[TowerTalk] Method of calculating phase delay variation when stacking two...

Sun Sep 1 15:44:20 EDT 2013

Eric, I don't see the answer being quite as complicated as all that. 

If one yagi feed point is forward of the second yagi, the number of  
wavelengths or degrees would be corrected by making the forward yagi feed line  
longer by the amount of the offset.

a 5 ft offset would need an increase in line length of 5ft electrically.  
Physical length in this case would be 5ft times the velocity factor. RG213 
has a  v.f. of .66 so the actual length increase is 3.3 ft. LMR400 v.f. is 
about .84 so  the actual length would be 4.25 ft.

No need for calculating degrees or wavelengths is needed.

This is assuming you want to make the two feed points in phase and the vswr 
 is low. Elevated vswr would introduce some phase shift but that is a  
whole new subject.

The spacing you mentioned should work out very well. It is enough to attain 
 near maximum stacking gain on 20m. While more than enough for the higher 
bands,  the only consequence is somewhat higher grating lobes in the vertical 
plane. The  grating lobes are many dB lower in amplitude than the ground 
reflection lobes  and probably not even noticeable.

73,
Gerald K5GW

In a message dated 9/1/2013 1:30:17 P.M. Central Daylight Time,  
k2cb at comcast.net writes:

Good Morning TT, 

I have a pair of SteppIR  yagis that I am attempting to run in phase. A 
DB36 at 91ft, and a 4E at  55ft.   

The main reason for the two antennas is  versatility to hop quickly between 
bands, or point in two directions at any  given time. Having the ability to 
phase them together is and added bonus.... I  understand that the stacking 
distance may not be optimal, but it "is what it  is" , dictated by the 
tower.    

I originally  ran equal lengths of cable between the feed point of each 
antenna and a  two port stack match box at the base of the tower, and ran with 
it. But I  recently did some more reading and came across a Powerpoint doc 
that WX0B  presented some years ago:  

http://www.pvrc.org/Powerpoint/wx0b_files/WX0B%20Stack%20Presentation%202000
.ppt  

In the presentation, he states that if there is a difference  in the driven 
element locations, you need to account for this and add a "delay  line" of 
sorts to the leading antenna.  So based on his document, I  measured a 5ft 
difference in the position of the DB36 and 4E driven  elements.  I ran the 
following calculation: 

Variation of  driven elements / wavelength x 360 deg.  = degrees of delay 
needed, s o  in my case the variation in driven element distance from the 
center of the  tower results in phase shifts as follows : 

10m - 5/32.96 x 360 =  54.6deg 

15m - 5/43.94 x 360 = 41.0deg 

20m - 5/65.92 x 360 =  27.3deg 

So based on the document, I  inserted a delay  stub of coax, as follows: 

5ft x VF(0.87 for LMR400) = 4.35ft  

I then also took my MFJ269 analzer (for what it is worth) and  measured the 
electrical length (not the physical length) of both cables using  the VF 
value of 1.00 per the manual , and they measured out with the 5ft  difference, 
with the DB36 measuring 5ft longer electrically. So I thought I  had it 
accounted for properly...... 

Fast forward to  this morning, and I now ran across the following page by 
WN9O,  

http://www.qsl.net/wn9o/Phaspat.htm#Figure 4    

in which he describes stacking different Force12  tribanders.   His 
calculation method for his delay line lengths  seems to vary greatly from what I 
see in the WX0B presentation  document.  Whereas the WX0B document simply has 
you account for the  spacing variable of the driven element for both yagis, 
and convert  the physical distance to the equivalent coaxial electrical 
difference, and  insert this length into the leading antenna feedline, the WN9O  
document seems to rely on detailed modeling to first determine the  phasing 
variation in degrees, then use the following formula to  calculate the 
required coaxial delay line: 

Coax length =  Wavelength  / 360   *   degree value from modeling calcs *  
VF of coax 

In his case, for 10m (since 10m is worse of all bands)  = (984/28.4) / 360  
* 114deg  * 0.78 = 8.55ft 

I  noticed that WN9O used 984/F for the wavelength calc, whereas the WX0B 
doc  used 936/F based on his 21MHz system.  I tried to apply the WX0B formula 
 against the numbers used in WN9O's antenna system to see if I could come 
up  with the same 8.55ft for the delay line , but I came up with a value  of 
3.5ft based on his 4.5ft of difference between driven elements and a  0.78VF 
coax. And a phase delay of 49.15 degrees.  See the WX0B link, page  36 for 
reference. 

So my question at this moment is -  which is the proper method for me to 
calculate the proper delay line length to  properly phase the two dissimilar 
SteppIR yagis? 

Use the  simple WX0B method of just measuring the variable between driven 
elements in  the vertical plane, and multiply that by the velocity factor of 
the coax I am  using, or do I need to have the two antennas modeled and then 
u se the WN9O  calculation method?     

Unfortunately, the  NEC and modelling software is somewhat above my 
knowledge level at this time.  I have tried to play with Ez -Nec a few times , but 
I just could not get my  hands around it.   Please don't bash me for not 
trying...  I  have too many other irons in the fire to attempt to master 
modeling software  at this stage in life.  And my skills in other areas of the 
hobby more  than make up for my lack of modelling skills. I can troubleshoot 
and  repair  radios and  amplifiers to component level , just don't ask  me to 
model an antenna! 

If the latter is the case, would  anyone care to offer some assistance and 
model the DB36 vs. the 4E in my case,  and enlighten me on how much of a 
delay line length I actually should be using  (electrically)? 

Looking for comments or  suggestions.... 

Thanks 

Eric 

K2CB  

_______________________________________________

_______________________________________________
TowerTalk  mailing  list
TowerTalk at contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk