[TowerTalk] Playing with my new VIA Analyst (long)

[Jim Lux]

At 10:50 AM 11/11/2005, Robert Chudek wrote:
>Merry Christmas to myself, early. I'm like a kid with a new toy using my 
>new antenna analyzer!
>
>"You guys" pointed out I needed a 1/2 wavelength (electrical) run of coax 
>to my 160m inverted vee to get true values on my meter. I took this as an 
>excellent educational opportunity. So I bought a 500' roll of cheap RG-8x 
>coax that I didn't mind hacking up into pieces. (btw, cheap = $60 for the 
>roll).
>
>I created an Excel spreadsheet which shows the steps I took to tune my 1/2 
>wavelength feedline to 1830 KHz. I posted this spreadsheet to 
>http://chudek.aberon.net in the "AEA VIA Analyst 160m Charts" folder. You 
>can download it from there, if you are interested.


Excellent work!
So what else did I learn? Well, I swept the feedline up to 54 MHz (the 
limit of this model). I saw additional resonances or minimum VSWR's all 
along the spectrum. I zoomed in to analyze the frequencies of interest on 
other ham bands. I thought a 1830 KHz resonant feedline would be useful on 
other harmonically related ham bands. For instance, I expected to see 3660 
KHz as a resonant point. In reality, I measured a dip at 3718 KHz. My 
spreadsheet contains information about all resonances that are within (or 
close) to the other amateur bands.

>Another thing I discovered was the higher in frequency I went, the 
>"shallower" the dip. For instance, on 10 meters the minimum VSWR only 
>dropped to 1.30:1. I speculate this is due to cable attenuation or loss. 
>But I do not know how to turn this number into a meaningful value, i.e. X 
>dB per 100 feet at X MHz. Anyone?

There is an equation for this (improvement in VSWR with attenuation).  I 
don't remember it off hand (I work in dB return loss usually, where it's 
trivial).


>And I am puzzled by the original observation, why does the minimum VSWR 
>frequencies "stretch out" and not hit on exact multiples of the 
>fundamental frequency?

I'll speculate a bit.. your open end would ideally provide a perfect 
reflection. However, it's very hard to make a precision open, because of 
the fringing fields at the end (why calibrating a network analyzer with 
Thru, Short, and Load is popular).  You may have, in effect, a transmission 
line connected to a reactive load (which itself varies with frequency).