[CQ-Contest] Stubs - 40 & 15 meters

Tom W8JI w8ji at w8ji.com
Sat Jul 21 04:46:47 PDT 2012

> The Elsie coax is "ideal" and as such it will show really huge attenuation
> notches.  Real stubs are going to be in the low to mid 20's.  But 
> otherwise,
> the plots from Elsie looks similar to what the VNA measures for the real
> thing.

Real stubs in real systems can have over 15 dB variation in attenuation, 
depending on electrical spacing from the antenna, and a lesser extent 
spacing from the amplifier or radio (which usually is different on receive 
and transmit). All this is dependent on source and load characteristics at 
the reject frequencies.

Smaller L/C filters can have a similar large performance variation.

This isn't much different than the insertion of baluns in lines, which are 
dependent on common mode characteristics and do not always behave like we 
**think**. For example, an air-core balun can actually increase common mode 

This is one reason why some systems, in totality, work much better or worse 
than others.

But a 30 degree shorted stub parallel with a 60 degree open stub forms a 
parallel resonant low-Q circuit (low-Q because stubs are somewhat lossy), so 
it passes signals at the 90 degree length point just as it does at the at 
the 3rd overtone of the stubs. For example the Q of a 60 degree open stub 
(forming a capacitor) is well under 100, as is the Q of an inductance formed 
by a 30 degree shorted stub in typical sized transmitting cables like RG8 

Stubs are great as a little band-aid for a specific problem, just like air 
core baluns can work in specific limited applications. But for filters, a 
little coil of wire (with a Q in the 300's or higher) and a little mica 
(with a Q in the thousands) might be a better alternative, and only pass the 
band we want.

73 Tom 

More information about the CQ-Contest mailing list