[Amps] IMD

[Fuqua, Bill L]

    The human ear does not hear waveform or phase shifts. In fact it a bunch of overlapping narrow bandpass filters that only respond to 
amplitudes at their resonant frequencies.  Phase information, as far as the human ear is concerned, is not important.
However, in the case of data or video it is. 
   Just a piece of coax cable produces different phase shifts for different frequencies even if it is a perfect delay. That is because
for  a fixed delay for all fequencies there will be a phase shift for each frequency proportional to frequency. That is why an ideal
filter has a rectangular frequency response and a linear phase response ( constant delay).
73
Bill wa4lav

________________________________________
From: amps-bounces at contesting.com [amps-bounces at contesting.com] On Behalf Of Jim Brown [jim at audiosystemsgroup.com]
Sent: Tuesday, January 10, 2012 1:02 PM
To: amps at contesting.com
Subject: Re: [Amps] IMD

On 1/8/2012 6:35 AM, Ron Youvan wrote:
>   This "different delay for different frequencies" is called "group delay."
> It is something that messes up dial up modems.  (so you don't get 56 kB/s)

In the world of pro audio, where I made my living as a consultant for 30
years before retiring several years ago, we had to go back and study the
fundamentals of circuit theory that says that any variation in the
amplitude response is accompanied by a corresponding change in the PHASE
response.  ANY filter produces phase shift in the part of the spectrum
where its amplitude response is changing -- that is, its skirts, AND in
the passband NEAR its skirts. The greater the number of poles (circuit
elements) the greater the phase shift.  Years ago I used a K2, mostly
for CW, but occasionally on SSB. It was a decent SSB RX if you used the
TX filter, which it did for widest RX bandwidth, but it was an AWFUL RX
for narrower bandwidths, which were accomplished by stagger-tuning the
crystal filters. The response looked like the Rocky Mountains, and the
resulting phase distortion made signals quite difficult to copy.

This phase shift is also present in filters produced by digital signal
processors DSP) --  DSP filtering is simply a simulation in the digital
domain of analog filters, and the phase shift of an analog filter in the
digital domain will be the same as that filter in the analog domain. One
advantage of filtering in DSP is that if you have enough DSP, you can
use more complex filter topologies that have less phase shift, or less
problematic phase shift.  The best DSP rigs can produce excellent
response, and sound very good. The K3 and Orions are in that category.

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