DIY TV Antenna 2.0

[wakibaki]

Originally Posted by tubelab.com

Nope, no math, just a stupid idea to prove a point. This isn't a true diamond antenna since the diamonds are seperated and not even the same size. I couldn't find two hangers that were exactly the same.

Yeah, what point is that?

By diamond do you mean a quad? This is just a variation on a fan dipole or capacitive-loaded dipole, it gives a bit of length over the feed-to-end length due to capacitive loading and some broadbanding. You'd get more broadbanding if you brought the split point closer to the centre and made the 'wings' thinner isosceles triangles instead of diamonds. For a quad you need to connect to the open ends. The fact that you're picking up anything is down to signal strength and blind luck. You've got no idea what the impedance is at the feed point. You could 'design' it by looking at a local Yagi and pulling the dipole to size by eye. A bit shorter than the 'driven' element. You'd probably get a better match to the co-ax without a transformer, it'll be ~75ohms @ 1/2 wavelength.

w

[tubelab.com]

Given enough signal, anything will receive,

That is EXACTLY the point I am trying to make! There are plenty of rather expensive "DTV" antennas on the market. Some of them are nothing more than amplified coat hangers, with minimal engineering effort expended. Many don't even have good amplifiers.

Did you mean spectrum analyzer, or network analyzer?

Not many of us have a network analyzer on their home workbench. I have on on my bench at work, but here at home I will settle for connecting the coax up to my nearly 50 year old HP spectrum analyzer to see what's out there.

For those that are not familiar with a spectrum analyzer, I offer this simple explanation. An oscilloscope displays amplitude VS time. Both scales are usually linear. A spectrum analyzer displays amplitude VS frequency. The amplitude scale is usually logrithmic (in dB's). The FFT analyzer often used to display harmonic spectra IS a spectrum analyzer. The FFT analyzer we are used to works at audio frequencies, but they are available into the GHz range. My RF spectrum analyzer is an old tech swept LO analyzer. It goes from 0 to 1250 MHz.

The first observation I got when I connected the coat hanger antenna to the spectrum analyzer was a surprise to see strong signals from 88 MHz all the way up to 930 MHz.

The first photo shows the FM broadcast band (88 to 108 MHz). There were many strong signals in the -50 to -60 dbm range.

The second photo shows the VHF high TV band (174 to 216 MHz). Channels 7, 10, 12 and 13 are visible. 7 and 10 are Miami channels, 12 and 13 are in West Palm Beach.

The third photo shows a portion of the UHF TV band (470 to 698 MHz). Channels 18, 19, 20, 22, and 23 are visible.

[tubelab.com]

You've got no idea what the impedance is at the feed point.

You are right. I have access to good equipment at work but I am hesitant to bring this into work and admit that I might have an interest in antennas. I have too many responsibilities as it is.

You'd probably get a better match to the co-ax without a transformer

Maybe, maybe not. I could do some real experiments, but I needed a quick antenna using stuff that I had on hand. You could analyze it to death and pick apart all the flaws, but I would rather have the TV working. It does.

You could 'design' it by looking at a local Yagi and pulling the dipole to size by eye.

A Yagi by design has a very narrow bandwidth and is quite directional. I wanted exactly the opposite. In this case I "designed" it by stretching the coat hangers to fit the piece of wood that I had.

The fact that you're picking up anything is down to signal strength and blind luck.

Granted the signals are relatively strong. However multipath can be a killer with the blind luck approach.

The first photo shows the upper end of the UHF TV band. Channels 46, 47, 48 and 50 are visible. Channels 44 and 51 are in the noise. They can be seen at narrower span and bandwidth settings.

The second photo shows a few SMR carriers around 865 MHz, two WCDMA channels in the 870 to 880 MHz range and a few cellular carriers above them. They are from a cell site about 2 miles away.

There are some paging carriers in the 930 to 931 MHz range not shown.

[DF96]

Given that you are using it over a frequency range of 4:1 the transformer is probably helping, on average. You may get 75ohms at the first half-wave resonance, which the transformer will change to 19ohms but at the full-wave resonance you could get many hundred ohms (or more) so the transformer will bring it down.

I agree with your comments about many TV antennas: modern manufacturers seem to come up with some pretty arrangement of wires and then try to make up for the awful performance by adding a preamp (which could easily make things worse by adding intermodulation).

My guess is that in your situation the right answer is some variant of the fat dipole or discone, although 4:1 may require some form of log-periodic structure.

