Optimizing TV Antennas with Yagis: A DIY Guide for Better Reception

[jim hardy]

Background:

I never learned much about RF and antennas.

I've made a lot of folded dipoles for TV out of PVC and bare copper wire, with a 300::75 ohm transformer in the middle. They work great in town. My sister's TV (North of Tulsa) found twenty four stations with the antenna just sitting on her couch..My task is to receive two specific TV stations here in the boonies , over the air. They're the only ones within 100 miles that broadcast with significant power ..
I'm ~fifty miles from the transmitters and in somewhat of a valley..
Fortunately their directions are are only ten compass degrees apart.
One is RF Channel 8, ABC , center frequency 183 mhz, halfwave ~31 inches
other is RF channel 20, PBS, center frequency 509 mhz, halfwave ~ 11 inches
(Interesting, ratio of those two is almost e .)

Antennaweb site says i don't stand a chance
but FCC site says i do, should see -50dbm (which sounds high to me).
http://transition.fcc.gov/mb/engineering/dtvmaps/ I'm wondering if i can build two Yagis with folded dipole driven elements, one for each frequency, then couple them with a length of 300 ohm twin lead.
The twin lead would (hopefully) act as a variable impedance transformer to keep the 183 mhz antenna from loading the 509mhz one. It only needs to do that at 509mhz.What I've done so far:

Built a 31 inch halfwave folded dipole for ch 8, 183 mhz. It works but signal is weak, significant dropouts.

Lengthened it to 33 inches which is three halfwaves for ch20, hoping it'd still get channel 8 being only 6% too long. That made ch8 worse, and TV couldn't find ch 20.

So i knocked together a ten element Ch 20 YAGI,
figuring it wouldn't work and i could abandon the project.
Used this guy's calculating program
http://www.k7mem.com/Electronic_Notebook/antennas/yagi_vhf.html
Well, ch 20 booms in just beautifully with antenna only ten feet above ground.
So i have to finish the project if i can.

(Hmm we can't un-indent, can we ? )

So my question is
If i tack a 3 element ch8 YAGI behind the ch20 one on same boom,,
and connect the driven elements with twin lead
It seems to me the shorter 509 mhz antenna will be , at 183mhz, pretty high impedance because it's way too short to resonate hence won't load the longer one
but at 509mhz the longer 183 mhz antenna is only 8% away from third harmonic resonance(549), so has rather low impedance and i fear it might upset the short one.​

Now I've got so rusty at Smith charts i don't know how to approach calculating this :
What fractional wavelength of twin lead will separate the long antenna (183mhz, ~31 inches) from the short one(509mhz, ~11 inches) at 509 mhz. ? No other frequency matters.
I can trial and error it but would sure like to have an idea...

Did i express that clearly?

Maxim has a good tutorial on Smith Charts,
http://www.maximintegrated.com/en/app-notes/index.mvp/id/742
will take me a week to get back to where i was in High School...
I need to get back to that Arduino...

thanks for any advice,

old jim

31 inch(183mhz) folded dipole, stretched to 33 inches(three halfwaves@509mhz)

business end of 509mhz yagi, exceeded expectations

​

 

[skeptic2]

Jim, so many times I have wanted to answer a question only to find you had answered first and your answer was better than mine would have been. I only hope I can be of help here.

I'm not sure if just paralleling antennas will work. The problem I think is that at the frequency of antenna A, the impedance of antenna B will be a problem and vice versa. The antennas need to be kept separated and to do that you need a diplexer (not a duplexer). When I googled diplexer, under the images tab, there were various schematics of diplexers. I think you would be able to modify one of them to suit your purposes. Also Wiki has a fairly good article about them but no schematics.

http://www.n6rk.com/diplexer/diplexer.pdf

 

[jim hardy]

Thanks skeptic. Sorry if I've been a busybody, i do try to not be one. "Old fire-horse" instinct. My last ten working years were largely explaining stuff. I'm just out of my field here with the antennas... In high school we worked Smith Charts, again in an antennas course about 1968, not since... so I've lost the skill with them.Diplexer - i didnt know that term.
I guess i was hoping to use a piece of transmission line for a diplexer.

