[HBR] vacuum tube ham transmitter

Sun Feb 7 21:32:26 EST 2010

Any ideas on a receiver of this type

On Feb 5, 2010 5:33 PM, "Ian Wilson" <ianmwilson73 at gmail.com> wrote:

This is a shameless cross-posting of a posting to Glowbugs. My apologies if
you see it twice.

--ian

Ever since I read the 'KCS Compactron' pair of articles in the GE Ham News,
I've
been interested in what could be done with the last gasp of receiving-tube
technology.

About 6 months ago, I started idly collecting Compactrons, and playing with
tube
lineups that would be useful for an HF receiver. A very useful summary of
Compactrons
can be found here:
http://www.junkbox.com/electronics/CompactronTubesIndex.shtml

It is not easy to allocate the functions of a superhet across a number of
multi-function
tubes while maintaining isolation between stages, etc. After some
experimentation, I've
arrived at something workable although it's still a work in progress.

Since I am remarkably poor at metalwork I wanted to try something different.
This RX
uses a number of modules built on single-sided 0.060" PCB. This snaps
cleanly if you
deeply score both sides, works FB for ugly construction with lots of things
being
grounded, and allows screening to be added as you go. One benefit I hadn't
thought
about before was that you can use those 'useless' PCB tube sockets. A couple
of my
Compactron sockets have a central metal tube that works well as a partial
screen and
anchors the socket firmly when soldered to the ground plane. Pictures of the
current
set of 'modules' here:
http://www.flickr.com/photos/17477299@N04/4327375853/in/pool-glowbugs

The receiver is a single-conversion superhet with 9MHz IF and a 'premixer'
VFO.

The modules are detailed below. I will TRY to find time to draw up the
schematics and scan them
some time this week. The interconnects between module 1 and 2, and module 2
and 3, are at
150 ohms and 9MHz. Most of the inductors are toroids, and it's easy to add a
few turns to step
down/up to/from the 150 ohm intermodule impedance.

1. First mixer (Pullen) and premixer VFO
   6AC10 (triple hi-mu triode) cathode follower and mixer. The 6AC10 has
reasonably high
   gm, which is desirable in a mixer. One section of the tube is available
for use as a calibration
   oscillator.
   6BH11 (dual medium-mu triode; sharp cutoff pentode) VFO, crystal
oscillator and premixer.
   This took lots of experimentation. The following works reasonably well:
      - triode Colpitts VFO, 5-5.5MHz (very happy with the stability -
high-quality components with
        the tank LC in a shielded box several inches away from the heat
sources)
      - triode xtal oscillator (currently only 11MHz xtal in place. Should
be using a 21.5MHz xtal)
      - pentode mixer (VFO to grid; xtal oscillator to cathode) with
double-tuned over-coupled 16MHz
        bandpass filter
      - there is lots of space to add xtal/bandpass filter for other bands
but only 40m is there at present
   9MHz homebrew xtal filter

2. First and second IF
   6AR11 (dual remote cutoff pentode). Both stages AGC controlled. Output
for S-meter available from
   one of the cathodes. In order to tame this (lots of gain at 9MHz in a
small space), I needed to lower
   the Q of the tuned circuits (only single-tuned circuits used, since the
9MHz filter provides the main
   selectivity); add shielding between the input, middle, and output parts
of the circuit; and screen-
   neutralize the tubes (see the second IF amp of the KCS circuit for an
example of this). The amplifier
   will oscillate with some input/output terminations but seems happy now.

3. Pullen product detector and BFO
   6U10 (dual medium-mu triode; high-mu triode). This is a great tube for a
Pullen configuration, since
   you can use one medium-mu (high gm) triode as the cathode follower and
the high-mu (low gm) triode
   as the mixer, without having to lower the plate voltage on the second
triode (which you have to do if using
   identical triodes). The plate current of the high-mu triode is only
1.2mA, so you can use a high value plate
   resistor and get lots of free audio gain. My measurements showed voltage
conversion gain of about 20.
   The BFO uses an identical xtal to those used in the front end filter in
a Colpitts oscillator. A relay is used
   to switch either a capacitor or an inductor in series with the xtal for
LSB/USB. This works fine with the
   capacitor (frequency pulled up) but with the inductor value required,
the output level is lower than ideal.
   (Getting the BFO right has been a consistently difficult area in
receiver building for me, so far).
   I plan to add a couple of Ge diodes to as a cheap-and-nasty AM detector.
The BFO can be turned off
   for AM.

4. Audio amplifier and AGC
   6T10 (sharp cutoff pentode; beam power pentode). Now working FB thanks
to (Brad T. and numerous other
   helpful suggestions from other list members). This is not going to be
the ultimate in hi-fi. A 6T9 triode-pentode
   would probably be a better choice, but I don't have any of those, and
they are probably all being used as
   QRP transmitters anyway <g>.
   6U10 (dual medium-mu triode; high-mu triode). This is intended for an
audio AGC amplifier, CW filter, and
   gain stage for AM if required (currently none of this is implemented).

I might add a power supply underneath the audio module, and to stack the
IF/product detector modules.
This would result in an 9.25" wide x 11" deep footprint (so an 11" piece of
PCB will work OK as a front panel).

This could also be extended to an all-Compactron transceiver by adding a
mixer (premixer output mixed with
9MHz BFO and lowpass filtered gets you there), buffer, and PA stage (7984,
8150 or whatever). However, I think
that this would work out better if all the frequency generating
oscillators/mixers were implemented in a single
module. Now where's my pencil....

73, ian K3IMW
______________________________________________________________
HBR mailing list
Home: http://mailman.qth.net/mailman/listinfo/hbr
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:HBR at mailman.qth.net

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html