Stacking, Ganging, Combining TV Antennas

[roger1818]

Originally Posted by Cottager

I understand how adding a second antenna pointed in the same direction doubles the gain - that is a gain of 3 dB. What I don't understand is where the second 3 dB of gain comes from. How can a splitter become more efficient and go from introducing loss to adding gain?

When combining the signal from 2 antennas, there are signal reflections (both back to the originating antenna and to the other antenna) that result in a 3dB loss (plus any additional loss due to inefficiencies of the combiner). However, for reasons I don't totally understand, when the two antennas are pointing in the same direction these reflections end up canceling each other out and the 3dB loss ends up disappearing completely.

[mclapp]

This is kind of a simplification of what happens and alot of stuff is left out but here goes.

When 2 signals meet together totally in phase from identical antennas you have 2 times the signal. For every doubling of the signal you gain 3db but due to cabling and combiner losses it will be somewhat less (2 - 2.5db typical).

If the signals meet out of phase there will be some sort of signal cancellation which could cut the signal by as much as half. Half the signal will be minus 3db and the combiner and cable losses still apply so it could be more like minus 3.5 - 4db. The max negative effect almost never happens with directional antennas because it would be difficult to get 2 identical signals coming from different directions with equal strength.

There is also the impedance mismatch to consider which could work for you or against you. When you combine 2 antennas the combined antenna impedance will be about half of what the antennas would be alone. Combining (2) 300 ohm antennas will give you 150 ohms where they combine not counting the interaction of the antennas due to their close proximity to each other. Most antennas are not a true 300 ohms across the whole TV band so there could be spots where the antennas are actually closer to 600 ohms which when combined together would give more than 3 db gain net and other places where it may be 100 ohms which it could make it almost nothing. That's one of the reasons if you look at the net gain curves of a 4221 and 4228 (which is just (2) 4221's stacked side by side) you will see there are spots where the 4228 isn't any better than a 4221 and other places where it's 3db or more better.

[roger1818]

If the signals meet out of phase there will be some sort of signal cancellation which could cut the signal by as much as half.

There are a few flaws with this explanation. First of all, phase cancellation can cut the signal by more than one half. If the signals are equal strength and 180 degrees out of phase the signal would theoretically be cut completely. That is why when ganging two antennas together pointed in the same direction you need to make sure that the two antennas are in phase, otherwise you won't get any signal.

Having said that, phase canceling is not a significant factor when the antennas are pointing in different directions since the signal strengths from the two antennas for a given station will be significantly different. Assuming the signal from the desired antenna is at least 10dB stronger than the one on the other antenna (likely the case), the maximum possible signal loss due to phase canceling (i.e. 180 degrees out of phase) is about 0.5 dB. If the signals are less than 90 degrees out of phase, you will get positive interference and the signal will actually be stronger (though you may still have phase distortion, which could cause other problems).

[mclapp]

Originally Posted by roger1818

There are a few flaws with this explanation. First of all, phase cancellation can cut the signal by more than one half. If the signals are equal strength and 180 degrees out of phase the signal would theoretically be cut completely. That is why when ganging two antennas together pointed in the same direction you need to make sure that the two antennas are in phase, otherwise you won't get any signal.

Not true ganging 2 identical antennas 180 degrees out of phase will result in a 2 lobe pattern with a gain near the same as a single antenna. If they are vertically stacked you will probably get some sort of weird radiation pattern but the signals would still not cancel.

If you were talking purely signals not antennas maybe they could cancel but there is way more that goes into it than that.

http://www.hdtvprimer.com/ANTENNAS/ganging.html

I've experimented with this and it works.

[roger1818]

Originally Posted by mclapp

Not true ganging 2 identical antennas 180 degrees out of phase will result in a 2 lobe pattern with a gain near the same as a single antenna.

You are forgetting that the path from the broadcast to receive antennas is part of the signal path that contributes to phase of the signal at the combiner. The reason for the 2 lobe pattern is in part due to the 2 antennas receiving the signal at slightly different times and resulting in a phase shift. Thus at those lobes, the signals are no longer out of phase. The exact position of the lobes is dependent on (among other things) the distance between the antennas and the wavelength of the signals received.

If you were talking purely signals not antennas maybe they could cancel but there is way more that goes into it than that.

I agree. Antenna theory is very complicated, and anything I am saying on here is a gross simplification.

