why not point each antenna directly at a transmitter?

In the analog days the answer to the proceeding questions would of been; ghosts otherwise known as reflections. The antenna pointing in the wrong direction for the desired channel would receive reflections quite well. Special filters were required to eliminate these reflections. Digital TV is very good at cancelling out these reflections (relevant post). As a result, multiple antennas can work quite well here in the 21st century.

You can combine the signals with a splitter/joiner:

If your antennas are pointing in more or less the same or opposite directions you should make the cables between the antennas and the joiner the same length. That is in the hope that the two signals will add together rather than cancel. In our right angle case it doesn't really matter. If you have matching transformers on one of both of the antennas you should reverse the antenna connections on one of them to see if things get better.

That takes care of channels 7 and 13 leaving 27, 35, 40, and 51. Here we finally get lucky. With 7 and 13 being in the VHF-hi band and 27-51 being in the UHF band we can use a band separator/joiner. In this case we want a UVSJ (UHF, VHF). The band separator/joiner will completely isolate the 7 and 13 antennas which allows us to consider the problem of receiving 27-51 with no consideration of what we have done so far. Channel 35 is the only remaining transmitter that is not right next door so we just use a low gain UHF band antenna with the hope that we can find a magic angle that will work for all 4 channels. This might work because the UHF band is at a higher frequency where there is less woman made noise.

We end up with this:

This was a very specific example. The point here is that this can be a reductionist process. That can make this a lot of fun for those that are into that type of thinking.

We now have a configuration we can build and try out...

posted at: 15:04 | path: /tv | permanent link to this entry | Comments (0)

OK, what do we need to know to design our over the air (OTA) TV antenna system?

What channels do we wish to receive?

There will likely be compromises. It is a good idea to spend some time thinking about what sort of compromises you would be willing to make. If you are doing this entirely for the technical challenge, then the answer to the question might be "All the channels that are humanly possible to receive from my location." If you are all or partially doing this for the TV then you will likely end up with three lists; Mission Critical (I need these to consider doing the project), Optional, and Don't care.

Where are the transmitters for these channels?

This is normally a matter of public record. The appropriate government agency web site will likely have the coordinates of the local TV transmitters that you can plot on a map. You want to have some idea of where you have to aim the antenna.

How strong are the signals from these transmitters?

The signal gets stronger when; the transmitter is closer, the transmitting antenna is higher, the transmitted power (measured as ERP in Watts) is greater. Signal strength is hard to predict so this is going to be a guess even with the best available information. If you have any TV antenna using neighbours you can ask them what channels they see. Online OTA enthusiast forums can be sources of reception reports for your area. In general you need larger antennas placed higher above the ground to receive weaker signals.

In places with enlightened government agencies that provide TV transmitter data in a usable form people will create web sites that display this data in a way most useful to those messing around with TV antennas. As an example, my favourite site of this kind is TV Fool. After giving the site my location it gave me this:

This site does some sort of modelling and comes up with a signal strength as well as a direction and distance. This signal strength information can be quite handy even if it is not entirely accurate but for the purposes of this example I will ignore it.

What channels can we expect to receive? The people who make TV antennas often claim that their very best efforts can receive signals from as far away as 100km. This is under conditions few will ever see but it makes a nice extreme limit. The distances given on the site are in the delightfully quaint units of "miles" so I will translate into something a little more mainstream as I go along. There are 6 transmitters within 100 km on channels 7, 13, 27, 35, 40, 51. Annoyingly, KNRR on channel 12 is more or less exactly at the 100 km distance (63.2 mi). I happen to know from the reports of others and personal experience that this is a very difficult signal to receive at my location ... so we will drop it. Receiving such a marginal signal is well beyond the level of this discussion anyway. We will place all 6 channels on the Mission Critical list to keep things challenging.

There is a tendency to locate TV transmitters that serve a particular area in more or less the same place. That makes it possible for everyone in that area to point a single antenna to receive everything. Let's see how they did in my area...

