Antenna Spacing Guidelines

[GroundUrMast]

Here is a single page guide regarding the distance antennas should be from adjacent antennas.

Antenna Spacing

Scoure: TonerCable.com

[ghz24]

Here's a site along the same lines.
You can see what happens if you stack at non-optimal distances or get one of your leads reversed/crossed.
How to use the graph is a little hard to follow until you follow the examples. Near the bottom see the two antenna trick.

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

I guess there's an indoor antenna description to see first
http://www.hdtvprimer.com/ANTENNAS/silver.html#TAT

[thom]

for bay antennas on the same mast, do you measure the distance from the centers or from the lower edge of the upper to the upper edge of the lower? (or something else?)

[GroundUrMast]

I would err on the side of caution, measure from the bottom most part of the upper antenna to the upper most part of the lower antenna.

The reflector and metallic structure of one antenna can have a detuning effect on not only the active element(s) but the tuned reflector and director elements of the neighboring antenna.

[thom]

Thanks for the advice. It looks like there is some conflict on optimal spacing between kyes and tonercable. One says 2/3 wavelength the other says .94.

http://www.kyes.com/antenna/stackluge.html

http://www.tonercable.com/assets/fil...ce/antenna.pdf

Interestingly, kyes also says that spacing greater than one wavelength will actually reduce performance which would explain what I've observed with my own stack.

[GroundUrMast]

I should ask... are you targeting different signals with each antenna? Or are you combining two antennas for the purpose of increased gain? Or are you trying to null out an interfering signal?

If you're simply stacking a UHF over a VHF, more separation will generally be better than less. The exact distance will not be critical. Combining the two outputs is also easy. UVSJ and HLSJ combiners are simple and inexpensive. http://www.solidsignal.com/pview.asp...nna-%28UVSJ%29 http://www.solidsignal.com/pview.asp...-%28ZHLSJ-1%29

On the other hand, stacking two antennas cut for the same band requires some precision. Combining the two outputs also becomes an engineering exercise where fixing one problem can create two or three new ones. http://forum.tvfool.com/showthread.php?t=1024

[thom]

Two DB4e's on the same heading, identical length coax to a Perfect Vision PV22233 combiner, then onto a PA-19 preamp. Looking for increased gain, but if what kyes says is correct, they've been too far apart.

Tried the different direction configuration and that was fascinating if not terribly productive.

[GroundUrMast]

Ahhhh...

Consider, if both antennas receive identical amounts of power and then deliver equal signal power to their respective combiner ports, the combiner is being driven with twice as much power (+3 dB = an increase of power by two times) than that of a single antenna. Sounds great, we should hope that the increase in power may help reception... right? But alas, the splitter/combiner has 3.5 to 4 dB loss between the input and output ports. So, the net gain is between -0.5 and -1.0 dB when compared to the net gain of a single antenna.

There is a way to combine two antennas using transmission line as a combiner: http://www.kyes.com/antenna/stackluge.html. However, expecting more than 2 dB net gain is unrealistic. Antennas that have balanced 300Ω outputs lend themselves to ganging via 450Ω twin-lead harness rather easily. Unfortunately, many antennas are now equipped with integral baluns which would require some sort of modification to the antenna if one wanted to connect twin-lead.

[thom]

Thanks again for the info; I appreciate the feedback. You've hit upon the one thing I consistently find divergent information about. I completely understand that a theoretically perfect splitter will cut the signal strength in half because that's exactly what it's supposed to do.

However, when used as a combiner instead of a splitter, weird things start happening with (typically) out of phase signals getting from antenna A getting pushed back up the wire of antenna B and transmitted out of antenna B and vice versa. If the downleads to the splitter are of identical length, the signals will be in phase and you should (other things being as they should be) get that 3db gain less any efficiency penalties.

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

But you ask, “How can the power at the load equal the total input power if some of the power was reflected backwards?” The answer is that superposition applies to currents as well as voltages. The phase of the reflected currents is such that they subtract instead of add. The reflected currents will cancel each other out completely.

