TV indoor antenna mysteries

[Peabody]

I'm about to head down the cord cutting road, and will need to get local channels OTA.  I have an old indoor antenna, an RCA combination rabbit ears and UHF loop.  It's shown in the attached picture with the rabbit ears fully retracted.  I have three real VHF channels (8, 10 and 11), and three real UHF channels.  All lie in the same direction, which is convenient.

What I've found with the antenna is puzzling.  The picture shows how it needs to be to get good UHF reception.  If the rabbit ears are extended, or if they are removed entirely, UHF reception is poor.  Perhaps the rabbit ears being so short, they just become a second UHF antenna, producing additional gain.

Of course retracted rabbit ears don't work for the VHF channels. But even if they are extended, I need to completely remove the UHF loop to get good VHF reception.  It's as though the loop is interfering with the VHF even though it's a completely different frequency band.

If I run my TV's autoscan using these two antenna configurations, I get either great UHF reception and lousy VHF, or great VHF and lousy UHF.  And this is confirmed trying to watch all the channels.

So it seems to me that I need to spend another $9 to get another antenna just like this one, so the cables would be the same type and length.  And I would need a splitter that works in both directions to combine the two antennas.  I would leave the existing antenna just as it appears in the picture for UHF, and the second antenna would have the loop removed, and the rabbit ears extended to a length that works well for VHF.   That would be a really inexpensive solution if it works.  I guess I would need to be sure the two antennas are not too close to each other.

Does anyone see any problem with that plan?

[tkamiya]

Hi,

Over the years, I've seen and used all sorts of rabbit ears.  What I found out is they are all about looks and not much about science, technology, or performance.  It just need to LOOK like it's going to work well.  Worse yet, we put those things on top of a very electrically noisy box.  So those working well is pretty slim.

What I did was to install small outdoor antenna in attic.  It works VERY well.  I've also mounted one high up on wall.  It works the second best.  I think, the key to good performance is get it AWAY from the TV itself.  I don't know where you are, but Amazon has all sorts from low 10 dollars.  My suggestion is to give up on rabbit ears and go with small real antenna.

[Peabody]

Thanks for the reply, but I've ordered another antenna like the one I have.  I will leave the one in the picture as is for UHF, and configure the new one in VHF mode without the loop but with ears extended.  Then the problematic part is combining the two using Cox's splitter in reverse.  There may be too much loss doing that, or it may not work at all.  These will be a couple feet above the TV, or in a closet nearby.  I'm saving the attic as a last resort.  Anyway, I figured the second antenna for $9.75 was not risking too much.  It might even work.  I'll post the results when the new one arrives.

[cdev]

(I'm sorry this is too long, it needs some editing.)

What you likely need is a single broadband antenna thats horizontally polarized. A so called fan dipole (a bunch of different band dipoles all fed in parallel- would work. But don't use a simple dipole with coax without a balun.

From my own experience, I can tell you right now that the thing you really need the most is to use coax for the antenna run and to use a balun at the far end between the antenna (which can be two equal length pieces of wire) and the coax.

That said, you can do better than that with a commercial rabbit ears or with something like a bowtie (or a horizontal variant of a planar disk antenna) A Bowtie is basically two triangles of equal size. They could even be made out of aluminum foil, stuck to a large piece of cardboard with some elmers glue on the back to make it stick.  Use two pieces of copper tape to provide a small 'solder-able' area for the feed point.

But make sure to use a balun. To decouple the feed line. (and reduce noise)


The impedance is not that important so a 4:1 balun is fine.

You could even make your own balun if you have some of the right (for VHF/UHF) ferrite cores lying around. But most people over 40 or so usually have a few TV baluns lying around the house. They cost as little as $0.25 on ebay.

There are a bunch of different antenna designs on some of the cordcutter sites.

Are the stations you want to reach all in the same compass direction or scattered all around you? 

There are sites where you can get a printout of what directions the stations you want to receive are from you. Having that information would be helpful.

Here are three possible scenarios.

1.) All the stations in the same direction or close.

2.) All the stations in different directions completely.

3. All the stations within 120 degrees of one another and a building behind you that blocks signals from that direction.

Here is a PDF I stumbled upon recently that gives some technical background on the "rabbit ear" type antennas.  https://www.americanradiohistory.com/ARCHIVE-RCA/RCA-Engineer/1975-02-03.pdf

So I guess your original idea sounds like a good workable one.

Putting the rabbit ears in the attic on a 75:300 ohm balun and running coaxial cable down to your TV. 

You can also find cheap low noise amplifiers if you are in a fringe area.