So ... not so great. There are transmitters directly to the east, transmitters directly to the south, and channel 13 by itself directly to the west. This is starting to look bad. But let us now group the transmitters in terms of band to get an idea of what sort of antennas we might need. We only have two bands in use. VHF-hi (channels 7 and 13) and UHF (channels 27, 35, 40, 51). That's good. Let us now consider each band separately. On VHF-hi we have channel 13 directly west and channel 7 directly south. In other words, at right angles. That's bad. On UHF we have channel 35 pretty much at right angles to all the other channels. This is also bad but I note that the UHF transmitters other than channel 35 are only about 3km (1.7 mi) away. At that distance it is unlikely that we could do anything to prevent the reception of those channels.

We now understand the problem. We must have enough information...

Edit: channel 13 -> channel 35
Edit: removed extra period

posted at: 22:48 | path: /tv | permanent link to this entry | Comments (0)

Here I will do a show and tell with the important components used for over the air (OTA) television ... but first let's talk about the past...

OTA television was the first kind of television. A kind of transmission line called twin lead was used to get the signal from the antenna to the TV. Twin lead is poorly shielded so the cable TV people used a kind of transmission line known as coaxial cable to keep their private radio from interacting with the public radio in the open atmosphere. This coaxial cable (AKA coax) eventually became popular for OTA antenna work and the use of twin lead is minimal these days.

This situation would seem to imply that there should be components intended for use with with coaxial cable in cable applications that could be used for OTA applications. This is mostly true.

This is what the coaxial cable might look like:

This type of cable is known as RG-6 or RG-6U. There are other types of cable that will work but the discussion is a waste of time for pretty much everyone. Just use RG-6.

You can buy your RG-6 from the local home improvement store without too much fear of poor quality. Just ensure that the cable you get is mechanically reasonable and has a black plastic covering (UV resistance). The outer conductor should entirely cover the inner conductor. This is often done with an extra layer of aluminium foil. Do not kink or sharply bend your cable as this can damage it. If the person at the store seems confused just tell them that you are installing a satellite dish and you will likely be given RG-6.

Front and rear views of a RG-6 compatible connector:

This is known as an "F" connector. The connector does not supply a centre conductor, instead the bare copper conductor of the cable is used directly. This makes the connector very quick to attach to the the cable. It also makes the F connector very susceptible to corrosion. That means you you need to find F connectors that are specifically designed to exclude water and corrosive atmospheres. The connector pictured achieves this with silicone grease in the slot where the cable is crimped and a rubber O ring on the face of the connector. This is the crimping tool used with this sort of F connector:

Note that the die on the crimping tool is as wide as the crimpable part of the connector. This tool is relatively expensive so there is always a temptation to use some other sort of tool to do the crimping. Don't. The tool is pretty much impossible to damage or wear out so it is a good candidate for some sort of shared cost arrangement. Go to your favourite video sharing web site and watch other people attach F connectors, then do a few practise attachments.

The F connector is threaded and will accept a wrench for tightening. It is very easy to cause damage by overtightening this type of connector. If you are not sure of yourself it is OK to just finger tighten an F connector. Poor quality F connectors will often be hard to thread on and/or will cause the cable to twist during the tightening phase. This will not make any difference to the electrical performance of the connector but it can be annoying.

There is another popular type of F connector called a compression type available in a weatherproof version. These are reputed to work well and should be investigated for new installations. They use their own unique crimping tool.

So ... you want to ask the person in the store for F connectors intended for outdoor applications that are compatible with RG-6 cable.

Here is a somewhat obscure part:

This is a dead end F connector with a resistor in it. It is called a "terminator" as its primary purpose is to prevent electrical reflections caused by the discontinuity at the end of an unconnected transmission line. It also keeps the weather out of the connector and the thing the connector is attached to. When you are done with your antenna installation just screw one of these onto any open connectors.

Here's a "balun":

These are also known as matching transformers. The one shown is the type intended for outdoor applications. A balun bridges the world of classic twin lead transmission line and the more commonly used coaxial cable. Many antennas are still designed for the use of twin lead. If your antenna has screw terminals instead of of an F connector then you need to add a balun like this:

The F connector on the balun is covered by a rubber boot intended to protect it from the elements.

This type of weatherproof balun is only used for OTA antenna systems. This means that the manufacturer is expected to produce something with adequate performance for the application. Unfortunately, poorly performing baluns are still fairly common.

