I would be interested in hearing what kind of phasing is being used on
the K9AY four square. When I looked at doing this I was never able to
achieve performance as good as I thought might be possible. Using an
optimized phasing (maybe mine wasn't really optimum) produced better
performance than quadrature phasing but this complicates the electronics
significantly when switching bands.
In either case, the front-to-back performance of this system is
exceptional, however the front lobe is wider than I had hoped for, and
there were no nulls at 90 degrees from the forward direction when using
quadrature phasing, although I was able to achieve a null at 90 degrees
when using optimized phasing. The system with ANY phasing is much better
than a single K9AY.
I wonder why the system is not being used on 160 meters. It should
perform well on 160. The gain will be significantly lower, but that can
be taken care of with a good preamp. Performance on 40 should be good,
but not exceptional, because the take-off angle is getting too high.
Joe is correct about there being many chances to make a mistake in this
system and cause the performance to go to nothing. I spent considerable
time analyzing a 4 element EWE array looking for ways to optimize it,
and evaluate the contributions of all the misc items like coax lines,
metal supports, ground resistance, transformer common mode rejection,
cable decoupling, and interaction with other antennas. I found some
very surprising results. I'm sure most people don't give much thought to
some of these items, and this can be the reason the antenna doesn't work
well for them. Many of the items I found also apply to the K9AY and
other similar antennas.
After looking at several alternatives, I ended up building a 4 element
EWE (two elements used at a time, fed end-fire). When comparing this to
a K9AY four square using quadrature phasing the EWE was better. That is
the reason for my question about what phasing is being used for the 4
square, maybe I missed something in the analysis. Construction of the
EWE was also simpler than for the 4 square. The first version of this 4
element EWE was a compromise size for operation on 160, 80, and 40. For
the second version I optimized everything for max performance on 160,
giving up 40 entirely, and some performance on 80. Well, I didn't
optimize EVERYTHING, I limited the size considerably. I tried to make a
best estimate of the performance of the EWE compared to a beverage. I
found the performance to be closest to that of 750 foot beverages on 160
meters. Unfortunately I took down my 750 foot beverage last summer and
didn't put it back up so I can't provide experimental data to back that
up. The conclusions were made by comparing calculated response curves.
Please note, as mentioned above, it is easy to make this performance go
to nothing by overlooking the misc items. Making a beverage perform well
is super simple compared to making these small phased arrays work up to
Joe said that in his observations any wire longer than 60 meters beats a
single K9AY on 80 meters. That agrees with my analysis also. It is also
interesting to note that the S/N of a K9AY improves as the frequency
goes down, whereas the S/N of a beverage improves as the frequency goes
up (within a limit).
Joe Szeliga wrote:
>K9AY is great antenna. But big improvement in reception
>is obtainable to form it to 4 square.
>20 set's was build and 13 are in use with big success
>(7 sets are sill available on my stock.)
>K9AY & 4SQUARE ...??? Yes !
>The magic boxes including Phasing Networks and Switching Directions of 4
>Square K9AY Array don't make magic its work perfectly,
>remotely controlled from shack via feed line
>Jay was nice to see you in Friedrichshaffen
>cu soon in USA.
>I'm happy that all with the Hamation bandpass filter unit
>has his happy finale.
>Joe Wojciech Szeliga, SP9P
>>The K9AY new loop is still Gary's great antenna design.
>>It does require a ground but not 1/4 WL radials only some that are say
>>16 feet long if you need them at all with the required ground rod.
>>In the latest incarnation of the K9AY loop that were doing not only do
>>we have a ground rod provision but include a nice milled radial ring
>>under the fiberlass poles in the mast kit to allow radials to be easily
>>attached to beef up the ground if needed, like coral reefs, rocky soils,
>>top of your garage roof, etc.
>>It will perform as the spec. says, and it will allow DXpeditions and
>>postage stamp backyards to play with the big boys with beverages if
>>The loop is a great antenna idea and is like a short two element
>>veritical with 25-30+ db of F/B. So the noise is reduced and you can hear.
>>It needs to be set up with consideration to the TX antenna like any RX
>>array. Furthar away is better. But it still works and doesn't cost an
>>arm and a leg!
>>Next step is to take this very directive element (array) and make it
>>into a larger directive array. Which should open some eyes since were
>>starting with a directional antenna and not a whip. Gary has agreed to
>>help with this design. It doesn't suffer from being an open FET voltage
>>sampler antenna, it is already a terminated matched device so it matches
>>to your feedline without problems and doesn't need special inductors/
>>swamping resistors inside. Another advantage to those of you who know
>>about preamps and band pass pre-filters.
>>It also is not an open FET or transistor input, it has the properly
>>matched band pass filter ahead of the gain stage, so IMD is very small.
>>We test it for this.
>>Don't believe the negativity of Gary's design that you read on TTalk.
>>Please. This is a well designed arrray, Gary knows his stuff. Many DX
>>peditions have used it to hear your signals.
>>And I have compared it to my 6 direction 600 ft beverage arrays. It's
>>almost as good. Just like any antenna, if it is poorly erected it will
>>Merry Christmas to all.
>TowerTalk mailing list
TowerTalk mailing list