[TowerTalk] Comments on Array Solutions "dissimilar antennas"

[Jim Lux]

----- Original Message -----
From: "Dan Levin" <djl at andlev.com>
To: <towertalk at contesting.com>
Sent: Thursday, May 12, 2005 10:13 PM
Subject: Re: [TowerTalk] Comments on Array Solutions "dissimilar antennas"


> Jim Lux had a number of theoretically interesting comments on the Array
> Solutions app note that suggests checking the phase of two antennae using
an
> oscilloscope.  Since I actually used this technique, and wrote about my
> experience in another app note on that same site ( "Notes on Stacking an
> open sleeve feed antenna and a direct dipole feed mono-bander"
> http://www.arraysolutions.com/Products/opensleeve.htm ) I thought I'd
weigh
> in.
>
> Many of Jim's comments are accurate, but can be dealt with with only a
small
> amount of effort.
>
> In my case, I asked a ham who was only about 700 yards away to provide the
> test signal.  I also asked him to run 1500 watts FSK.  That gave me a nice
> loud signal to play with.
>
> Next, we did this on 10 meters, and we did it at night while the band was
> dead.  That helped to ensure a nice clean signal.
>
> Finally, I put a fairly high quality 10 meter bandpass filter (that I had
> carefully checked for phase distortion using the same dual input 500 mhz
> digital 'scope) in the line, which ensured that I wasn't seeing a signal
> cluttered up with stuff from other bands.
>
> In fact, the signals coming into the 'scope were crystal clear and clean.
>
> Then I fed the signals from the two antennas into my 'scope, and laid the
> two traces on top of each other on the display - not above and below, but
> right on top.
>
> It was trivial to see almost any difference using this setup - the human
eye
> is very good at discerning differences between two curves laid on top of
> each other.  You don't have to measure the difference, only see if there
is
> one.

This works very nicely if you're looking for exact matching.  If you're
trying to get some other phase shift, it's quite difficult.



>
> On top of all that, +- 5 degrees is more than good enough for most ham
> applications.  Stacking gain goes down very slowly as the phase angles
> diverge.  10 degrees of error amounts to around .05 db in lost gain - way
> below my threshold of pain.

The stacking thing is, I maintain, not about eking out gain, but where that
gain is pointed.  And for that phase errors do make a difference.  Whether
it's important is another story.



>
> All in all, the technique (suggested by WX0B) worked extremely well.  It
was
> quick, easy, and plenty accurate enough for my purposes.

My original comments were that the procedure (as presented on the web page)
had some problems.  You've addressed most of them.