I build last week a Bazooka vertical for the 40m (see link here) with good results.

This inspired me to build a second one and implement a two phased vertical array system.

I have chosen for a Christman phasing system : simple and easy o build

ON8IM Christman 

Now, praticaly ...  simple, yes... but ... 

I'll give you here how I made : probably there are better methods. I'm far to be a specialst !

First : start to build two identical vertical bazooka antennas

  • Central frequency : 7.050 MHz
  • Radials : system of 4 elevated radials 2m above the ground 
  • Coax used to build the antennas : used H100
  • Top extension of the antenna (part C in the illustration below) : 300 Ohm Open Line shortcuted on both ends
  • Measured coax velocity factor (*) : 0.82

(*) Don't trust to much the given velocity factor of the coax cables you have : the wise man will verify it with a antenna analyser (I use a MFJ 259-B), sometimes the results are verry surprising -espacialy with old cables laying in the shack since years, or little oxydated or often, the cheap cables coming from 'exotic manufactures'.

 

I used the formulas given in the page about the vertical bazooka :

on8im vertical 40m

A = (230/F)*0.305 (meters) = (230/7.050)*0.305 = 9.95 m
B = A * Vf = 9.95 * 0.82 = 8.16 m
C = (A - B) = 1.79 m
S = +/- 10 cm (not critical)
R = A * 1.05 = 9.95 * 1.05 = 10.45 m

 

Both antennas where built identical, mounted along fibergalss fishing roots and installed at abt 2 m above the ground
The SWR of both antennas didn't exceeds 1.2:1 on the entire 40m band.

 

The phasing lines ... another storry.... 

Let's apply the 3 steps rule.

Electricaly,:

90° =  0.25 λ
-> 1° = 0.25/90 λ
-> 84° = (0.25/90) X 84 λ

on the same way ,

-> 71° = (0.25/90) X 71 λ

But this way doesn't brings anything to determine the physical lenght of the lines.
If whe want to work with a antenna analyser, whe do have to work with quarter or half wave stubs.
 
So let's find the equivalent frequencies on wich the stubs 84° (at 7.050 Mhz)  and the stub 71° (at 7.050 Mhz) are reacting as quarter wave stubs.
->for a 84° stub  = (7.050/90) X 84 =  6.58 Mhz
->for a 71° stub = (7.050/90) X 71 = 5.56 Mhz

 

Now, in function of the velocity factor of the used coax (here I used RG-58 with a measured coef of 0.65), whe can determine the lenghts :

this wil give for the 84° stub=>  (75/6.58) X 0.65 = 7.29 m
this wil give for the 71° stub=>  (75/5.56) X 0.65 = 8.77 m

 

Now, I cut pieces of coax a little bit longer that mentionned, I mounted a connector on one side, and left the other side open and determine with the MFJ-259B the correct quarter wave lenghts. You should find a dip and a value of X=0 or close to 0.

To be sure :  I checked at conrresponding frequencies where my lenght of coax react as  3 or 5  quarter waves lenghts, I observed that the X =0 values are more prounonced at higher frequencies.

Example  : if your dip is around 6.58 Mhz and indicate a X=1 or 2 , I try at 19.7 Mhz (6.58 X 3) : and I check if X=0 on this frequency, re-adjust if necessary. Repeat those measures several times before cuting in you coax.

Working at higher frequencies allow the work more accuately.

Once the coax pieces ready, I mounted the phasing lines and I used a coaxial relay , I didn't use the K1 relay like illustrated.

 

Last remarqs :

Your best ennemies :  

the "T" SO-239 connectors (especialy the cheap ones comming from "exotic"manufacctures). 
the PL-259 connectors, plse use professional quality (i.e. Amphenol) ,
incorrect assembly method of the connectors.

 

Making short : how more sophisticated, how more it must be assembled 'in the state of the art'. 

Measures :

The measured SWR of  phasing in both directions does not exceed 1.3:1 on the entire band.

 

Practical results :

A strong direction effect and a good F/B is observed.

See this demo video here  

I made contacts with US stations, using only 50W SSB  with succes. Not so bad ....