[Elecraft] Why QRPers should HATE clicky rigs.

[Masleid, Michael A.]

I've been trying to understand the key click issue.
I don't think the simple RC circuit that I've seen
in the ARRL will stop key clicks.  I've done computer
simulations of various wave shaping circuits, and
looked at the resulting clicks through various
receiver filters.  Trouble is, I don't know what is
good enough for close spaced stations.

I'm a new ham, so I don't understand some of the
language.

-- S51 station -- that's a prefix, not an "S" number?

-- cascaded 400 Hz INRAD filters --

This is a really sharp flat filter, so you see anything from
7000.55 to 7000.95 kHz equally well?

Key clicks are at S8 from a 7001.25 kHz station who was really loud.

So, this station's center frequency is 300 Hz away from the edge of
your pass band.  Assuming that a really loud station is at
perhaps S9 + 40 dB when you are tuned to 7001.25 kHz, and
one S unit is 6 dB, Then the key clicks in the INRAD filter
are down 46 dB from the carrier when measured over a frequency
range from 300 Hz to 700 Hz below the carrier center frequency?

With 40 meter background noise fairly low, you would be able to work
perhaps an S1 or S2 signal level?

This would be at least 6 "S" units below S8, or another 36 dB down.

So, to be a good neighbor, the K*2** station would have to have
side bands that are 82 dB down from the carrier over the frequency
range between 300 and 700 Hz from the K*2** center frequency.

Now, the K*2** station needs some bandwidth to transmit.  Say the
transmitter was allowed 300 Hz bandwidth.  If the transmitter uses
a base band filter with a pass band out to 150 Hz, and a "stop band"
starting at 300 Hz, then the filter must cut in 82 dB attenuation
between 150 Hz and 300 Hz.  An RC filter gives 6 dB per octave, but
it looks like we need 13 times that?

OK, so an RC filter will never work, or we've got to cut the transmitter
bandwidth by a huge amount.  Lets figure 24 dB per octave is practical
with a 4 stage filter.  Then we need to divide 300 Hz in half 3.5 times.
This allows perhaps a 23 Hz side band width for the transmitter.

If we went with the simple RC circuit, we need to start to roll off
the filter 13 octaves before 300 Hz.  That would be terrible.  Even
a 1 Hz side band is only 7 octaves below 300 Hz.  A 2 Hz wide filter
would blur even a 5 wpm signal.

I guess I'm wandering off a bit.  It does sound like being a good
neighbor is not simple.

It does sounds like the use of a 400 Hz wide filter on the receiver will
require the use of a very much better (than 400 Hz) filter on the
transmitter if the two are to be good neighbors at 500 Hz spacing.

73 de Michael, AB9GV