Google took me here
http://www.maximintegrated.com/en/app-notes/index.mvp/id/3700
and look at this
it splits frequencies when signal flow is left to right

even has a bill of materials.

if it's bidirectional it might give 20 db of separation...

 

[Baluncore]

Yes you can parallel the yagis. You can also do it on one boom.

The best way might be to build two yagis, then place one boom on top of the other. Take your balun and down feedline from the smallest folded dipole. Break the smallest folded dipole again where it crosses the boom and connect those two terminals to the bigger folded dipole drive point. The smaller folded dipole will look like a short, straight through to the longer wavelength folded dipole, while the larger folded dipole will reflect shorter wavelengths. The position of the driven elements sets relative position of yagis. If it works OK, then use one boom with all parasitics in same positions as on the two yagis. The parasitics and reflector for the two bands cannot see each other because they are high Q elements.

 

[jim hardy]

Thanks Baluncore ! I'd have not thought of that.

Break the smallest folded dipole again where it crosses the boom and connect those two terminals to the bigger folded dipole drive point.

The wire connecting the two folded dipole elements, length -
should it be just short as possible
or a specific fraction of 509 mhz wavelength ?

I'm studying up on Smith Charts again...
http://www.antenna-theory.com/tutorial/smith/chart.php

Thanks !

 

[Baluncore]

I would make it as short as possible. Sometimes the two can be folded from one wire.

 

[jim hardy]

Thanks ! Looking for suitable wire now...

 

[jim hardy]

Thanks Baluncore
on a 60 inch pvc boom
i got a six element 183 mhz and an eleven element 509 mhz yagi, connected the folded dipoles as you suggested.

509mhz elements are #12 copper and the longer 183 mhz elements are #6.

TV found both stations. It has no signal strength meter(at least not one that I've found) so will take a few days to know if we'll get dropouts in bad weather.

But initial results look great !

Thanks again,

old jim

ps i used this guys yagi calculator.

http://www.k7mem.com/Electronic_Notebook/antennas/yagi_vhf.html

A purist would have stared from NBS Technical Note 688 at
http://www.k7mem.com/Electronic_Notebook/antennas/yagi_vhf.html

ooops make that http://tf.nist.gov/general/pdf/451.pdf

 

[davenn]

ps i used this guys yagi calculator.

http://www.k7mem.com/Electronic_Notebook/antennas/yagi_vhf.html

A purist would have stared from NBS Technical Note 688 at
http://www.k7mem.com/Electronic_Notebook/antennas/yagi_vhf.html

you put the same link for both references ... the lower one should be the NBS site addy

for transmitting, it pays to be a bit more "purist" for receive only things are not quite so critical

Dave

 

[davenn]

Jim
A quote from your opening post

I'm wondering if i can build two Yagis with folded dipole driven elements, one for each frequency, then couple them with a length of 300 ohm twin lead.

just some random info note for you ...

whilst a folded dipole does indeed have a 300 Ohm feedpoint. that impedance drops significantly
when it is part of a Yagi array. For practical applications, it is common to use an impedance of 200 Ohms and a 4:1 BALUN
to match to 50 Ohm coax. this of course works out quite well for commercial and amateur radio comms that use 50 Ohm coax

also a straight dipole is nominally a 75 Ohm feedpoint and once again when used in a Yagi, its feedpoint impedance drops
down to around 25 Ohms. It's common to use gamma match or delta match to match 50 Ohm coax to these feedpoints

Dave

 

[Baluncore]

But initial results look great !

That is good news. It is such a simple fix.