This might be a part you will want to buy online. Online retailers that specialize in OTA parts will sometimes provide performance claims (the lower the loss the better). Baluns branded by makers of TV antennas are usually intended to make the primary product sold work well. As a result "antenna brand" baluns can be expected to be not all that terrible. Online OTA enthusiast forums can be a source of tests of popular brands.

Here are some splitters:

The thing that is "split" here is the signal. These are normally used to distribute a signal from an OTA antenna to several TVs and/or tuners. Those of us hung up on the first law of thermodynamics will have to acknowledge that this process has to cause a loss of signal on each leg of the splitter. This sort of thought produces the result that a 2-way splitter has a minimum loss of 3 dB (half power) and a 4-way splitter has a minimum loss of 6 dB (quarter power). A good quality splitter should not be responsible for more than a few extra dB on top of the theoretical best performance. A good 2-way splitter could have a a loss of 3.5 dB (67% of the signal remains) and a good 4-way splitter could have a loss of 7 dB (45% of the signal remains).

There is normally little uncertainty about the performance of splitters found at random local stores. In most places splitters are mostly used to split cable TV which involves very high signal levels. If the resultant splitter lets any small fraction of the signal through the customer will be happy. It is not like the buyer is going to have the equipment to check. As a result these things are sold with performance that ranges from the terrible to the hilarious.

My comments about having to order your baluns online thus apply to splitters as well. Sometimes an electronic retailer will have two levels of splitter with one of them having a price that is two to three times the other for no apparent reason. The expensive one is possibly intended for OTA use. The splitters that are rated to work up to a frequency of 2000 MHz are intended for satellite TV use. They will work but are normally a waste of money. Sometimes you might be forced to buy the satellite TV version if it is the only one available with a good claimed signal loss specification.

A band separator/joiner:

This is actually a sort of analog signal processor. When a signal is fed into the common input (separator) the device will produce two new signals with the channels from the wrong band filtered out on each side. It causes very little signal loss while doing so. For example, a UHF/VHF separator/joiner will produce two outputs; one with all the UHF channels, and one with all the VHF channels. This was very useful in the days when TVs had separate UHF and VHF inputs as it allowed the use of a single cable to carry signals from both bands from the roof.

These days they are more often used in the other direction for their joiner function. For example, a UHF/VHF separator/joiner makes it possible to combine the channels received by a UHF antenna with the channels received by a VHF antenna with virtually no loss. Extraneous signals from the band that each respective antenna is not designed to receive would be filtered out.

Here is some silicone grease:

This is more of a consumable than a part. The cool kids like to smear this stuff on and in any connectors/connections in an attempt to delay corrosion. The grease shown is actually intended to seal laboratory glassware. It was expensive but some random internet commenter claims that this stuff is the best ever for antenna work. That's good enough for me.

I will say one thing for the "DOW CORNING(R) high vacuum grease". It is impossible to entirely remove remove it from anything. That included me. You pretty much have to wait until the thin film on yourself and your surroundings gets spread out enough so it is no longer perceptible. I guess that means it is actually good or something.

Another consumable:

If you are planning to make holes in a roof as part of your schemes it is a good idea to have a small supply of a patching material appropriate to the type of roof. Eventually you will do something less than optimal and things will suddenly turn into a basic shelter issue. This is boring and can detract from your enjoyment of the primary activity. This is particularly true if there are other people living in the house that do not share your enthusiasm for this activity.

Edit: silicon -> silicone

posted at: 15:40 | path: /tv | permanent link to this entry | Comments (0)

Among other things, hacking is seeing the value in that which is considered worthless by less insightful people. This section is about making old TV antennas work in the exciting new world of digital television.

In most places in the world the same frequency allocations are used for digital TV as was used for analog TV. This makes sense as TV takes a lot of bandwidth. There isn't really any place to move it to. In North America some of the higher channels (part of the UHF band) have been reassigned to other purposes. Some of the lower channels (all of the VHF-low band) have been more or less deliberately abandoned by the broadcasters. That is because noise tends to be stronger at lower frequencies and there tends to be a lot of noise around these days. This means that in North America we are normally looking for antennas that work in the VHF-high and the lower portion of the UHF band.

