[TowerTalk] Stacked Skip Log and Normal LP's and the Miligned L Network

[K7GCO@aol.com]

In a message dated 12/6/00 8:22:32 PM Pacific Standard Time, 
wx0b@arraysolutions.com writes:
<< 
 Thats not really a good idea Ken,  I would point out that an L network
 has a very small bandwidth. I suggest to do it a different way.  Use a
 matching nework that is broadbanded. Like the StackMatch wideband UN-UN
 transformer.
 
 Then the logs will have the proper impedance matching over their entire
 bandwidth.
 
 So this the technique.
 
 Run equal lengths of 50 ohm coax to a central point between the logs
 place your Stackmatch here, it will match the logs properly to select
 U/L/Both. Then run a line of hardline down to the shack. It can be CATV
 to save more money and have less loss.  Match it with a couple of 50:75
 ohm UN-UN Xfrms.
 
 L networks are work at one frequency.  And I know that K7GCO knows this,
 he probably just was in a hurry.
 
 Jay, WX0B
  >>
Jay: I will take some exception to your views based on a lot of use of the L 
newtwork (and I was in a hurry). I have a great sytem for stacked beams that 
you would even like but didn't have time to describe it.  Without attachments 
even a simple circuit is hard to describe without a bunch of E-mails to 
follow.  I will have a Web Site In SD as the SD Antenna Club with my old call 
W Zero Lice, Mice & Bedbugs for material like this and some of the mags like 
AntenneX which every TTer should read.  Cebik had a great LP with F/B like I 
have never seen and will build.

First off it's assumed the Skip Log or Normal LP's that are horizontally 
spaced (in this example) wide enough to minimize the affect each has on the 
other when used indiviually and the different Z's pesented at the feedpoints 
due to the difference in heights off the ground.  For this installation one 
would have to live with the vertical pattern affects and Z's on each other 
individually when only one or the other is used and the different Z affects 
on each one from the ground.  A compromise spacing was suggested not for 15M 
but 17M just to minimize this coupling problem.  I've never read anyone 
warning of the affect one stacked beam has on the other when used 
individually as I can do with my switching harness.  I saw the interference 
in EZnec.  Most use both together full time which is a serious mistake.  The 
spacing I suggsted is most likely too wide on 10M.  On 10M it sharpens the 
free space vertical lobe even more but doesn't increase the gain and 2 ears 
start to form.  On 20M the spacing is too close and the full 3 db gain is not 
obtained but it's less of a compromise than if they were optimally spaced for 
15M.

The pattern distortion when either one is used alone (to make use of the 
different angles of radiation between the 2 beams) is a problem that I treid 
to minimize.  Fortunately this problem is mimimized the higher the gain of 
the beams used in stacks for 2 reasons.  The vertical patterns are 
progressively sharper with gain which means less coupling between them 
vertically when used separately and the spacing also has to be wider to 
obtain the max gain with stacked beams--both +'s.  The LP's would couple to 
each other more as their vertical patterns are much broader (why they work so 
well close to the ground), and a closer electrical spacing is required when 
stacked.  So the assumption is made that with equal length feedlines the Z at 
the end of each feedline will be very close to each other and you will have 
to live with it.  I also use my "Magic Length" of  91' 2" (.66 VF) or 
multiples as it will be a 1/2 wave or multiples at 3.562, 7.125, 10.67, 
14.25, 17.81, 21.375, 24.937& 28.5 MHz.  Resonant coax lengths reduce 
reactances and repeats the R value of the Z at the feedpoint.  If longer 
lengths are needed, you can add lengths of 45' 6.9" and the 1/2 wave 
relationships will still hold. 

I have a very simple matching harness that goes on a coax switch for this 
application that allows full power to each beam indiviually or both together 
using 2-75 ohm 1/4 wave stubs with a unique switching feature.  It has a big 
bonus of giving addtional bandwidth when both are used due to the opposite 
reactance the 75 ohm 1/4 wave stubs create above and below the center 
frequency.  I've got a lot of rave notices of this sytem (at least worth a 
dinner) from many on it including DXer Lou Gordon K4VX.  I didn't mention it 
as the 1/4 wave stubs would have to be changed for each band.  

Now for those who suggested the typical V stacking where both feed points are 
tied together for simplicity and 1 feedline, this is a bad application for 
this concept when using short boom Skip Logs.  It's the most usefull and 
effective for very long boom LP's.  This physical configuration essentially 
maintaines the same and optimum electrical spacing for the entire frequnecy 
range and has a minimum V angle.  With short boom LP's, the beams would have 
to be up to 90 degrees or more to each other to optimize and balance the 
electrical spacing on the upper and lower bands.  Furthermore the part of the 
vertical pattern pointing forward from each Skip Log would be down several dB 
and pattern addition straight forward is not at all optimum--forget it.  
Note! When 2 long wires are used in a Vee Beam configuration, they are 
aligned so that each has a lobe going forward and parallel and only 3 db is 
obtain in an optimum configuration.  I have such a V log here for the TV 
Channels of minimum elements using also a staggered element elimination 
design and the Vee stacking didn't do a thing for it.  Another compromize 
failure.  We need more of those who maximize their designs (I've been known 
to lean that way), not this damn "Minimal Design Bankruptcy" constantly 
jammed down our throat that wastes everyone's money.  

My L networks seem to be a bit broader than yours or I didn't think they were 
narrow on applications just like this 25 proposed ohm load.  I also have what 
I call "One Knob Antenna Tuners" in 4 switchable L network configurations, 
shunt and series capacitance and inductance. It does lots of things.  
Visualize this.  The variable Xc (I use them a lot) is on the same shaft as 
on the inductor.  As I rotate the inductor through each turn, the Xc goes 
through 2 complete cycles and in about 5 seconds, I can tune through the 
entire range of many many Z's it will match--in 4 different configurations.  
So I find 1:1 SWR real fast and can touch it up quickly off frequency if 
needed.  