[mclapp]

Originally Posted by roger1818

You are forgetting that the path from the broadcast to receive antennas is part of the signal path that contributes to phase of the signal at the combiner. The reason for the 2 lobe pattern is in part due to the 2 antennas receiving the signal at slightly different times and resulting in a phase shift. Thus at those lobes, the signals are no longer out of phase. The exact position of the lobes is dependent on (among other things) the distance between the antennas and the wavelength of the signals received.

I know what you're saying since the antennas are not aimed directly at the transmitter the wave will strike one antenna before the other which creates a phase shift between antennas then the feed line phase shift will put them back in phase depending on distance between antennas and frequency (simply put).

To say that combining 2 antennas out of phase will result in no signal is only true in that there will be no signal somewhere in the reception pattern at a certain frequency.

To say that the signal could be cut by nearly half as I said is also not true in all situations either and as pointed out could be much more. I guess that's what happens when you try to simplify a complex situation

[stampeder]

From the UK, an excellent resource from a HAM with essential theory that also works for stacking TV yagis:

http://www.ifwtech.co.uk/g3sek/stacking/stacking2.htm#1.

[300ohm]

I was going to post that site too, heh.

Note the "shape" of the yagi aperature, its a horizontal oval that extends beyond the reflectors.

Ive read that the aperature shape of a parabola is a circle that is LESS than the diameter of the parabolic reflector, go figure.

IMO, the aperature shape of the Grey Hoverman or Bow Tie antennas is a vertical oval that extends a little beyond the reflector.

[mlord]

Here's an even better article dealing with the complexities of stacking yagis:

http://www.grantronics.com.au/docs/StkYagis.pdf

[mlord]

Originally Posted by mlord

Here's an even better article dealing with the complexities of stacking yagis:

http://www.grantronics.com.au/docs/StkYagis.pdf

The K7MEM yagi calculator seems to use the same (or very similar) stacking formulas as what this article describes, which saves a lot of fuss for me at least.

Also in that article, is the magic formula for designing a feedline of a given impedance, something my earlier educations omitted:

Z = 120 arc cosh(D/d),
or approximately Z = 276 log(2D/d),
where D is the centre to centre spacing,
and d is the diameter of the wires.

Handy, that!

When stacking two yagis vertically, the custom is to connect their two feed points with appropriately sized/spaced metal rods to act as feedlines, thus wiring the array and stiffening the structure at the same time.

The impedance of the feedline has to match that of an individual antenna, so for a folded-dipole (in a yagi) it's usually about 300ohms. Impedance at the middle of the connecting feedline (between two antennas) will be half of that, or about 150ohms.

(post restored to it's proper location by mlord, minus the DBGH reference that caused Stampeder to move it to a new thread).

[holl_ands]

Additional Stacking references found here:
http://www.kyes.com/antenna/stacking.html

and a couple from Cebik's website, incl stacking separation effects:
http://www.cebik.com/content/a10/yagi/stacksup.html
http://www.cebik.com/content/a10/vhf/2mstack.html

[mlord]

Originally Posted by holl ands

and a couple from Cebik's website, incl stacking separation effects:
http://www.cebik.com/content/a10/yagi/stacksup.html
http://www.cebik.com/content/a10/vhf/2mstack.html

Note: a userid+password is required -- those pages require sign-up before they can be viewed.

[300ohm]

Note: a userid+password is required -- those pages require sign-up before they can be viewed.

Yeah, its free, but they spam your email a little with ads. Not bad though, compared to other websites.

Another good link for stacking: http://www.kyes.com/antenna/stackluge.html

[weehaggis]

I read through the FAQ's but didn't see any mention of this (unless I missed that part).
I have 2 identical 4 Bay antennas (I wont mention the name for fear of ridicule !
I am planning to stack them facing the same direction.
In the FAQ's it talks about wiring the 2 antennas with equal lengths of 300 Ohm twin lead. My antennas have the baluns built into the back. Does this mean that I simply install 2 equal lengths of RG6 through a slitter ?
Because I am using 75 ohm cable, is there no phasing issue ?
Thanks

[stampeder]

Post #16 in the FAQ is the one about combining antennas. I've edited that FAQ post to include the following:

To answer your questions, in your case you would:

1. stack your antennas one above the other so that their reflectors touch
2. combine their output with exactly equal lengths of RG6 into a reversed high quality splitter
3. before nailing anything down permanently, test on an analogue station (easier to see the immediate results than trying for a digital station) to make sure your antennas are in phase
4. if they are not, just switch the balun leads on only one of the antennas and test again.

It should be fine after that, and Bob's your uncle.