You can often get an idea of the sort of raw material available by simply looking for abandoned antennas on roofs. Here is a common sight in the city of Winnipeg:

From having lived through the era I know that the top antenna with all the elements is likely optimized for receiving a faint signal on channel 12. That antenna was also expected to receive a strong signal on channel 7. Both of these are in the UHF-high band. I want to receive digital signals on channel 7 and 13 so the top antenna is of interest. The bottom antenna was used for receiving channels 6 and 3 in the VHF-low band. My local broadcasters have, as expected, abandoned this band so the lower antenna is of no interest to me.

In most cases you will not know the history of the old antennas you find scattered about the landscape. Then you need to do a bit more work.

TV antennas are made out of conductive rods called elements. The elements go perpendicular to the transmitter and are the active part of the antenna. Parts of the antenna that point directly towards the transmitter are inactive and are usually just there to support the elements. Often times the entire antenna is electrically connected. This doesn't make intuitive sense so it is important to realize that the centre of an element is normally electrically inactive.

One of the elements will be connected to the wire that goes to the TV. First measure the length of the connected element closest to the TV in metres. This element will normally be a half wavelength corresponding to one of the channels the antenna is designed to receive. Plug your measured length into this formula:

             150
Frequency = ------
            Length

The result is frequency in Megahertz(MHz). Consult with a table that maps your local channels/bands to frequency and use it to find a channel and band (TV channel frequencies). If your antenna has a bunch of elements that are more or less the same size (like the ones in the preceding picture) then you know the band and have a pretty good idea of what channel the antenna is optimized for. If the antenna has a bunch of different sized elements then you really only know one of the bands it works at. Chances are it is supposed to work over that entire band. If there are other elements connected to wires then you can measure them and assume that the antenna works for the band(s) you come up with as well. If the antenna has wires connected to all the elements (log-periodic) then you can try measuring the longest and shortest to get some idea of the range of the antenna. If none of the elements are anywhere close to the length required for any of your bands of interest then you can reject that particular antenna without further thought. You can try cheating by looking at pictures of antennas on a manufacturer web site (Winegard, Channel Master, Antennas Direct) relevant to your part of the world.

TV antenna elements are usually made from aluminium. As a result they will last pretty much forever in a non-corrosive outdoor environment. The best forgotten antennas are those that have never been outside at all. These can be found in the attics of houses with previous owners that were interested in not falling off the roof even once. Estate sales and older relatives are possible sources of these antennas.

In the case of an antenna that has spent a lot of time outside, pretty much everything other than the elements themselves and the support structure will be suspect. You will normally have to replace any nuts and bolts that are electrically important. Stainless steel hardware should be used if possible. Any connections will likely have to be redone. Any elements that are non-contiguous (these are sometimes arranged to fold) should be checked for continuity. Continuity between the elements and the support structure is normally unimportant.

I recently was made aware of an intriguing idea for antenna connections at a site I have also recently lost the reference to. The idea is to use stainless steel flat washers to prevent the wire and the aluminium element from coming into direct contact (causes corrosion). The only catch that I can see is that the flat washers would not be able to produce really high contact pressures by themselves. Thus the bolt would have to be made very tight. Here is how I managed that on one of my antennas:

After splitting the plastic insulator due to excess pressure I ended up having to use a nut for tightening and a nut to hold the end of the element in place. That allowed me to tighten the connection to the point where the copper wire just started to deform. Starting from the left end of one of the terminals shown in the picture you are seeing; screw head, flat washer, wire, flat washer, element, self locking nut, insulator, self locking nut. All the hardware shown is of course stainless steel.

Fiddling with connections ends up being the primary maintenance task for an outside TV antenna. Thus it is well worth the time to explore ideas that could potentially increase the number of years between trips to the roof.

Most digital tuners/TVs will show you some sort of signal strength indicator for the currently tuned channel. If you can run a cable to a relatively open area you can get some sort of idea of the performance of your antenna by swinging it around. If the signal is significantly higher when you point the antenna at the transmitter then that is good. If the antenna works pretty much just as well pointed straight up (for instance) then that is bad.

Edit: proceeding -> preceding
Edit: will -> well

posted at: 10:13 | path: /tv | permanent link to this entry | Comments (0)