[Top] [All Lists]

[TowerTalk] 80M 4SQR data revisited from N0AH

To: <>
Subject: [TowerTalk] 80M 4SQR data revisited from N0AH
From: (
Date: Mon, 3 Sep 2001 12:28:33 EDT
This is a follow up to my posting earlier this week in which I felt I 
improved my 80 meter 1/4 wave vertical arrays performance by using offset 
verticals vs one's more perfectly matched in an earlier version.  My original 
version had four 44 foot tall verticals tuned at the same (+-5KHz) resonant 
point while the offset version had three 44 foot tall verticals with one 
tuned at 3.600MHz, two at 3.550MHz and one at 47 feet tuned at 3.500MHz.  

After seeing the effects of what seemed to be a broader bandwidth and more 
efficient array using slightly off set verticals in my 4SQR, compared to my 
original model notes I built two years ago, I modified the verticals back to 
where the matched exactly +_5KHz at 3.550MHz, as with my original version, to 
see which array was better.  (A lot of work but I had to know!)  Below is a 
data table with the following measurements of power dumped into the dummy 
load using 100 watts into the array.  Power dumped shows the effectiveness of 
the array as it measures the unused power due to resonant issues.   The less 
power dumped, the better the array is tuned for that frequency and the better 
the F/B is.  The array is using top loaded verticals, made from 2 1/8 down to 
1 inch aluminum tappered 1/8 of an inch in various lengths. I was using a 
Comtek Hybred coupler with 4 antenna hook ups and a power dump port to hook 
the dummy load to using an inline power meter at my desk.  (Visit for more details on how this phasing system works-) Ok, so 
here are the numbers:

                                     Power Dumped      Power Dumped
Freq in MHz                    Offset Array          Matched Array
4.0                                      13%                    11.5%
3.975                                   12                       11
3.950                                   11.5                    11
3.925                                   11                       10.5
3.900                                   10                         9
3.875                                     9                         8.5
3.850                                     8                         8.5
3.825                                     6.5                      8
3.800                                     5.5                      8
3.775                                     5                         6
3.750                                     4.5                      5
3.725                                     4                         5
3.700                                     4.5                      5
3.675                                     5                         5.5
3.650                                     6                         6
3.625                                     7                         8
3.600                                     9                         9
3.575                                    10                       10
3.550                                    11                       11
3.525                                    13                       14
3.500                                    15                       16

Not much of a difference and the F/B was better on the matched array, 
although at the time, it did not seem that bad on the offset array.  The goal 
of the offset array was to improve bandwidth on similar size verticals, but 
in comparison, both arrays have almost exactly the same.  

I was very interested in this comparison as the initial results of the offset 
array seemed to show better performance in comparing notes to the 44 foot 
version I put up two years ago.  But since the revised 44 foot matched 
version is also better on paper then the original array, there does not seem 
to be any benefit to using offset arrays to increase bandwidth. Basically, I 
got all excited over nothing-

 The only difference between the original 44 foot matched array I built two 
years ago and this one, is the top hat material and the fact I used small 
inductors to bring down to resonant point of the array apprx 100KHz to be a 
better player in the DX phone window.  The old array had stranded insulated 
wire with the new version is using copper weld.  So here are some conclusions 
from all of these comparisons:

1.  If you use inductive loading to try a match your verticals, you can bring 
down the array's resonant point but I believe my tests over the last two 
years shows that you will narrow the usable bandwidth of the array and 
slightly increase the minimum power dump level-Without inductors, I got the 
array's minimum power dump to 5%, with the coils, I could not get it down 
below 6.5-7%.  (Who's going to hear that any way???)

2.  From verticals ranging in length from 44 feet to 52 feet, the array's 
performance is dramatically increased in broader bandwidth and less minimum 
power dumped by  maximizing the length of the vertical element and keeping 
the top hat lengths to a minimum.  In other words, go full 1/4 wave length vs 
top loading if you can wing it- 

3.  Offsetting the verticals in an attempt to broaden the array's bandwidth 
does not work.  The only reason I think my minimum power dump was slightly 
less using the offset verticals is that one of the verticals was 47 feet tall 
with the other three being 44 feet tall.  This little bit of height on the 
one vertical seemed to effect the array, slightly enough, to lower the 
minimum power dumped, as in the case of using longer elements.  But this was 
to no benefit on bandwidth.  When I redid the 47 foot vertical and made it 44 
feet, to match the other verticals, and modified the one 44 foot vertical 
with slightly longer top hats to bring it's resonant point down from 3.600 to 
3.550MHz, the results were 4 verticals with very well matched resonant points 
that only increased the minimum power dumped from 4% to 5%.  Hardly 

4.  In every 80 meter 1/4 wave vertical array I have built, the power dump 
always increases at a slightly faster rate at frequencies below the array's 
resonant point vs frequencies above.  Therefore, I would suggest if you are 
using shortened verticals for an array, you may want to tune 20-40KHz what 
you think is the mid range if you want to try and play both the SSB and DX 
windows on 80m meters.  For example, tune for around 3.625MHz vs 

5.  When you put up the array for the first time, be ready to bring it down 
again- For until you know just how much the array's resonant point will jump 
from where the antenna's are tuned, you will a lot to do it you attempt to 
make it a permanent first time set up-  I was told expect a 100KHz jump, but 
all 4 of my array versions jumped +-175KHz- Thats what got me started on the 
inductive coils on the original version in the first place!  I built the 
array to be most resonant at 3.800MHz, so I built each vertical to be 
resonant at 3.700MHz......when it jump to around 3.875MHz, the journey began- 
 Now my semi final version has no inductive coils, but I really do need to 
bring down the individual verticals resonant point some more from 3.550MHz to 
3.500MHz.  This way, I could be dumping 8% of power into the dummy load in 
the phone window and 11% in the lower portion of the CW window versus the 7% 
average in the phone window but 14% average in the CW window- I'm getting 
If I ever started this project over again, I would use Rohn 25 towers- the 
aluminum just has too many variables to keep the array up in the air- With a 
file 4 inches thick, I think I'm done learning any more about 1/4 wave 
vertical arrays for awhile- 

73  Paul  N0AH     

List Sponsored by AN Wireless:  AN Wireless handles Rohn tower systems,
Trylon Titan towers, coax, hardline and more. Also check out our self
supporting towers up to 96 feet for under $1500!!

FAQ on WWW:     
Administrative requests:

<Prev in Thread] Current Thread [Next in Thread>
  • [TowerTalk] 80M 4SQR data revisited from N0AH, <=