[Elecraft] Alligators and Bats

[Paul Manis]

Tree,

In response to your comments, I had to go look at the extended tests for
the in-band IMD (e.g., within the passband of the receiver - two tones
100 Hz apart) on the K2 versus the 870 (the 850 apparently came before
the extended tests, so I looked at the 870 for comparsion). The 870
shows pretty terrible performance with the AGC fast (3rd order IMD only
-35 db; 5th order -32), which with more than 2 signals is basically
going to make a lot of fluctuating background hash corresponding to
comodulation of the signals - and I don't think comodulation masking
release (yes, that's a real phenomenon in auditory perception) can even
help that. It looks like AGC slow is a lot better on that radio, but
that seems counterproductive when operating CW. In contrast, the K2
specs out with the AGC set fast at 3rd order -43 dB and 5th order -47
db, and slightly better for slow AGC. It would be interesting to make
the measurements with the AGC off, but that doesn't seem to be part of
the protocol. So, when listening to a CW pileup, or to the activity
during a contest, you'd expect the K2 to be a fair bit cleaner sounding,
and provide your ears with a representation of the signals that can be
more easily dissected. Ultimately, this should mean less fatigue for the
contester. I can only compare it with an ICOM 730, and yes, the K2 is
cleaner and not as mushy or muffled (I've always thought there was too
much AGC action in the Icom). Even the TenTec Omni's in-band IMDs are in
the 37-39dB range. 

I have to agree, one wants a very clean sounding receiver; like Rick
Campbell's R1/R2 direct conversion receivers (they sound _really_ nice -
no agc!), or what Jay Rusgrove described in QST back in the 80's. This
lets your brain do the best job it can, without having to also deal with
receiver IMD and the effects of DSP. It might be interesting to put the
output audio stage of the R1 in a K2, for a little extra punch - but
maybe not for field use.

But, where it should really shine is with the "close-in" IMD for signals
more than 2 kHz away, as compared to the other rigs using roofing
filters that are 15-50 kHz wide. I'd like to know your impressions after
a strenuous domestic contest like SS with decent antennas at the front,
or in the DX contests with everyone running KW's to stacked Force-12's
right next door to the V51 running 100W to a wire.

Now, some people may think that my ramblins so far are off topic, so let
me just say this about some kinds of bats - the ones that use sonar
signals for echolocation. They have a very interesting specialization in
their cochleas that lets them have a really sharply tuned, high
sensitivity region near the echolocation frequency. This is a lot like
the swept frequency 3rd order IMD plots for the K2 - a nice sharp
filter, right up at the front end where it can do the most good. I
wouldn't push this analogy to far though.

Paul,  NC3G.