[TowerTalk] 1/4-1/2 wave nonsense

[Tom Rauch]

I'll try to "be good" and  bypass all the personal attacks needless 
included, but I do think I should respond technical content in 
K7GCO's lengthy post:

> that reduce these impedance variables and math.  If a long 75 ohm hardline
> is used on 10M with Phone/CW wide excursions of frequency connected to a
> 50 ohm load, expecting a 75 ohm feedline of 1/2 wave multiples to bail you
> out and reduce the over all SWR, resistive and reactance variances in
> particular when connected to a 50 ohm output final--is just not good
> thinking.  You can get a good match at one frequency where it's a 1/2 wave
> multiple and 50 ohms at the antenna.  

Using multiple 1/2 wl sections of a mismatched line will reduce the 
useful bandwidth of an antenna. I don't think anyone argues against 
that. The more 1/2 wl sections and the greater the mismatch, the 
more restricted bandwidth becomes.

As Wes correctly pointed out, the impedance gradually moves 
closer to line Zo because of loss.   

But the fact remains, that system works.

> 50 ohm finals reduce power out with loads other than 50 ohms and then at

Everyone agrees with that, so far as I know. I doubt anyone will 
want to modify their rig to normalize it at 75 ohms output. There are 
much more useful and simple solutions, such as transformers, 
stubs, or Q sections. Using a multiple of 1/2 wl is the most simple 
solution, and only has the drawback of a reduction in VSWR 
bandwidth. 

It seems contradictory that we were told to use an open wire line in 
series with a capacitor to feed a conventional 80 meter dipole to 
operate the whole 80 meter band (a system with a 9:1 mismatch 
that reduces bandwidth)..........and then at the next turn the same 
person ridiculed several of us (and me in particular) for using a 
multiple of 1/2 wl that only has a 1.5:1 mismatch responsible for 
reducing bandwidth!  
 
I wrote the following about another system, where ladder line is 
used to center feed an antenna:

>     If feedline length isn't "planned", it's quite possible to have an
>     impedance that is nearly impossible to match on some bands. Lengths to
>     especially "avoid" are odd-quarter waves on bands where the antenna is
>     1/2 wl long.

To which K7GCO replies:

>  ****** Yes and No.  End fed or center fed??  A very poor "unplanned"
> example--use specifics!  A 1/4 wave of open wire line will give a Hi-Z
> load at the end of the feedline from the center Lo-Z of a 1/2 wave. 
> Without resonant lengths of either or both you can indeed have reactive
> loads that can be difficult to match. 

"Reactive loads" aren't the problem. Reactive loads can be very 
easy to match. The problem, as I pointed out, are extreme values 
of impedance when open wire lines 1/4 wl long are used to feed a 
1/2 wl dipole...and that antenna is used booth on the 1/2 wl 
frequency and harmonics.

As I pointed out, that combination results in impedances well over 
the impedance limit of most tuners.    

> avoids that totally.  It's a "resistive load" that is easy to match when
> both the antenna and the feedline are resonant even reasonably close and
> that is why either 1/4 or 1/2 wavelengths are suggested by all
> authorities--to reduce or eliminate reactance.

That suggestion is in error. It isn't the reactance that is generally 
the problem. The problem is matching an impedance of several 
thousand ohms on multiple bands....even if the reactance is zero. 

> "Follow Instructions".  NOTE! I should like to point out that the Johnson
> Match Box will handle higher reactive loads than most tuners due to the
> use of a dual differential variable capacitor--it allows balanced ir
> unbalanced operation also.. Learn how to use it.  That's specifically why
> it was developed.

That is incorrect. The differential capacitor actually restricts 
matching range. If it is replaced with a conventional capacitor, 
matching range of the Matchbox increases. I'll post a message 
from Walt Maxwell on this topic, that he asked me to post. I was 
waiting for it to come up again, so this is a good time. 
 
 I wrote:
>  For example:
>      With a 450 ohm line and a 50 ohm (dipole) load, using a 1/4 wl line 
>       (lossless case) input impedance will be over 4000 ohms.

K7GCO relies: 
>  *****The match above is close but shy by 50 ohms--no big deal.  

That isn't correct. The line input impedance will be nowhere near 50 
ohms.

A 1/4 wl 450 ohm line feeding a 1/2 wl dipole will be around 4050 
ohms at the input. While many T networks will handle, the baluns 
won't "like" that impedance. The Johnson Matchbox won't handle 
that impedance on many bands!

I wrote:
>      On the second harmonic input impedance of the system feeding the
>      above 
> dipole,  
>      impedance could be as high as several thousand ohms.  

K7GCO relied:
 >  *******In order for the 450 ohm line to invert the antenna impedance to
>  say 
> 7000 ohms, the antenna R has to be 28.929 ohms to be exact using the 1/4
> wave stub formula.  

Again, that is clearly incorrect. At the second harmonic, the line is 
1/2 wl long. The dipole is two half-waves in-phase.

The end-impedance of a single half-wave with normal wire 
diameters is several thousand ohms. Feeding two half-waves 
results in twice that impedance at the antenna end of the feedline. 
1/2 WL away, that impedance is repeated (with a slight error from 
line losses). Now the tuner would have to match several thousand 
ohms.    

>      Very few tuners, including the old Matchbox, will work with such a
>      load.
> 
> ******* "Absolutely Nothing" could be further from the truth.  The MB was
> designed especially for these antennas and--Johnson didn't screw up.

That was already addressed. The most "restricted" matching range 
tuner I have are my KW Matchboxes. And yes, Johnson did "screw 
up". I'll let Walt Maxwell's post speak to that.  

Rest of post deleted, since it basically repeats itself.


73, Tom W8JI
w8ji@contesting.com

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