[Top] [All Lists]

[TowerTalk] RE: 80 meter 4 square array notes from N0AH

To: <>
Subject: [TowerTalk] RE: 80 meter 4 square array notes from N0AH
From: (
Date: Mon, 30 Aug 1999 18:31:56 EDT
Greetings All-

I have recently constructed a 4 square for 80 meters. 

It has performed wonderfully so I thought I would share some notes about it-  
First, a disclaimer-  I'm not an engineer-  my data is from what I remember 
reading from others and what I've experienced from putting one of these 
antennas together- 

If you want to put one together, here are some notes I hope you will find 

Ok, so whats a four square??  A four square earned its name by the way it is 
put together. It consists of 4 verticals,  placed 1/4 wave apart, in a square 
pattern, that together, act as an array.  You can aim your signal in 4 
directions, diagonally across the radiators, using a phasing system (I use 
Comtek's hybred coupler).  The phasing system should provide power 
distributing and phasing to obtain a directional pattern.

Two verticals can be used but you get much improved front to back, front to 
sides, and forward gain using four verticals.  

The antenna designed was formulated by Fred Collins, (W1FC) and Dana Atchely 
(W1CF) in the 1980's.  Steve Davis, K1PEK helped and the three designed a way 
to phase the array so that it would work well- and be available for others to 

Through various reports, the array provides a reported front to back gain of 
around 20-25db.  The forward gain claims I have read about range from  4-8 
db. A lot of factors go into these reports such as phasing methods, materials 
used for radiators, full size verticals vs. shortened verticals, radials 
systems etc...So needless to say, it is not balck and white on exact gain but 
needless to say, it has a punch- 

There are a lot of ways to construct the radiators.  A lot of hams use wire 
hanging from long side arms off the top of towers.  Others, like me, use 
aluminum radiators while others still use commercial verticals and/or 
insulated towers.  

Ok, so now some notes on how I put mine together-

Based on a design in the Com Tek manual, I used 44 foot tall aluminum 
radiators with three 20 foot long top hats connected at the 39.5 foot level.  
The first aluminum section had a diameter that started at 2 1/8 inches, and 
is tappered down 1/8 of an inch every 6 feet so that the last element has 
about a  1 1/4 diameter.

I used black nylon cord to guy each vertical.  They are guyed at 20 feet, 35 
feet, and at the 39.5 foot level using the tophats pulled out with nylon 
rope. Each antenna uses about 600 feet of nylon rope- I used 2,400 feet in 
all to guy the antennas.  

I used 60 1/4 wave radials per antenna using insulated 16AWG wire.  About 
15,000 feet all together.  My radials lay on top of the ground and are pinned 
in place using landscape pins.  I considered using raised radials to conserve 
wire but other 4 square users suggested a minimum of 60 ground radials- Lots 
of data on rasied radials but none of seem conclusive that a 1-4 wire raised 
radial system would be better than a 60 ground radials per vertical system 
used for an array-

The radials that criss cross inside the square are spliced onto a bus line 
made up of 4AWG copper wire. This is said to prevent unwanted currents from 
developing as a result of criss crossing radials- It is a common practice in 
commercial arrays- (Although several 4 square owners I spoke to just lay the 
radials accross one another- and a few radial guru's told me that it was ok 
to just lay them out)   

The antenna array takes up about 1 acre of land, apprx 46,000 sq ft, and I 
use another 40,000 sq feet to keep everything at least a full wave length 
away from the tips of the array's radials-  Thus a total of 76,000 sq feet 
were used to accomodate the system. 

To get the antenna to work properly, it is important to phase the system so 
that your signal goes where you want to and unwanted signals are rejected by 
a strong F/B, and power put into the array is properly distributed with 
little wasted.

There are a lot of ways to set up phasing on the antenna.  ON4UN's 3rd 
edition on Low Band DX'ing addresses several ways.  But to save time, and to 
go with what had already been a proven product, I went with the Com Tek 
hybred coupler to control the phase relationships.

The hybred coupler sits right in the center of the array, on a small post.  
It has 6 coax ports.  One for the rig's input, 4 for each of the verticals, 
and one for the dummy load which dissapates unused power resulting from 
resonant issues.  Because of the dummy load, the SWR of the antenna is masked 
by the coupler-  So how do you check the antenna's performance?

You simply insert a watt meter between the dummy load and the dummy load port 
and measure the power being dumped by the system.  For example, if you run 
500 watts into the array, and the dummy load shoiws 50 watts being dumped by 
the array, you have 10% of your power not making it out- Not bad-  

But most 4 squares owners, including me, try to get the minimum power dump 
not to exceed 3%.  The antenna will work well with less than 10% of the power 
being dumped. Once you start going up from there- your signal will still 
radiate well but you will notice a decrease in your front to back gain.

To feed the verticals using the hybred coupler,  you need to use 1/4 wave 75 
ohm feedlines.  The feedline should have a velocity factor of at least 78%. I 
used RG11 coax.  A lot of coax is at 66% and if you cut your 1/4 wave feed 
lines taking this into account, you wont't have enough coax to get from the 
coupler to the antenna- 75 Ohm seems to work better than 50 ohm for 
feedlines- Some try 50 ohm but report better results using 75 ohm.  

I feed the system using a 400 foot piece of 9913.........

