Amps
[Top] [All Lists]

Re: [Amps] W3NQN bandpass filters, question.

To: amps@contesting.com
Subject: Re: [Amps] W3NQN bandpass filters, question.
From: Jim Brown <jim@audiosystemsgroup.com>
Reply-to: jim@audiosystemsgroup.com
Date: Mon, 06 Dec 2010 12:43:26 -0800
List-post: <amps@contesting.com">mailto:amps@contesting.com>
On 11/27/2010 9:56 PM, Larry Benko wrote:
>   I have a
> plot that I can send you of the measured differences but unfortunately I
> don't have both units to compare the schematics.

It's important to remember that a filter is a passive network, and like 
ALL passive networks, its behavior depends on the source impedance and 
load impedance AT THE FREQUENCY OF INTEREST.  Bench measurements are 
nearly always made with 50 ohm source and load Z, which is an Alice In 
Wonderland view of reality.  Few amplifiers have a 50 ohm output 
impedance, and few antennas look anything like 50 ohms resistive at 
harmonics of their intended operating frequency (unless they're a 
multiband antenna, of course).  A filter that provides 40dB of 
attenuation with those bench conditions might provide only a few dB in a 
real antenna circuit (if, for example, the load impedance was much 
higher than 50 ohms, or highly inductive). The same would be true of a 
power amp with a tuned input.

A far more meaningful measurement, albeit a far more difficult one, is 
one that measures the transfer function of the filter IN THE ACTUAL 
ANTENNA circuit -- that is, with one probe bridging the input side and 
another bridging the output.  Another useful measurement would be an 
insertion loss measurement -- that is, measure the out of band content 
with and without the filter in place.

On receive, the problem is somewhat different -- input circuits tend to 
be in the 50-100 ohm range, and can be FORCED closer to that range by 
adding a simple resistive pad in series with the RX input (something 
like 10dB).

73, Jim K9YC
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps

<Prev in Thread] Current Thread [Next in Thread>
  • Re: [Amps] W3NQN bandpass filters, question., Jim Brown <=