Most Noteworthy
changes:
Many traces must be cut, and
some added. These are not documented, but you can likely see some of them on the
bottom-side photo of the printed circuit board. Other cuts are on the top-side
that youll have to figure out yourself.
The
1N4004 varactor diode was replaced with a vari-cap of greater range, a 1N1404.
Four turns extra were wound onto a molded 3.9uH commercial inductor to give a
tuning range that includes the bottom of the 17M band.
I used 4.000MHz parallel resonant crystals in the I.F. and in the NE602
oscillators. This choice makes the VXO range stretch rather far at the lower end
of 17M. Id suggest using 4.000 MHz.
series-resonant crystals instead youll lose a bit at the top of the band,
but wont have to stretch as far at the bottom. I found the top end extended
into SSB territory a waste of bandwidth. With this super VXO, a ten-turn
tuning resistor was needed. Youd find that a one-turn pot gives much too quick
tuning.
I
have the notorious 38S thump pretty well beat, by moving one of the CMOS
switches past much of the audio gain, between the two op-amps. Moving this
switch here requires a good bit of hacking PCB traces, but is worth the effort
in cleaner transmit/receive switching. A PNP audio preamp reduces audio noise coming from the 7808
regulator.
One CMOS switch was eliminated, at the output of the 2nd
NE602. Two outputs are required one for the transmit signal at 18MHz, the
other is receive audio. Since the NE602 has two similar outputs anyway,
switching isnt required.
I found that some of the small ceramic capacitors in the early audio
stages caused microphonics. Any coupling capacitors should be mylar, polyester,
or polyproplyene.