Towertalk
[Top] [All Lists]

Re: [TowerTalk] Bandpass Filters

To: towertalk@contesting.com
Subject: Re: [TowerTalk] Bandpass Filters
From: <donovanf@starpower.net>
Date: Thu, 10 Jan 2008 07:30:38 -0500 (EST)
List-post: <mailto:towertalk@contesting.com>
The W3NQN filters are excellent.  I use them between my transceivers and all of 
my amplifiers and they do an excellent job of eliminating out of band spurious 
signals and noise from the trnasmitted signal.

However...

Traditional laboratory filter measurements are extremely misleading with 
respect to the actual performance of filters when they are installed in our 
stations.

Why?  Because the out-of-band impedance of our transmitters (on both transmit 
and receive), amplifiers, and antennas are often wildly different than 50 ohms. 
 There are a few exceptions; for example, wideband amplifiers (not multiband 
amplifiers), triband antennas and log periodic antennas.

We often see spectacular results if we terminate filters in broadband 50 ohm 
impedances in our laboratory measurements, especially if the filters were 
designed for broadband 50 ohm terminations.  

The spectacular laboratory measured filter performance is always degraded when 
we install these filters in our stations, where they are
significantly misterminated in almost every case.  Some filter designs
are extremely sensitive to mistermination, while others (for example, the W3LPL 
receive filter) were specifically designed to perform
excellently when misterminated.

Its not hard to evaluate filter performance in your station.  On receive, find 
a very strong signal with the filter switched off, then insert the filter and 
observe the results.   For example; if you're evaluating a 40 meter filter, 
signals on the 40 meter band should not be noticably attenuated, while strong 
signals on every other band should be driven into the noise.

Its a little more complicated to evaluate transmit performance.  One good 
approach is to use a cooperating nearby ham the listen for your
transmitter harmonics with the filter removed, then measure the 
amount of reduced harmonic signal level after the filter is installed.  

Don't be surprised to find disappointing results from these tests, especially 
the transmit tests.  Sometimes the results can be improved by lengthening the 
cable between the filter and the transmitter.

73
Frank
W3LPL


---- Original message ----
>Date: Wed, 9 Jan 2008 19:43:06 -0600
>From: "Ethan Miller K8GU" <ethan@k8gu.com>  
>Subject: Re: [TowerTalk] Bandpass Filters  
>To: towertalk@contesting.com
>
>Alfredo,
>
>Yes, the W3NQN filters are the best.  But, you will need to design
>sufficient radio-to-radio isolation into your switching system in
>order take full advantage of them.
>
>The I.C.E. and Dunestar filters are NOT the same.  The Dunestar
>filters I've measured with a network analyzer are considerably better
>than their I.C.E. counterparts.  For instance, a Dunestar 600 on 40
>meters attenuates the 14-Mhz harmonic about 55-60 dB, whereas the
>I.C.E. is in the 35-40 dB range.  For reference, my homebrew W3NQN
>40-meter filter pushed the harmonic deep into the noise floor of the
>network analyzer (in this configuration, >90 dB).
>
>73,
>
>--Ethan, K8GU/9.
>http://www.k8gu.com/
>_______________________________________________
>
>
>
>_______________________________________________
>TowerTalk mailing list
>TowerTalk@contesting.com
>http://lists.contesting.com/mailman/listinfo/towertalk
_______________________________________________



_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk

<Prev in Thread] Current Thread [Next in Thread>