[Elecraft] Opinions please.

[Geoffrey Mackenzie-Kennedy]

My K2/100 is s/n 3255, the same "vintage" as yours. For USB/LSB I set the
BFO frequencies the "hard" way by manually plotting the KSB2 filter's
response, and placing the BFO where best sounding transmit and receive audio
was obtained. It turned out that unwanted sideband (close in and far out)
attenuation was more than adequate. This is with an unmodified filter. Did
the same for narrower bandwidths, and even at 1.5kHz filter bandwidth it is
possible to copy when the frequency sounds like a pig's breakfast. Reports
when using the KSB2 filter are good.

Regarding the Post Mixer amplifier. If a bias change does not change the
signal handling capability of the Post Mixer amplifier as a stand alone
stage, and its IP3out is much greater than the IP3in of the filter and
following stages, while you will get Xdb increase in overall gain the IP3in
of the receiver will suffer by about Xdbm. If however, as is more likely,
the bias change reduces the IP3out of the Post Mixer amplifier as a stand
alone stage, the IP3in of the receiver will suffer by more than Xdbm. I have
not included the effects of the mixer and its IP3out, nor distortion added
by tuned circuits in the front end.

A serious problem occurs if the Post Mixer amplifier becomes "transparent",
and the wild input impedance variations of the filter at its passband
corers are seen by the mixer.

It is my hunch that the 2N5109 was chosen for use in simple but proven
circuits while producing the required performance. I feel certain that Wayne
and Eric were also keeping to a power budget while designing the K2. MOSFETS
can be good, but can introduce problems. If power consumption is not a
restraint, the options open up. A pair of MRF580s in pushpull "noiseless"
feedback Norton configuration for example provides a very strong amplifier
with very low noise figure, but it draws 120mA ! Also the device gain
bandwidth is in GHz, so circuit layout and construction must be precise or
else you have a UHF oscillator - not good in kits or most places.
Furthermore
there are four transformers to build.

73,
Geoff.
GM4ESD


> ----- Original Message -----
> From: <WA2JJH at aol.com>
> To: <elecraft at mailman.qth.net>
> Sent: Tuesday, February 22, 2005 11:11 PM
> Subject: [Elecraft] Opinions please.
>
> >
> >   I have had hours of great QSO's so far. I know see an option to open
up
> > the SSB filter on RX. I use the stock 2.2khz and I like the audio. I
find
> the
> > 1.8khz filter on SSB to be better than that of my Drake TR-7 and Kenwood
> > TS-850.
> >
> >   The 1.8khz on the K2 sounds full bodies on SSB.
>
> >
> >   I was also struck by the choice of transistor used in the RF  pre-amp.
> Why
> > a simple bipolar and not the industry standard Mosfet. I was  thinking
> about
> > trying the bais change in the post IF mixer to get an even lower  noise
> floor.
> >
> >    I like to operate pre-ampless when ever possible. I have found  this
> > reduces operator fatique.
> >
> > I do have to say that I am still impressed by the K2's RX. However my
> Drake
> > TR-7's pre-ampless design does give that almost zero noise floor sound.
> >
> >
>