Antenna - extreme multipath problem

[AHDTV Board]

A while back I posted a link to a web site about an anti-ghosting
antenna I developed for HD reception. I have created the second
revision of the antenna. Here's the web site;

http://www.prism.gatech.edu/~wn17/

The new antenna is at the link on the bottom of the page. I welcome
your comments and suggestions.

Bill

[AHDTV Board]

Bill:

You'll need some kind of cover for the open end, wouldn't you? To keep bees and birds from nesting up in the can?

Awesome work, btw.

[AHDTV Board]

I've still got the lid.

I guess I could spread some plastic window screening over the open
end. I used the same thing when a bird kept trying to roost at night
inside the throat of my satellite dish's feedhorn.

Bill

[AHDTV Board]

Covering the open end would be simple - just use more screen, only use
fiberglass or plastic instead of metal ...

- JD

[AHDTV Board]

BTW - I think you win the "world's biggest Pringles Can Antennae" award ...

[AHDTV Board]

Bill,

In the new rev, the bowtie antenna receives just as well off the back as
off the front (i.e. 0dB F/B ratio). It seems to me like this would be
the ultimate multipath as RF both hits the bowties from the front and
from a reflection off the back.

This path length is fixed and so it's probably only a problem around
specific frequencies, but in different installations you'd need to be
careful to adjust the spacing between the antenna and the back to
achieve minimum multipath on your different broadcast stations.

For instance, channel 39 has a wavelength around 50 cm. So, if the
antenna is 12.5 cm from the back, then the round-trip difference between
the front wave and the reflected wave will be a half-wavelength which
should result in a big notch somewhere. As your antenna moves toward the
front, you start having half-waves on lower frequencies and other
multiples at the higher frequencies.

Maybe in the real world this doesn't matter much. Your notches may be at
frequencies you don't care about and they may be sharp enough that they
don't cause much trouble. If you swept the antenna, though, I'll bet
they are there.

It's a neat idea. If you wanted, you could send a note to Dr. Steffes
and ask him if he's got any students who are looking for an independent
research project or something like that. With some time, a lot of
testing and a nice report could be put together.

Jeff

[AHDTV Board]

Jeff's comments raise another issue, but with more complicated
geometries. I'll try to explain my thinking as clearly as I can.

A few years ago I built a similar structure and put a Radio-Shack
Double Bow-Tie in the back. I'm not really sure what results I got
from it.

The thing is that the entire cage is reflective, inside and out. So
that while the outside may well block diagonally incoming signal
that would interact with the antenna, the inside will reflect
diagonally incoming signal that would have missed the antenna back
into it. If the inside of the cage could be lined with some RF
absorbing material, then this effect would be minimized.

I hope I've made that understandable.

Anyhow, looks like it's working for you, so great!

-CB-

[AHDTV Board]

I'm not sure that the sides are as big of an issue. If the can is deep
enough, then the incident angles which could enter the can and reach the
antenna with just a single internal reflection (for instance) are pretty
close to straight out of the can. So, a reflection from a mountain 20
degrees off of direct can enter the can, but it would have to bounce
many times in the can before reaching the antenna. Each bounce would
attenuate the signal, so the multipath is reduced. With a little math,
you could probably calculate the size of can needed for any particular
beam-width and number of reflections. So, if you want incident waves
with 2 internal reflections inside the can to have a beamwidth of 3
degrees, you can calculate how long to make the can. Near-straight
internal reflections won't introduce much of a path-length difference,
so I don't think the multipath would be too bad.

That rear reflection, though, looks like a killer to me. It comes
directly from the transmitting antenna, so it is full strength when it
hits the back wall and anything you do in front of the antenna to block
it will block the desired front antenna lobe, too.

The original design had an antenna with a high F/B ratio, which means
the rear reflection wouldn't do much anyway. Maybe a Yagi or LPDA in a
can would be a better approach. The sizes might get unmanageable,
though. In the pictures at the start of the web page, Bill blames two
buildings for his multipath. It looks to me, though, like you'd have to
have an incredibly narrow beamwidth to actually receive signal from
between the buildings but not the buildings themselves.

I wonder, too, if given the extra size and metal Bill put up, an
equivalent larger traditional antenna wouldn't have had the same effect.
Instead of putting up a huge can, put up a huge antenna.

Anyway, none of this is complicated and it would be interesting to see
what real-world effects are with measuring different things all on the
same day on the same roof. If I were a senior looking for an independent
study project, this would look great. Build stuff, play around on the
roof, and there's lots of opportunities for graphs and theory, too. I
know at one point the undergrad antennas class all built homemade
antennas and they measured their performance on the roof and compared
them to theory. It would be interesting if that class explored
non-traditional designs like this and let the students have an antenna
shootout.

Jeff

[AHDTV Board]

Jeff:

I had hoped that the metal bottom of the can would act as the back
reflector that I took off of the bowtie before I mounted it in the
can, both reflecting the wanted signals and rejecting the unwanted
signals. I used the same mounting hardware, so the distance between
the can's metal bottom and the original back reflector remained constant.

Thanks for your input. I'll keep it in mind for Rev. 3.

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>

The antenna elements are insulated from the metal of the can and I
grounded the can to a good earth ground. This (in theory) should
eliminate any accumulated signals from the surface of the can while
shielding the antenna elements.

Thanks for your comments.

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>

I tried putting up a larger antenna but a 360 degree sweep with my
spectrum analyzer showed so much multipath that I basically had signal
in every direction. That's what led me to trying the Faraday cage
approach in the first place.

I like the idea of getting some undergrad students to have a contest
though. Thanks.

Bill

[AHDTV Board]

What Bill has made is a basic waveguide antenna.
I found some info on calculating the frequency of operation here:

http://www.kiawahstyle.com/Antennas/index.htm

There is a lot of interest in building these for the 2.4 ghz band to improve
WIFI range. They use coffee cans instead of trash cans because of the
shorter wavelength

--
--
Dale