[Elecraft] K2 Filter Shape Factors

[Bill Tippett]

Hi Don,

At 04:54 PM 4/10/04, you wrote:

>It is difficult to set the filters properly using a single signal input to
>the K2 - it is much easier if you use wideband noise.

         Not for me...especially since I don't have a noise generator or
Spectrogram. :-)) Assuming the K2's filters are generally symmetrical,
it is very easy to adjust for center without any equipment other than
a stable signal source.  I first zero beat using SPOT, note that frequency,
then tune to either side noting the points where the S9>S7 bar becomes
stable at S7 (i.e. no flicker).  Assuming that the K2's level detector for
the S-meter is stable, this should tell me if the filter is centered.  For
example, using the XG1, I zero beat at 7040.02, measure the HF S7 point
as 7040.22 and the LF S7 point as 7039.96.  This tells me that filter is
260 Hz wide and is centered at 7040.09.  Then I set the CAL FIL BFO
so that I equalize displacements from zero beat to S7.  When it is set
correctly, I have 130 Hz above and 130 Hz below zero beat (e.g. S7 is
at 7040.15 and 7039.89 for a zero beat of 7040.02).

>Looking at your BFO settings, it seems you have the BFO on the same side of
>zero for both CW and CWr.  I just checked my settings to give you an idea -
>they are as follows:
>Filter BW   CW   CWr
>1.20           124    166
>0.75           117     163
>0.50           115     161
>0.10           111     160
>
>The above values produce a sidetone of 600 Hz - with your 400 Hz selection,
>your CW numbers will be a bit higher and the CWr values will be a bit lower
>(frequencies closer to the filter center).

         The numbers I quoted were "BW" (bandwidths), not the BFO settings.
My point was to show that the 1.50 filter has a strange peak in the passband
that is actually narrower at S7 than the 700 or 400, yet that filter then 
widens
out down the skirts of the passband.  Has anyone ever posted a plot of the
wide filters?  I'm just curious to know what they look like.

         The main point of my post was to ask what BW setting yields the
best 6/60 dB shape factor.  From KI6WX's plots, it appears that the 400 Hz
is better than 700 Hz, but I wonder what BW setting minimizes shape factor?

                                 73,  Bill  W4ZV