[Top] [All Lists]

[TowerTalk] problems with 10m COMTEK 4 sq unit

To: <>
Subject: [TowerTalk] problems with 10m COMTEK 4 sq unit
From: (
Date: Sun, 18 Mar 2001 14:03:02 EST
I finally got around setting up 10m 4 square (WPX SSB coming :-) and run into 
some problems.
First I checked the ACB - 10 unit on the bench, loaded the 4 antenna ports 
with 50 ohm resistors, attached 50 ohm dummy and using MFJ 259B checked the 
I obtained the following
F is min SWR frequency, following done without resistor at dumping port 
(looked better).

sw. pos.       F       R         X

1               28.6      58        5
2    NE      27.6      53        0
3               27.8      57        0
4               29.3      56        6

If I connect 50 ohm dummy load to the dumping port then I get

sw. pos.       F       R         X

1               28.2      70        20
2               28.2      67        16
3               28.2      70        19
4               28.2      66        22

When I checked my 160 and 80 m units, loaded with 50 ohm resistors at the 
antenna ports and 50 ohm dummy at the dumping port I get dips within the 
bands and R around 46 ohms and X = 2 to 3.

When I connected Comtek unit to 4 square of vertical half wave straight 
dipoles, bottom tip about 2 ft above ground (modeled individual impedances 
around 110 ohm, individual dipole tuned to 26.3 MHz) SWR is way out, 
impedances jump all over within the band. On receive I can see some 
directivity, but would not dare to load full power into the system.

It appears that Comtek ACB-10 unit is not working properly (judging by 
behavior of 160 and 80m units.) On 10m length of wires/traces (abt 5") 
between the hybrids, relays and connectors within the unit becomes 
significant and might contribute to imbalance. To be picky, the length of 
traces should be made known so the feedlines can be properly trimmed. It 
appears that there might be something wrong with toroid hybrids.

Nowhere in the instructions does it say to what impedances radiators should 
be adjusted to. Only note saying that you are supposed to use 75 ohm foam 
quarter wave coax between ACB and radiators and 50 ohm if Inv.L are used 
(present lower impedance).  Examples are given for 4 square of V shaped 
dipoles and mention that you can use the unit with 4 square made of dipoles, 
verticals, shortened verticals. 
I assume that that 75 ohm quarter waves are used to double the radiator 
impedances and connected in parallel when fed 0 deg. phase for middle 

Question of using current balun. When I add W2DU balun, that changes the 
resonant frequency of the quarter wave line from 28.3 to 21 MHz. Obviously 
balun adds some coax length. If I shorten the coax to compensate for balun 
length, looks like I will be short of coax, plus balun is 50 ohms. 
Question on the balun, if I take the feedlines exactly perpendicular to 
dipoles, do I need to use balun? What about impedance? (Checking impedance of 
W2DU balun at 28 MHz, I get around 50 ohm and X = 20.) Looks like balun is 
complicating things.  Would it be better to use bazooka balun?

I designed beautiful 4 square on EZNEC and optimized on AO, on paper works 
fine, I get some 30 dB F/B, F/S but when erected it in the backyard, thing is 
goofy. I wouldn't put full contest gallon to it.

I also tried triangular array using WX0B stackmatch, not too happy with it 
either, not much pattern. Looks like back to my trusty Quads and Razors.

Has anyone fooled around with 4 squares that high in frequency, any practical 
experiences? At VE1ZZ using 80m unit, there were no problems, but this "VHF" 
situation seems to be finicky.

Yuri, K3BU
trying to be NT1E in WPX SSB on 10m

FAQ on WWW:     
Administrative requests:

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