"The doubling of the output power is equivalent to a 3 dB increase in the signal. If the combiner is 90% efficient then a 2.5 dB gain is seen. Note the dichotomy:
· If the antennas point in different directions, there is a 3.5 dB loss at the combiner.
· If the antennas point in the same direction, there is a 2.5 dB gain at the combiner.
This is a 6 dB swing. 3 dB of this is just the adding of the second antenna, but the other 3 dB is from the combiner becoming a much more effective device."

The most compelling thing I've seen so far is this test of combiners:
http://www.antennahacks.com/CombinerComparison.htm
It seems that if you manage to do everything right, you can eek out a 2.5 db gain.

So, it sounds like it is possible to get added gain, just not necessarily easy without a spectrum analyzer or a DisplayMax 800.

[GroundUrMast]

Perhaps the easiest way to describe a practical low loss combiner is to simply have you look at a panel style UHF antenna. The DB4e is just as good an example as any... In the case of the DB4e, they built it from what are essentially two DB2e antennas. The older DB8 is another example of an antenna that is based on four DB2 antennas.

In both cases you can see the primary combining network is made of twin-lead (in the form of side by side heavy wire/rod). The parallel conductors form a transmission line with an impedance of about 1.5 times higher than the antenna terminal impedance.

The goal of these designs is to transform the terminal impedance of each dipole (bow-tie / whisker element) from its nominal 300Ω to 600Ω. This is done by taking advantage of the fact that a transmission line (coax, balance twin-lead, strip-line or even waveguide can be used) can serve as an impedance transformer. http://en.wikipedia.org/wiki/Quarter...ce_transformer

Once we transform the 300Ω impedance of two dipoles to 600Ω each, we can connect them in parallel, which results in a combined impedance of 300Ω. This process can then be repeated to form a set of four or even eight combined dipoles. In theory, one could continue on to sixteen thirty-two...

If there's a downside to this method of combining, it would be that the 1/4 wave long transmission-line / transformer is only 1/4 wave long at one frequency. So, the 1/4 wave transformer becomes less and less efficient as the frequency difference increases. A 1/4 wave transformer is essentially a tuned devise, which makes it less than perfect for application in a broadband system.

In practice, combining antenna elements with a broadband splitter/combiner comes at the expense of loss in the combiner while retaining bandwidth. On the other hand, combiner networks based on 1/4 wave transformer made of transmission line offers lower loss at the expense of bandwidth. If I am combining for the primary purpose of added gain, I would not use a broadband combiner such as a hybrid splitter in reverse. However, if my goal was to combine two antennas for the purpose of nulling out an interfering signal, the loss of a splitter would be of little concern to me.

[thom]

So if I read you right, to get a bona fide DB8e, I need to pull off the baluns, put the DB4e's side by side and tie them together with lateral transmissions lines of equal length, and put a balun in the center of that? Would one of the DB4e baluns work, or would it not be calibrated for the 2X configuration?

In other news, having much better luck with them stacked closer together. Currently 18 inches between the edges of the reflectors, and I have a dramatic improvement in signal strength: the LA stations have become watchably reliable. That puts me close to 2/3 wavelength if I measure from element to element. I'll be doing some more experimenting this weekend to see what yields the optimum spacing. My AntennaCraft Y5 just showed up too, so I'll add that to the mix. I've read some interesting things about a VHF antenna in close proximity reducing the front to back ratio of the UHF antenna (without reducing the front gain) which could improve my reception on the San Diego (back) side.

http://imageevent.com/holl_ands/stac...fqmq81.panda_s

http://www.tvfool.com/?option=com_wr...e77959121bce5a

[GroundUrMast]

If your goal is maximum gain, yes you need to eliminate the combiner losses to the extent possible. However, simply combining side by side antennas can give you a more directional antenna regardless the combiner losses. That can buy you a significant boost in signal quality when faced with multipath.