Tuning and the Effects of Mutual Coupling

When you make the first vertical, you need to test it as a solo antenna with 
it's radial system attached.  You want the antenna to be around 100KHz below 
the desired resonant point because when the other three antennas are up, the 
mutual coupling will raise each single vertical's resonant point by apprx 
100KHz......Thuis, the entire array will be most efficient around 100KHz 
abvove the original vertical's resonant point-

For example, if you want the array to be most resonant at 3.775MHz, you need 
to tune the first vertical constructed to be resonant around 3.675MHz.  

Now this is a touchy subject because mutual coupling does not always produce 
a 100KHz jump.  You might get only a 50KHz jump or have one as big as 200KHz- 
such as the case with mine-  But 100KHz tends to be the average and the 
standard suggested by many 4 square operators-

Of note, you must make a single vertical first, test it, tune it, then make 
each of the other three verticals exactly like the first one.  This way, each 
vertical should act in a like manner as the first, and you will have a well 
balanced array-  Just think of seeing the first vertical in a mirror with 
four images- and that is what you want to produce-

Some four square owners report as little as 50 KHz jumps due to mutual 
coupling on compromised 1/4 verticals such as shortened, top loaded, 
antennas.  Full 1/4 antennas tend to have a bigger jump- but a lot of factors 
will influence this jump such as the radial system, environmental factors, 
materials used for radiators etc...But if you go in expecting a 100KHz 
adjustment, it is a somewhat a safe bet that you be close- 

Ok, so what were some of my results-  Well, first of all, my resonant point 
jumped almost while the feedlines and spacing between the 
verticals are cut and measured for the phone DX window-  I need to lenghten 
my verticals to bring them down a bit-  Right now, my minimum power dump is 
5% at 3.900MHz and around 8-10% in the DX ssb window-  it is a whomping 20% 
in the CW band but I can fix this by reducing the current array resonant 

>From a perfomance standpoint, K0RF, who holds a lot of 80 meter contest 
records, has looked the array over and feels it has as good of F/B and side 
to side rejection as his 3 element yagi-  Measured front to back is as great 
as 25dB on DX contacts and 20-25dB so far on a lot of stateside results.  
Front gain looks to be 6-8dBs based again, on testing the array with DX and 
stateside contacts and using a Butternut HF2V for comparisons-   

Since putting up the antenna just last week, I have worked HL5FUA, 7J4AAL, 
RA0CG, VK5NJ (QRP), VK2XN (dailey it seems with one QRP contact), EA8ZZ, 
CT4NH, and UA0FF...most of these contacts have been in heavy QRN as the 
season is still a bit early- 

The antenna does seem to hear better than the beverages in most situations- 
which is a good test per other 4 square owners- and even though it is a 
vertical array, it is really quiet compared to the Butternut HF2V 

As for tuning, I tested each of the individual verticals and found them all 
to bottom out around 3.900 they are are acting like clones although 
the bandwidth of the SWR bell curves vary by as much as 40KHz.............

I need to add a slight bit of inductive loading or lengthen the top hat wires 
to bring the antenna's most resonant point, as measured my the power dumped 
into the dummy load, down about 100KHz.......but over-all, I could just leave 
it alone and I doubt the 3-4% gain in power I'll get will make much of a 
difference on the phone portion of the band--  it will however bring me 
closer to operating in the CW window a bit more effectively-

You can make the antenna operate over the entire band usaing full size 1/4 
wave verticals per many reports- but you won't be as effective covering the 
entire band vs working inside a specific window- about a 250KHz bandwidth 
with a power dump of less than 10-15 percent seems most likely with my array- 
- but larger bandwidths are reported using more effective materials-   

Materials used and budget:

Aluminum from Texas Towers........................... $450 which included 
400 feet of RG9913..........................................   160
220 feet of RG11.............................................     88
Army surplus parachute nylon for guys 2,400 feet  160
Wood supports and concrete.............................    50
Hybred Coupler................................................   350
Misc supplies...solder, connectors, etc..............     30
Dummy Load...................................................     35
Dummy load feedline 400 feet of the cheap stuff-)   100
Radial wire 15,000 surplus 16AWG insluated.....    200 

Not bad considering a 3 element yagi is around a gazillion $$$ and the four 
square is almost as good- not as good, but close thus far in many tests-

Ok, so if you have any questions, please let me know-  This was an involved 
project but a great learing experience.  I had over 30 hams give me feedback 
on putting this together- and while i didn't do everything they suggested, I 
never could have finished this project with out their support- too many to 
thank but there are a lot of them-  Locally, N2IC, K0KE, K0RF, W0YG, N4VI, 
W0AH to mention a few- and down south, W4AG, Stan, really made my day many 
times as he had almost the exact design for his array as I used in mine- just 
a little variation in vertical height- he really provided a good model to 
build around-

K4SQR, Jim, at Com Tek makes a great product and provided a lot of ideas- it 
was nice to have a product actually work out of the box ...the hybred coupler 
and Jim's design have worked great!

IIf I get some energy, I'll be writing an artical for some magazine on it in 
the near future- any feedback for this would be greatly appreciated-  I'm 
still learning about this antenna array and lack a considerable amount of 
smarts before I claim to be an expert on them-- and I know I may have missed 
a few important points- but from a ham's point of view- I hope this story 
helps others considering such a project- 

- 73  Paul  N0AH    Carpenter, WY

FAQ on WWW:     
Administrative requests:

<Prev in Thread] Current Thread [Next in Thread>