[CQ-Contest] Re: ON4UN and the Ten-Tec Orion

[Richard Ferch]

I think the reason ON4UN could get flawless copy with a 150Hz filter setting
is explained elsewhere in his post, where he mentions the less than
spectacular shape factor. As I recall, the 6 dB b/w was 150 Hz, but the 60
dB b/w was 450 Hz.

So, with the mark and space signals 170 Hz apart, they were both outside the
6 dB b/w but the filter slope is such that they would only be attenuated by
around 8 dB. In other words, using a narrow but soft filter like this you
simply subtract an unnecessary 8 dB or so off the signal strength of both
the mark and space signals.

This is an oversimplified analysis. The mark and space signals are not
single-frequency signals, they have their own sidebands due to the 45-baud
modulation, and this means you have to look at the slope of the filter
skirts.

The mark signal, for example, occupies a frequency range from around 40 Hz
to 130 Hz above the center frequency (or is it below? doesn't matter, just
swap mark and space and the same reasoning applies). The inner sideband
would hardly be attenuated at all, but the outer sideband would be
attenuated by as much as 25 dB or so. The waveform would look quite
distorted (or "mangled", as Brett said). Nothing the demodulator can't cope
with on a good strong signal, but as N4HY said, don't do this with weak
signals and expect to get good results.

In fact, it looks from John's post as if the 250 Hz filter setting would
avoid most of the attenuation and distortion introduced by the narrower
setting, but would have pretty much the same effect on adjacent-signal QRM.
The 60 dB b/w of the filter is almost unchanged between the 150 Hz and 250
Hz settings.

This brings up an interesting question: Which is "better", a narrow but soft
filter (b/w 150 Hz, shape factor 3) or a wider but sharper filter (b/w 300
Hz, shape factor 1.5)? I'm no expert, but for the sake of argument I'd guess
that for CW, the narrower softer filter would do a better job on close-in
QRM, while for RTTY the copy on weak signals would be better using the wider
filter with steeper skirts.

73,
Rich VE3IAY

PS: Brett, thanks for the QSL card that showed up in my mailbox recently!