If you get the impedance matched correctly, you can reuse one of the original DB4e baluns. You can also use a common ferrite transformer such as the TV-2900.

I consider holl_ands work with 4NEC2 antenna modeling software to be among the best available. There's an impressive brain trust gathered together as the group of regular contributors to the antenna modeling/build forum at digitalhome.ca

[thom]

No kidding, a TV-2900 would do the job?

By 'matching impedances' do you mean between the two antennas? How does one go about that?

[GroundUrMast]

Originally Posted by thom

No kidding, a TV-2900 would do the job?

By 'matching impedances' do you mean between the two antennas? How does one go about that?

I attempted to describe that in post #10 of this thread...

An important concept to read up on... develop an understanding of is: A section of transmission line (coax, twin-lead, etc.) that's 1/4 of a wave length long can act as an impedance transformer.

Knowing the actual impedance of the antenna element is critical to the design process. However, the nominal impedance of a folded dipole in free space is about 277Ω... pretty close to 300Ω. If I connected two of those elements in parallel with no transformation of impedance, the combined impedance would be roughly 150Ω. But, if I use a 1/4 wave long section of transmission line with a characteristic impedance of about 450Ω, a 300Ω antenna element can be made to look like a 600Ω source/load. (Two 600Ω impedance's in parallel have a net impedance of 300Ω.)

Antennas that have balanced two terminal outputs (ex. Antennacraft Y5713) have a nominal impedance of 300Ω. 4:1 balun/matching transformers (ex. TV-2900) transform the impedance down to 75Ω (300/4=75).

Forgive me if I fail to condense a year and a half of college into a few paragraphs.

[thom]

Thanks, makes sense.

So I should be able to measure the impedance on the notional DB8e with multimeter contacts on the points where the balun/transformer would attach?

(and likewise with a stock DB4e which would tell me the ratio of its balun)

[GroundUrMast]

Originally Posted by thom

Thanks, makes sense.

So I should be able to measure the impedance on the notional DB8e with multimeter contacts on the points where the balun/transformer would attach?

(and likewise with a stock DB4e which would tell me the ratio of its balun)

Unfortunately, no. A Multimeter or DVM is going to apply DC (0 Hz) to the devise when measuring resistance. The antenna operates in the 50 to 700 MHz range... A DVM does not.

An RF impedance bridge is needed... In current practice, you would be looking at an RF Network Analyzer. These types of test equipment generate RF and measure power, voltage or current reflected from the devise under test.They start at around $5000 used.

For poor folks like you and I, modeling with 4NEC2 software then build, test, modify, test... is the cost effective option.

[mars1162]

I had my brother buy some antennas and a preamp. I put the antennas and preamp on a tower for him also. What I need to no now after the fact is how far should the antennas be seperated. One antenna is a Winegard hd9095 uhf the other is a winegard ya1713 vhf. They are both mounted on the same mast pipe and in the the same direction above the tower. The tower does have a ham 4 rotator in it. I dont know which antenna should be on top or bottom. I also had him get a preamp which is the winegard ap 2870 because it has 2 75 ohm inputs one for uhf and one for vhf. The preamp is now discontinued. My thoughts were that with the 2 inputs I would avoid having to get a combiner or diplixer. I realize this preamp has considerable noise figure about 3 db on both uhf and vhf. If my memory serves me right I think the antennas are about 4 feet apart with the shorter of the 2 on top. Any thoughts on a different amp and maybe a low loss combiner would be appreciated. I am not satified with the results of present setup.

[GroundUrMast]

The minimum separation between the UHF and H-VHF would be about 3', more is better but after about 5' there's very little interaction.

There's not a hard fast rule re. which antenna should be on top... Go with what works best in you application. If there's no difference in reception reliability, the smaller (usually the UHF antenna) wind load would go on top to reduce the stress on the tower/mast.

Please post the rest of your enquiry in Help with Reception and include a link to the TV Fool reception prediction for the specific location. Thanks

[mars1162]

Thanks for the info Ground ur Mast and I will take the post to help with reception