Hi George,
> Losses through one of the popular commercial switching bandpass
> filter assemblies range fro 0.5 dB minimum to 1.0 dB maximum.
> The higher losses are at 160 and 10 meters and the lower losses
> are at 40 and 20 meters, typically. The rejection of unwanted
> frequencies range from 18 to 55 dB.
It that a bandpass filter that rejects all frequencies above and below the
operating frequency?
One dB amounts to about 20% of applied power. At 1500 watts, filter
dissipation would amount to about 150 watts minimum and 300 watts maximum.
The filter better be equipped with a cooling fan, large heat sink and/or
some very large components, or it had better be a low power unit!
What was the power rating of the filter?
Low power bandpass filters I build or measure typically have only tenths of
a dB loss, and high power bandpass filters are down below .05 dB loss. Even
filters with crummy (low Q) components parts used for receiving
applications barely get to the one dB loss area. Something is drastically
wrong somewhere either with the filter or the measurement method!
> Loss introduced by a shorted quarter wave stub made from RG-213
> is 0.06 dB at 14 MHz. Rejection of 28 MHz is 31 dB.
That stub only rejects one frequency, or even multiples of one frequency.
That's a lot less taxing filter requirement than a multi-section bandpass
filter. A single frequency pass-reject L/C filter can have very low loss
and excellent attenuation.
Not only that, attenuation in a single stage (or two stage) filter is
source and load impedance dependent. You might very well "see" 31 dB with a
50 ohm source and load, but there is no guarantee the rig or antenna
supplies 50 ohms of filter source or termination impedance at the undesired
frequency. As a matter of fact, we can be virtually certain the rig and
antenna has an impedance much different than 50 ohms at undesired harmonic.
Even the length of cable between the rig and stub greatly affects
attenuation, with different line lengths optimizing things for different
rigs.
That's a big problem with measuring and rating filter attenuation. 31 dB
might be measured, but it is ONLY 31 dB in the test setup used. The real
world attenuation may be much more... or much less... in the real world
application.
> Loss introduced by an open 1/2 wave stub made from RG-213 is 0.05
> dB at 28 MHz. Rejection of 14 MHz is 34 dB.
The dissipative resistance of 1/2 wl of RG-213 measured at 30 MHz measures
about 3400 ohms.
3400 ohms in parallel with 50 ohms dissipates about 1.5% of the applied
power. That figure makes sense.
> Double stubs will typically more than double the dB rejection.
As will changing line length between the rig and stub!
73 Tom
--
FAQ on WWW: http://www.contesting.com/towertalkfaq.html
Submissions: towertalk@contesting.com
Administrative requests: towertalk-REQUEST@contesting.com
Problems: owner-towertalk@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|