Application Note No. 47899.14  Antenna Gain Measurement

(previously posted tech-note)

I went out this evening to experiment with my 52in see if and how
much better it performs than my loopers on 2304 es 3456.  I had borrowed a
spectrum analyzer from work to measure dBs at the 144 MHz IF, and found a
convenient antenna range near the new Best Buy in Sterling.  The near-field
aviodance criterion requires about 250ft spacing between the source and the
dish, to allow the transmitted EM waves to "flatten out".  ( it is
2*(D^2)/lambda).  If you get too close, the waves will be curved, and you
will measure less gain than your supposed to get. (there is fancy software
that does near-field antenna measurements, but we can't afford it...and
besides, I met a couple of other hams out on the range.

I used a DEM weak signal source for the transmitter, with a little PC board
log-periodic antenna (WA5VJB 2-6GHz), and planted it in a flat field, about
250ft away.  I got a nice peak on the signal on 3456, and then connected the
SA to the IF spigot.  I was surprised to get plenty of S/N, making the SA
measurements pretty easy.  I was able to put the SA in the 1dB/cm mode, to
better see small differences.  I was expecting to need more IF gain, since
most normal spectrum analyzers have a 50dB noise figure. (Owen has one of
them new-fangled low NF units...very nice!).

Anyways, was I surprised!  The 52in dish measured 8.2dB better than my 76 el
looper!  It was only supposed to be about 6dB, from theory.  (we don't need
no steenkin theory...we got measurin radio apparatus...)  I'm not sure if
the looper is sick, or if the dish is awesome, but at least I know what I'm
dealing with.  I wish the dish wasn't such a pain to erect whilst roving.

On 2304, I was expecting about 2.5 dB improvement over the 76el looper.  I
measured 4.5 dB!  Previously, I had measured slightly better performance
with the looper, but my source was too close, and my front yard is
multi-path heaven.  Out on this nice field, I placed the source right at
ground level to minimize ground reflections, which cause gain variations
with test antenna height.

I fooled with the dish focus a little on 2304, but it seemed to be pretty
close to that predicted by mechanical measurements.  I didn't have time to
mess with focusing on 3456, as I had previously found it to be OK, and my
generator was getting low on gas.  (you can't power an instrument with EMI
filters from an AC-inverter...the inverter just shuts down with even slight
capacitance on the load...what a drag!)

My dish used a dual-band feed made from a Campbell's soup can soldered on to
a 1 lb coffee can (tnx to WB3LJK).  I had read from the measurements of others, including
W5LUA that the dual-band feed costs you at least 1.5dB as compared to a good
single-band feed. As a rover, I'll take the loss to avoid erecting yet
another monstrosity.

Anyways, that's it.  I seems to me that a good dish is definitely the answer
above 2GHz...if you can deal with the mechanical aspects.  My dish is
perforated Al, but with my overkill mounting bracket, it is a little
cumbersome.  Maybe I can jury-rig a method of hinging it up off the roof of
the van.

Right after making the measurements, a whole flock of geese came running on
land over to my source antenna, to see what all the fuss was about.  This
was strange.  They were heavily engaged in significant goose-talk, making a
strange gobbling sound.

Future experiments may include a comparison of the dish to a 12ft looper on
1296 motorcycles.

Ain't science grand?

Home, Jeeves