25 years ago I used the original DL6WU yagi design code and rewrote it for my own use. I then generated data files that went to NEC modelling for optimisation. I was always surprised at how well the DL6WU code worked without optimisation. It was rare to get more than one dB of improvement.

The double folded dipole was shown to me for comment way back then on a rare commercial product by a TV antenna guy. I have been looking for a good use for the method ever since. It has always worked in the NEC models, but has never been mass produced by a client as they are always far too “style” conservative. They knew that colour TV antennas had coloured plastic bits and came in coloured boxes.

UHF yagis have a smaller aperture area than longer wavelength VHF yagis. The UHF yagi therefore needs more parasitic elements, so higher gain can make up part of the aperture difference. Unfortunately it is a wavelength squared aperture cross section so, for the same boom length, you will never get back that loss of signal at UHF compared with VHF. For that reason you should optimise your double folded dipole for the UHF signal.

I assume your antenna is for digital TV, so being in a valley, passive reflections from the “RF illuminated” hills behind you do not cause ghosting. You may notice a change in signal due to refraction over the forward ridges after rain, seasonal vegetation change, or new construction.

Sometimes aiming at a passively reflecting tin shed on a nearby hill will give better results than the direct signal. Again, wavelength squared comes into it, but also the fourth power RADAR reflection equation. Passive reflectors do not work well with multiband antennas because reflectors are usually wavelength dependent.

I am in a deep RF quiet valley. My biggest UHF TV signals came from half a dozen natural reflectors high on the opposite side of the valley. I used a high gain, shoulder mounted rhombic to identify the passive reflectors, by sighting along the bore/boom. The wire elements of the rhombic were tapered cages of three copper wires. I now have VHF digital TV, the ghosts of eras past have been exorcised and so are now spread into spectral noise.

 

[jim hardy]

thanks guys

fixed NBS yagi link

 

[jim hardy]

For that reason you should optimise your double folded dipole for the UHF signal.

Thanks Baluncore

yes it's for digital tv
My first try was your suggestion, cascade the two loops.

i found a signal strength display in the TV's menu and the set reported UHF signal a whole lot weaker than the VHF one.
So i disconnected the VHF folded dipole loop with no noticeable change on UHF but of course VHF got a lot weaker. Just the proximity of those longer VHF elements seems to couple that signal into the uhf loop. I didnt expect that , but it does. The loops are only about 2 inches apart

Then i re-connected the VHF loop but this time just paralleled them at the balun.
No change on UHF but VHF of course shows very strong .
If those undefined numbers in my TV display are db, i have weak signal , 5 to 10"units" and dropouts on UHF(509mhz, us ch 20), but mid-fifties and no dropouts on VHF(183mhz, US channel 8) .
Will continue tweaking. Didn't even think about aiming at my neighbor's big metal workshop...

more experiments tomorrow.

Thanks for the help... Nearly there, i think !old jim

 

[dlgoff]

If those undefined numbers in my TV display are db, i have weak signal , ...

Do you have a way of measuring the field strengths? If not, I could send you this "beauty" to use for a while.

 

[jim hardy]

wow, what a treasure that is !

I'm still "getting the feel" of the TV's strength meter
so before you go to the trouble to pack it, let me see what happens in this weekend's experiments.

Both signals got a lot stronger as soon as the sun went down

Since i found the TV's signal strength menu i feel more confident. It was hidden under the "Product Support" telephone icon...
it reports SNR_db and AGC_%

So as i get a feel for those numbers this should get easier.

Spoiler: But,

,,,, it would be nice to anchor all those computed numbers to something real like microvolts...

old jim

 

[Baluncore]

Jim. There is another reason for optimising the UHF. The UHF yagi has more elements, so it has a narrower beam than the VHF yagi. If your transmitters are close but not co-located, aim the yagi to optimise the UHF signal, then settle for a few dB less than optimum on the usually stronger VHF signal.