[wakibaki]

If you want horizontal polarisation then a biconical arrangement will do 3 octaves, maybe 4. and give you a constant impedance, probably not too bad a match to 75 ohms. You can build it with a wire frame if you don't exceed 0.2 lambda max spacing. The slant length of the cone is the free-space quarter wavelength at the lowest frequency. Wire all the ends of the 'cone' together. Move the cones in and out to get best response, spacing probably an inch or less at these frequencies, but not critical. Profile of the cone is an equilateral.

w

[tubelab.com]

make up for the awful performance by adding a preamp (which could easily make things worse by adding intermodulation).

Intermod is a much bigger problem here than in many areas. City hall is line of sight, and about 1200 feet away. There are UHF and 800 MHz transmitting antennas on the roof. There is a cell site 1000 feet away, but it is 1.9 GHz. There is a mega cell site about 2 miles away that serves the eastern everglades. It has every cellular carrier on it plus several SMR antennas and it is cranked up. When the right combination of them transmits I lose TV reception on some channels, without an amp! Whacks my cellular reception too. Part of my job is investigating public safety to cellular and cellular to public safety interference.

Many high priced antennas have a rather useless amp. In fact the spectrum analyzer revealed that my Radio Shack "80 channel" distribution amp is actually costing me 3 channels. Time to make a new amp.

Connecting the antenna to the TV nets 54 channels. Adding the amp nets 51 chanels. Why? It chops off everything above 620 MHz. The antenna picks up virtually nothing below 50 MHz, but I see all sorts of "stuff" below the FM broadcast band when the amp is inserted. These are intermod and mixing products. The frequency response problem is costing me channel 9. What?

If you want horizontal polarisation

All TV signals in the USA are and always have been horizontally polarized. Many FM broadcasters have started using circular polarization as a compromize between home antennas which are horizontal, and car antennas which are typically vertical.

TV signals in the USA were assigned "channel numbers" spread across 3 "bands" a long time ago. Channel 1 was taken away to become the 6 meter ham band. Channels 2 through 6 (54 to 88 MHz) are in the VHF low band. Channels 7 through 13 (174 to 216 MHz) are in the VHF high band. Channels 14 through 51 (470 to 698 MHz)are UHF. Channels 52 through 69 were taken away with the DTV transition for the 700 MHz commercial, public safety, and high speed data (LTE) bands. Channels 70 through 83 dissapeared 20 years ago to become the 800 MHz cellular, public safety and SMR bands.

Along with the DTV transition broadcasters got new frequencies, but kept their old channel numbers. So now the number displayed on your TV bears no resemblance to the actual frequency. The "mapping" of channels to frequencies varies with geographical location. So here channel 9 is broadcast on channel 44, beyond the passband of the Rat Shack amp.

the right answer is some variant of the fat dipole or discone.......a biconical arrangement will do 3 octaves, maybe 4. and give you a constant impedance

I had an old military surplus biconical antenna, but I gave it away years ago. It shouldn't be to hard to make another. I think a biconical or modified rhombic would be good choices for Antenna 3.0, but for now I have a working zero cost antenna which can be used as a reference. Let's see the unit of measure for all new antennas is dbch (db referenced to a coat hanger)!

I am going to concentrate on the amplifier next. I had built one about a year ago using a MMIC from Hittite Microwave. It has a noise figure of 1.1 db, an IIP3 of +20 dbm and a gain of 15 db. It is specified for 550 to 1200 MHz, but has some gain down to 160 MHz. It worked well, but drowned inside my leaky housing. I don't have any more and they are not generally available. They were left over from a work project.

There are hundreds of MMIC's out there. I will choose a new one based on IIP3, NF, and gain in that order. Based on what I see on my analyzer I will need a wider frequency range too.

[DF96]

With lots of strong signals maybe you need a tube preamp! Your problem is that with such a wide bandwidth you need to worry about IP2 as well as IP3. 6DJ8 running parallel grounded-grid?

I'm sure Wavebourn will have some wonderful Russian military valve he can tell us about!

[tubelab.com]

There aren't many tubes that will have any gain at 700 MHz. The 6DJ8 certainly won't.

Years ago I looked at using a 5670 for a balanced front end in an HF ham radio for intermod reasons but a pair of fets worked better.

[DF96]

My suggestion was slightly tongue-in-cheek! Yes, I guess suitable FETs would be best - but do you really need an amp? Since your two directions are back-to-back why not make an antenna with deliberately very poor front-to-back ratio but low sidelobe response? Some sort of vertical stacking to get gain. Two wideband dipoles suitable phased would do the trick.

[nic6paul]

I have had my antenna working for a week now. Thanks George!....now the girls can watch all their soaps in the break room