Under some kinds of long cable run situations an LNA would help a lot. Put it up in the attic between the antenna's balun and the cable. If you do this, make sure to get the kind that uses a signal injector downstairs to feed a low voltage to it. 

I have no idea what the quality of units that are sold is but the state of the art in LNA MMICs has improved so much in the last 20 years or so that the chances of getting a decent LNA (for TV use) I think are better now than they ever have been, if you try to be a bit discerning.

In any case they are cheap so it likely wouldn't cost much to try it.

If the rabbit ears are not adequate, you could build any of a large number of very easy to build horizontally polarized antennas which are varying degrees of directional.

Some are really easy to build.

The lowest frequency you need to receive determines the size.

I have one antenna which is basically two triangular patterns that can be cut out of aluminum foil to make a trapezoidal tooth style log periodic antenna.

They need to be either on two separate pieces of cardboard or on a simgle very large long one that is folded in half that applows the points to meet at the middle.

Tt needs to be fed at its front, where they come to a point ideally with a balun there.

The beauty of this is that the dimensions and variations in the substrate dielectric constant, etc (cardboard or foam core approximates air, i.e. 1) are totally unimportant, if you get the proportions right.

It has 9 or 10 db or so of gain.  A lot of gain for something that takes half an hour to make. (For VHF however it would have to be pretty big. Not so practical).

An easy to make antenna could be made out of two pizza tins in a sideways figure eight almost touching, fed right at the middle.  That could also work quite well at lower frequencies, indoors and out of sight. the lowest frequency is a bit lower than the two diameters of the pizza tins would be if they were a simple dipole. I forget the exact amount, but I think its around one fifth again, lets say if the dipole was cut for 100 MHz, two pizza tins of identical length would probably work down to 80 MHz. they should be so close they are almost touching and the leads on the balun should be as short as possible. If you have several balun cores you can also use multiple balun cores to increase the baluns effectiveness.

You could also make it into a switchable 1:1 and 1:4 VHF/UHF balun using a (small) DPDT slide switch. That would be useful to have.



http://www.antenna-theory.com/antennas/wideband/log-periodic-dipole.php

[Peabody]

All of the stations are in the same direction, in one tower farm.

I already know that the antenna I have will receive all the channels, but not in a single configuration.  If I retract the rabbit ears as shown in the picture in post #1, then the UHF channels come in fine, but not the VHF.  However, if I remove the UHF loop, and extend the rabbit ears a bit, I get all the VHF stations.  So it seemed to me that I just needed a second antenna of the same type, connected with a two-way splitter.  The antenna already has coax coming out of it, with an F connector.

It just seemed to me that logically this should work so long as the splitter works.  I still don't understand the antenna's behavior, particularly why the loop is "in the way" when receiving VHF.  The reception with and without the loop in place is very different, with the same rabbit ear extension in both cases.  It just seems to me that the loop should be invisible to the rabbit ears when they are extended, but clearly it isn't.

Well I would put a 50/50 chance on this working because of uncertainty about the splitter.  And maybe there is a better one of those I could use if that's a problem.

[cdev]

Thats great that they are all in the same direction, that makes it easy to address this.

If you use a broadband antenna like the kinds I am describing you can dispense with the splitter.

I would start with the balun, do you have any of those coax to twinlead (4:1) baluns?

You could also use two coat hangers (wire ones, the old style) to make two diamond shaped wings.

You could make a fairly decent gain antenna by putting that at the focal point of a reflector - two pieces of cardboard covered with aluminum foil describing a 90 degree angle. Google "corner reflector".

Does all of this make any sense to you? Is there anything that I'm saying that isn't clear?

One thing that may be non-obvious/non-intuitive. Its the width of the 'dipole' elements that makes the antenna broad band.

[drussell]

Most splitters work fine as combiners when used in reverse, though some have designs that make them not combine properly in reverse.  If you have one of those old VHF/UHF splitter/combiners, it might work even better.  I had to dig up an old one from my stash a few years ago to use on my aunt's TV to combine the VHF RF from her old combo VCR/DVD player with the UHF for our local OTA broadcasts (all our OTA is now UHF here) from one of those home-made "coat-hanger" type (though mine are made with real copper ) 4-element bowtie antennas. 

I had tried a regular splitter/combiner (was even marked as being able to combine) but it didn't work well at all.  The VHF from the combo unit was fuzzy and only a couple of the digital UHF channels were strong enough to tune.  The old-school VHF/UHF filter type works perfectly with no appreciable loss, all the local channels tune fine with the antenna just propped standing up right behind the TV (A 42" Panasonic Plasma.)  The one I used was perfect because it had screw terminals for the UHF side so I didn't need a separate balun for the bowtie, just ran straight to the UHF terminals of the splitter/combiner, feed through the VHF and connect the output.  Easy peasy.

You may find that trying to just combine the two antennas will still give you wierd signal effects due to being multi-pathed from the two antennas, even though they're not really "tuned" to that band.  Separating it out with one of those simple passive filters may be just the ticket in that case.

Good luck!

[drussell]

It just seems to me that the loop should be invisible to the rabbit ears when they are extended, but clearly it isn't.

It would work fine if the antenna had one of those passive filter type combiners in it as I described above, but by the sounds of the extreme interaction you're seeing, the two elements are probably just wired together and therefore affecting each other as one "whole antenna" rather than a two-band double set.  This is very common on cheap antennas.  Ironically, many old vintage antennas would probably work far, far better than most that are available today.

[cdev]

An additional loop in parallel with a dipole antenna is likely to make the pattern of the combined antenna change dependent on frequency, which in this application, where all the stations are in the same place, and where you would have to go upstairs to the attic to adjust it, you likely don't want.

The dipole is only going to have the classic dipole pattern at its resonant frequency, at other frequencies it may work and even work well, but its pattern is going to be different. Which for an analog TV application would likely mean ghosting and undesirable multipath but for DTV would likely just mean the signal would not come in clearly enough to watch.

[PA0PBZ]

I thought of a few comments after reading the first post but most of them are already mentioned by others.

It looks like the rabbit ears and the loop are influencing each other both because they are close to each other and because they are electrically connected.
You can imagine what happens when both the loop and the rabbit ears are picking up the same signal but one is 180^o shifted.

It also looks like all elements are connected together in the base without any high/low pass filtering so the loop is shorting the vhf signals, maybe you can open one up and post a picture?

Now - using 2 of these... You can certainly increase the distance between the loop and the ears but that is no guarantee that the problem will disappear. You still have 2 signals
from the 2 antenna's and by coupling them the signals can easily cancel out (and the coupler will give you at least 3dB attenuation).

While you could go with high- and low pass filtering this will not be easy to do so the only failsafe solution is to use an antenna with an element fit for the complete range of frequencies, examples given by cdev.

[Peabody]

@cdev, I understand what you're saying, but I think what I need to do now is wait for the second antenna to arrive and see if it works as I hope.  If it doesn't I can look for other options.

@PA0PBZ, attached are pics of the old antenna PCB.  It doesn't look like the rabbit ears and the loop are just directly connected.  I see lots of coils and caps.  I don't know if the new antenna will be exactly the same, but I'm only going to use the VHF part of that one.

With regard to the splitter, I understand why there would be a loss dividing  a single input into two outputs.  In fact the Cox splitter I have has "-3.5db" marked on both outputs.  But I don't understand why there would be a loss combining two sources into one (i.e. - using the splitter backwards as a combiner).

[ez24]

watching   

[Beamin]

Depend on how much time/work you want to do.

The spectorLabs FM range extender antenna works excellent for VHF and UHf but needs to go in the attic and is taller than me folded up so 5.5'X4", unfolded It can be easily cut to different wavelengths because of the way it made.


Another idea is make a fractal antenna there are some really good example and some really bad examples (an instructable called "girls can build antennas" or something stupid and she turns it into an art project that totally negates all the important parts like wavelength symmetry etc

The fractal is flat and very directional. You can get it away from the TV and stick it in a picture hanging on the wall. It works for both uhf and vhf because of the way the angles and repeating sides work through complicated math that makes it into an LC network. The one I made was 5 pointed stars on a piece of wood from a fruit box.

I think in UHF mode the VHF elements act like a ground plane. In VHF mode the UHF element probably couples to it. As stated they just try to make these look cool and fall for bigger is better or smaller is better marketing wank.

Good job getting rid of cable life is so much better and you don't have to watch commercials which I find beyond annoying when I go to people houses. Plus 1200$ a year extra and getting 1/3 seconds back spent watching advertisements is priceless. You are paying them to sell you shit.

[T3sl4co1l]

FYI, the bowtie (see above) is a kind of fractal antenna, in the way that is important to wide bandwidth (geometric self-similarity; namely, that a wedge looks like a wedge no matter how closely you look at its vertex).

A lot of fractals (e.g., Koch curve) are meaningless to EM waves, and are more about looking neat, than having a useful bandwidth, polarization or radiation pattern characteristic.

Tim

[drussell]

@PA0PBZ, attached are pics of the old antenna PCB.  It doesn't look like the rabbit ears and the loop are just directly connected.  I see lots of coils and caps.  I don't know if the new antenna will be exactly the same, but I'm only going to use the VHF part of that one.