 

[jim hardy]

Thanks Baluncore and dlgoff and Davenn

i tried Baluncore's second suggestion first
my prototype was two yagis on one beam.
It worked quite well and i should have stopped there.
The 12 gage 509 mhz elements were a bit flimsy
and , obsessive-compulsive slave to excess that i am,
and mildly dyslexic to boot,
i just had to try his first suggestion second.

Here's two booms ten degrees apart, Yagis for 183 and 509 mhz.

( disclaimer- Fair Anne is amazingly tolerant of my cluttered corner. But she says "I just gave up" )

everything is #6 copper grounding wire (55 cents a foot)
Folded loops are paralleled at the balun
unfortunately i placed the mechanical mount near the loops where my right hand is, not at cg ~halfway between my hands.,.
But that problem only showed its head later.

here it is in the yard, in "Down Periscope" position
booms point right at the towers ~50 miles away , and each is "peaked" according to microvolt meter

Side view show it nose-heavy as you'd expect
my "Plumber's Nightmare transfers the bending moment from 3/4 to 1.5 inch PVC and it seems sturdy enough.

It feels a lot more stout than it looks.
Quite rigid actually. It's the skew of the booms gives optical illusion of bent.

I'm getting to back of TV around 60 microvolts on UHF ch 20 ,509mhz, 180 microvolts on VHF ch 8, 183mhz.Anybody who wants to try a similar project will find this site useful
http://transition.fcc.gov/mb/engineering/dtvmaps/
give it your address and it shows compass direction to transmitters and expected signal.
Its prediction was more in line with my results than antennaweb.org, of course the latter want to sell fancy antennae.

This fellow's site is handy for calculating yagi dimensions.
http://www.k7mem.com/Electronic_Notebook/antennas/yagi_vhf.html
i wasn't sure what velocity factor to use for bare copper parasitics, so let the program use his default of 0.96.

My prototype is now on my neighbor's front porch at top of the hill. It delivers 1000microvoltsvhf, 300 UHF to back of his TV.Thanks guys for the help !

old jim

 

[jim hardy]

When thunderstorms a week ago blocked out satellite, our Yagi's stood tall !

 

[jim hardy]

by the way - weak UHF on prototype turned out to be a lossy length of co-ax that's now in the metal recycle bin.
Replacing it brought UHF signal from <10 microvolts to ~60, some 16db

Two on one boom works fine, as Baluncore said it would.

I learned a lot of different things . PF is a great resource. Thanks again to all.

 

[Baluncore]

The velocity factor on bare copper will depend on the thickness of the oxide / acetate that forms over time. That is modified by variation of the moisture content in the surface chemistry. I spent some time measuring the vf on antenna wires. Aged beryllium copper wires in unsaturated air had a typical velocity factor of 0.997
Long wire traveling wave antennas work better as you approach vf = 1. That is why you should not use thick insulation on those antennas.

The capacitance of the boom and other nearby elements will probably have more effect than your exact choice of vf.

 

[jim hardy]

I spent some time measuring the vf on antenna wires. Aged beryllium copper wires in unsaturated air had a typical velocity factor of 0.997

Thanks. There's a lot of speculation about it on the internet but not much in the way of measured results.

Best i could find was maybe .98 or .99 was a likely number
but my elements go through a PVC boom.
PVC has a dielecric constant around 3.
I put a piece in the microwave and it didn't heat up, so i figured i'd just use that ham's calculator.
He had the foresight to ask whether boom is nonmetallic , and says he based it on the NBS tech note 688 so i rather trust his work.Since a TV channel is 6mhz wide , 3% at 183mhz , i figured it'd be okay if i were a tad short on the VHF.
On UHF the vf adjustment would amount to around 0.3 inch difference . .
I held my parasitics to 1/16 inch so i might be optimized for 524 mhz instead of 509.,, two channels higher..
688 shows in fig 8 i'd rather be a bit short than a bit long.
This one's working quite well.
Next one i'll try vf = .997.How'd you measure that vf ?Thanks again

old jim