I've used this L Network configuration in a similar application.  Visualize 2 
yagi's on the same boom at right angles to each other in the "X 
Configuration".  Each yagi is fed with equal length feedlines.  CBer's had a 
similar beam in a + configuration where they selected either horizontal or 
vertical polarization.  The vertically polarized DE had to be a different 
length and had other adjustments than the horizontal DE used due to the heavy 
coupling to the mast.  The tower was excited and lift off of the vertaical 
pattern was substantial.  I used the X configuration to reduce this coupling 
and to balance it up so the load of each beam for the coax was the same.  It 
still coupled but I have a way to get rid of the tower interference now when 
vertically polarized.  I'm adapting it to a quad DE--when I get time.   

I have a switch box that allows me to select 6 different polariztions from 
this "X Beam."  I can select either beam for 45 degree polarization left or 
right (direct feed), horizontal polarization (1/2 power in each beam in 
phase), vertical polarization (1/2 power to each beam 180 degrees out of 
phase) and turnstile polarization CW or CCW with 1/2 power in each 90 or -90 
degrees out of phase.  To obtain 90 or 180 degree phasing I switch in these 
lengths of coaxes in series with one coax in the proper way.  For matching 
in-phase, the coaxes are paralleled for 25 ohms and matched with (you guessed 
it) an L Network.  It uses relatively small coils and variable Xc's for 1 KW 
levels without heating or arcing in a small space--3 of them.  Once the 
proper inductor of the L Network was found and tuned up in the middle of the 
band I seldom had to adjust the Xc in the band edges.  I had all the 
bandwidth I needed with the matching and phasing hardware in a small box and 
saw no need to look for a better system.  I could have used 2-75 ohm stubs 
for the matching part but didn't for reasons I don't remember--some 50 years 
later.  A CB mfg in Chicago tried to swindle the design out of me--but 
failed.  I seems that most mfgs want everything I got free despite contracts. 
 He could have had a great money maker for us both laid in his hands but out 
right greed cost him (and me) dearly.  He could have had several other unique 
designs also.  Industry is notorious for this parctice but there are 
exceptions.

Anyhow horizontally stacked Skip or Norma Logs is a similar Z matching design 
and by having a calibrated variable inductor and Xc you can preset for each 
band or inbetween frequency very quickly instead of changing 1/4 wave stubs 
for each band or frequency for a match.  I'm anxious to try my "L-Network 
Match Box" on horizontally stacked LP's myself.  I see no reason to use 
another matching system at this time but would like to see a better system 
for flexibilty.  Send me the details of the Stack Match and I'll see if I can 
adapt it.  I have the flexibilty of selecting the top, bottom or both LP's in 
the shack--without relays.  I like to maximize the full potential of any 
system.  If I can get more bandwidth I'm interested.  I think I'll have 
plenty of bandwidth as with the LP's I won't have the reactance swings 
typical with parasitic beams.  Now I bet you knew that. Your concern of L 
Network use has been deminished by LP's.  Long live the L-Network until 
something else comes along that's better for this application.  

Although I was in a hurry I arrived at my suggested simple desgin I posted on 
TT in about 5 seconds considering all these details I've explained here as 
I've "been there and done that" in similar applications.  It took about 10 
minutes to Post--but many hours to answer.  I've found that if I give too 
much info or too little (and without attachments) I have to justify or supply 
even more info--which I always do.  I would hope that you TTer's would by now 
have a little more confidence in my suggstions.  If I say it works--it works. 
 I'd like to remind you that you also got the info free of charge.  There are 
some with linited insights that make "great demands" for more and more free 
info as they can't derive it themselves. They are a form of "Technical 
Parasites" on "Technical Welfare."   I normally charge for technical 
information. (If I charged you for what I said, would you have more 
confidence in it?)   I thought I'd make a quick suggestion, I was properly 
challenged and I may and often learn of new techniques myself which I'm 
always open for.  

Horizontally Stacking LP's is a great idea that I started getting myself 
about 15 years ago with the availability of them from 14 MHz to higher 
frequencies and I knew how to match 2 of them stacked real easy.  About 40 
years ago I tested one of the first LP's of the first designers of it by the 
name of Dewight Isbel came up with.  It was flat from 60 to 150 MHz at 93 
ohms.  I still have a picture somewhere of it and Isbel.  He worked for 
Boeing in a fancy Antenna Lab and I didn't, yet I was testing it for him. 

The lower gain of LP's is no detriment--it's an asset most of the time when 
properly used.  It's been a victom of unifomed critics who have never used 
one. I'll have 2 M2 Skip Logs's stacked in SD (if I ever get there) 
matched--you guessed it--with L Networks or a better sytesm if there is one.  
This should explain why I do what I do (have I ever mis-led you?).  During 
the mean time there should be some serious study of the life and times of the 
"unseen vertical pattern" as you rotate the beam, its size relationship (it's 
wider) to the horizontal pattern, gain and Reflection Factors.  It's time for 
more of you to get on Eznec and read AntenneX mag on internet.  It's read in 
over 170 countries now.  k7gco
                                "Now you know the rest of the L Network 
story." 


--
FAQ on WWW:               http://www.contesting.com/FAQ/towertalk
Submissions:              towertalk@contesting.com
Administrative requests:  towertalk-REQUEST@contesting.com
Problems:                 owner-towertalk@contesting.com