Hmm... Interesting.  It does have filtering (but doesn't really act like it is working very well), and it doesn't seem like it will be providing a proper match to the 75 ohm coax output...  There probably really should be some kind of balun in there one way or another and there are no turns of wire though a ferrite.  Has anyone sketched out the circuit of the filters to see what they did there yet? 

It would be interesting to see how much the two elements interact with each other in the physical domain by alternately disconnecting the red/black wires going to the VHF side (to see how the UHF-only side performs alone) and then the UHF side (to see how the VHF-only side performs) versus how they are affecting each other electrically when connected through the filter board.

In any case, you will have many different ways you can play with this.  You can just combine them with both elements still connected to their filters, you can use an external UHF/VHF filter/combiner unit to connect the two antennas, with both elements connected or not.  You can try adding a balun on the main output of the one antenna.  You can try putting baluns on each of the UHF and VHF leads and feeding that to a filter/combiner.  You could try running twin-lead directly from, say, the UHF part of your second antenna, placed somewhat physically away from the other one into the filter board you already have so you don't double filter, or you can stack them for theoretically better isolation with the steeper roll-off, etc. 

There are many, many possible possible ways to experiment with connecting the four elements you're going to end up with once you get your second antenna.  A very interesting experiment.  Do you at least have a couple of 300-75 baluns around that you can play with?  I'd be doing that already even with just the one antenna.  What fun! 

[Peabody]

@drussell,  I think I already know how the two sections behave alone and in combination.  I didn't unscrew the contacts, but the rabbit ears unscrew right at the base, and of course the loop can just be pulled out.  For UHF, I found that the loop alone does not work nearly as well as having the rabbit ears there also, but fully retracted.  For VHF, the only thing that works well is rabbit ears alone, and extended, with the loop removed.

I have one balun in my junque box.  But is it possible that the existing PCB includes a balun circuit in addition to the combining/filtering?  I guess I don't know what a balun consists of, but just assumed it was coils in some configuration.

While waiting for the second antenna, I'm going to try using the Cox splitter backwards to see if everything still works at least on one side at a time.  I do have a 75-ohm terminator in case I need it for the unused input.

[NiHaoMike]

In my experience, the "DB4" design and arrays based on it performed the best, with the main disadvantage being that they are physically big for what they do. If size is important, the Yagi included with the old ATI HDTV Wonder performs very well. Any kind of antenna can take advantage of a good LNA as close to the antenna as possible. And it's useful to have a RTL-SDR handy as a cheap spectrum analyzer.

[Peabody]

The second antenna arrived, and I was able to get everything working using both antennas and the Cox splitter connected as a combiner.  However, what was needed to make it work was to remove the loop from the old antenna, and remove the rabbit ears from the new one.  If I left the rabbit ears connected to the new antenna, even if fully retracted, all VHS signals disappeared.  I guess that means that the VHS signals were out of phase, and were cancelling each other out.  One base is placed a bit forward of the other relative to the towers by maybe a foot.  Maybe if they were exactly the same distance from the tower, it would be additive instead.

When I originally tested the old antenna, I had to leave the rabbit ears in place, fully retracted, to get UHF.  But the new antenna doesn't require that.  That may mean the new one does UHF better than the old one.  They are not quite the same part number.  The old one is ANT112R, and the new one is ANT111R.

So now I have one base with only extended rabbit ears, and the other a couple feet away with only the loop, and the splitter combining the two signals.  However, I have found that reception is sensitive to the placement of the two lines coming from the bases.  While they have F connectors on the ends, these are skinny lines, not at all standard coax, so it's not surprising that I can affect reception by touching the lines, or moving them from one route to another.  I guess the solution for that would be replacing the lines with real coax, but that might also require baluns.  It's anybody's guess what the impedance of the skinny lines actually is.

Logically, it seems I should try disconnecting, say, the rabbit ears connectors on both bases, and connecting the ears from one base to the circuitry of the other, in the hope of eliminating the splitter and any losses it has.  But I don't know what kind of wire I should use to do that - it would be up to about three feet.  My guess is it would be 300-ohm twin lead.  But since it's all working as is, I'm going to leave it be for now.

[cdev]

The site which can generate a map for you is tvfool.com I was trying to remember that earlier.


Also, in terms of reducing multipath, its crucial to use a good balun to decouple the coax from the antenna. Minicircuits sells good SMT baluns.

It would be worth it to make your own very small PCB so you could use one.

While you were are at it you could also include the footprint for a small MMIC (low noise preamp) for the antenna end as well as a bypass so you could omit it.

I bet some of the commercial preamps now are basically just that.