A tale of two UHF antennas (can they be combined?)

[blueroomelectronics]

I've recently discovered the joys of OTA HDTV digital broadcasts. But the Canadian broadcasts are to my SW and the US feeds are SE almost exactly 90' apart. I would like to avoid at rotor so a dual antenna (a pair of DB4 antennas) with either a coax switch or better yet a 75ohm combiner. I realize two antennas can cause multipath problems but the DB4 is somewhat directional and there appears to be no overlap.
Anyone out there familiar with dual antenna installations?

 

[Nigel Goodwin]

Use a swtich - it's by FAR the best way.

If the two transmitters are on very different parts of the band though, then you can use tuned combiners to join them (basically a low pass filter from the low one, and a high pass filter from the high one). This prevents the reflections from the other aerial affecting it giving ghosting.

You could certainly try a simple passive splitter/combiner, and see what happens? - but technically it's not a good choice, and it may or may not work.

 

[stevez]

Bill - the amateur radio gang will connect multiple antennas together. Publications that might provide enough insight to help are;

ARRL Handbook for Radio Amateurs
ARRL Antenna Manual
books related to VHF and UHF antenna systems.

I have a handbook or two at home - IM me if you like and I can take a look.

 

[on1aag]

The radio amateur gang does combine multiple antenna's but
they work all at the same frequency and they all point in the
same direction. There's no need for us to combine antenna's
that are pointed in different directions because we all have
rotors.

on1aag.

 

[blueroomelectronics]

Rotors are what I'd like to avoid.
I'm not interested in the analog signals and when you face the antenna directly at the transmitter I don't pick up the other stations so I assumed the dual solution at 90 degrees would not have multipath problems (I'm not an RF guy so I'm probably wrong about that)

I'll buy a quality splitter on the weekend to combine the antennas and see what happens.

 

[Externet]

Please come back with results of using a splitter or paralleling the antennas. Digital television is another animal, there is supposed to be no ghosting or multipath effects.

Miguel

 

[blueroomelectronics]

I'll be happy to post my results. Seems RF is yet another "dark art" when most stuff is trial and error. Even simple RF splitters are not so simple.

 

[stevez]

I was thinking that the radio amateur approach might offer a place to start as the antennas are the same. What I see is some care on length of feedine from tee to each antenna - stuff like that. I've no experience with it so can't comment on if better than splitters, etc.

I do agree that length of feedline relates to frequency - and that kind of makes some stuff almost pointless for a wide-band system.

Bill - I would like to see what you come up with as I am dealing with the same problem - summer place is in Niagara County - across the lake from you. Lots of very good programming out of Toronto, Buffalo and other cities. Like you - I'd prefer to avoid a rotor.

 

[blueroomelectronics]

A worthwhile visit **broken link removed**

It'll tell you where to point your antennas and what to expect as far as signal strength. Works for both Canada & US. Includes a very cool Google Earth plugin for transmitters.

As for feedline I'm using identical 3' lengths of RG6/U and am trying to find a low loss high quality splitter to feet it to. From there it goes directly to a four output distribution amp (at least on paper)

 

[RadioRon]

Feeding two antennas directly into a splitter/combiner doesn't make any sense at all. Its not that the directionality would suffer, since the two directional patterns are pointing in such different directions, so that would be OK. The problem is that in receiving signals off the air the struggle is all about getting the best carrier to noise ratio and putting a combiner/splitter in the wrong way ruins all the hard work that the receiver and antenna designers have done to give you the best carrier to noise ratio (and hence the best sensitivity) possible.

The problem begins with the concept of noise. Noise is caused by many things, not the least of which is brownian motion of atomic particles. We learn that this random movement of atomic particles generates noise power within conductors and semi-conductors and resistors. This noise is something we cannot escape, it is in everything. It is in the coax cable coming from the antenna. When you receive a signal, your ability to decode its meaning, the intelligability if you will, depends on how much noise is corrupting that signal. Just like when you are trying to hear someone speaking, if the noise around you is too strong you cannot understand what they are saying even though the sound they make is entering your ears just fine.

The amount of noise from the antenna is predictable and we call the amount of noise power the "noise floor". It is called a floor because it is big and relatively flat and forms the bottom of our intelligability range.

So the noise power from your antenna is at a fixed and irreducable level. When you receive a signal, you need to receive more signal power than this noise power. But when you put a passive splitter/combiner in your coax, you decrease this signal power but you do not decrease the noise power. Why? Because the splitter output has the same amount of noise power or noise floor as the antenna output. Yet, it will attenuate the signal you are trying to hear because combiners are not perfect, they suffer some losses. So what you are doing is decreasing the ratio of signal to noise (the well named "signal to noise ratio" or SNR) when you put a combiner/splitter on your antenna feedline. You can never get the signal back with all the amplifiers in the world because amplifiers cannot remove that noise power, they can only add noise power.

The effect of putting a combiner in the antenna feedline is to lower your sensitivity by about 3.3 dB. That's a big hit. So, don't do it.

If you must combine in this way, the way to do it is to first amplify the signal from each antenna individually, using an amplifier with a very low noise figure. Now you may combine the two amplified outputs together in a combiner and you won't suffer that big loss of sensitivity.

It is likely that you can buy amplified splitters but these work in the reverse way to what you want. You need an "amplified combiner" with each of the two inputs first amplified then combined. Not sure you can buy this sort of thing. So buy them as separate elements and hook them up yourself. Minor problem of how to get power to the amps, but you can feed DC up the coax if you use a splitter that can pass DC.

 

[blueroomelectronics]

Thanks Ron, I knew this was a dark art

Getting power to the amps is easy, figuring the setup out not so easy. The 3db loss might be acceptable as the signals are very strong when pointed correctly.

 

[stevez]

Note that directional antennas are far from being perfect and if you were to look at a polar plot of the gain you'd see variations - peaks and nulls. Greatest gain is usually where the antenna is aimed. There may be some gain/sensitivity in other directions as well.

All this means is that you might find that the directive peaks are such that one antenna with amps that have been described are sufficent. I know that in my location I can't make this work.

 

[blueroomelectronics]

As I mentioned the transmitting towers are 90 degrees apart from my location. If I point at one there is no signal from the other (well non my TV can see)

The preamp on each antenna sounds good, but will I still need a distribution amp once the signals are combined?

 

[RadioRon]

It depends on how many branches of distribution you plan. The individual amps on each antenna should boost your level enough to tolerate some additional splitting. I'll assume your antenna preamp will have a gain of about 20 dB which is not unusual. The cascade gain equation must be used to determine the net effect of further splitting on noise figure. For example, if you amplify for each antenna with 20 dB and 1dB of noise figure, then combine with a loss of 3.5 dB and then split again to 4 outputs with a loss of 7 dB then you end up with a gain of 9.5 dB overall and a degradation of noise figure of an extra 0.34 dB over that of the preamp or 1.34 total. But, if you split 8 ways instead of only 4, your noise figure goes up to 1.85 dB which is still pretty good. When we add in a few dB of loss for your distribution cable I think it would be quite safe to use only the antenna amps as long as you restrict your splitting to 1:4 or maybe 1:6. However, amplified splitters are not very expensive and it wouldn't do much harm to go with one as long as you are not receiving very strong signals. Amplifiers may overload if you have more gain than you need, in the presence of strong signals.