Tom, et al:
In terms of group delay, there is probably a tradeoff
between shape factor and delay. Whereas the 60dB
points in a filter don't contribute significantly to noise
bandwidth, they have a great deal of bearing on group
delay. Could it be that between two filters with the
same noise bandwidth, that the looser skirt filter is
actually easier on the ears in a situation like that of
Earl's test where their are no strong signal present?
Also, I never really thought about before, but does
anyone (like INRAD for instance) attempt to delay
equalize their IF filters?
It would interesting to find out if Earl tried the running
the Ten-Tec Orion with the narrow analog filters
switched out while at the same time cranking down the
bandwidth of the DSP filters. DSP filters generally
have linear phase (e.g. flat group delay), so it would
be interesting to compare the receiver "sound" when
running with tight analog filtering vs. tight DSP filtering
in a weak signal environment like the one Earl
described.
Mike, W4EF....................
>
> How many filters have you seen rated for group delay errors, or how often
is
> it mentioned? I'd bet almost no one looks at it, even though it can be
> critical. My 751A totally blows away a stock R4C or 75A4 for selectivity
and
> overload by close signals and looks much better for noise floor, but the
> 250Hz filters are unusable for working noise floor signals if there is any
> rough noise present. The 500Hz filters in the 751A are barely acceptable.
> Because of that, I never use the 751A for working weak signals. All the
> sensitivity in the world does no good when the filters ping on every noise
> pulse, and CW sounds mushy.
>
|