[TowerTalk] Stubs and rx noise

[Tom Rauch]

> I will have a look. My second rig is an Icom 736 and the
main rig is an Icom
> 746 Pro.
> Sylvan

Sylvan,

First, you don't have the best radios for handling this
application. Both are not really very good for "noise".
That means you not only have to trap the composite
transmitter noise, you also have to trap the receiver for
fundamental level from the transmitting rig.

Second, single stubs aren't normally as good as you measure
if S units are actually 6dB. What throws most people is they
think an S meter is 6dB per S-unit, but the can range from
less than 1dB per S unit up to a typical high of 5dB per S
unit. Some are 6dB, but most are not and are also very
non-linear at the lower end and highest end of the S meter.
My IC751A is less than 1dB per S unit near S1, and increases
to about 5dB per S unit at S8.

Third, systems with loss are not necessarily bilateral, and
simple traps or filters can have extreme attenuation changes
as loads or sources are changed. This is especially true
with single section traps or filters that merely add a shunt
or series reactance or resistance to a complex system.
Depending on source impedance characteristics and load
impedance characteristics the passband and attenuation can
be all over the map. A shunting PASS stub to a receiver
presents some R and  J ohms at the unwanted transmitter
frequency. It is  possible to have a complex impedance from
the receiver or the antenna that actually "likes" the stub
to be there! It might not change signal levels in operation
like it does in a test with other equipment, as a matter of
fact the stub *won't* have the same effect with a receiver
as it would if you substitute a 50 ohm load. It won't even
have the same attenuation if you do something as simple as
varying cable lengths in the system! If you   change bands
to look at the transmitter, and the complex input impedance
of the receiver changes, the attenuation of the stub will be
different.

Say you have a stub to pass a 20 meter signal, and you have
a 15 meter transmitter you are nulling. The attenuation on
20 meters WON'T be the same with the receiver operating on
20 as it would if you switched the receiver to 15 to measure
the attenuation of the stub if the complex impedance of the
receiver changes. I hope you get this point, the attenuation
you think you have is almost certainly not be what you
actually have.

Fourth, anything arcing or with an erratic connection in or
near the transmitting OR receiving system can generate
actual broadband noise that is radiated on the receiving
frequency. It could be a fence, a gutter, a pressure
connection in an antenna, or anything else after the TX
filter and ahead of the RX filter, and the filters would
make no change. In other words you may have a "noise
generator"  that converts the desired TX RF into noise that
is between the "filters", rather than between the TX-RX
equipment and stubs.

At my house, I can hear a wire scraping a fence 500 feet
away as a "scratching noise" on 40 meters when I transmit on
160. I think you get the idea.

73 Tom




Check out ARRL test reviews and