Chuck Guzis wrote:
  My first exposure to very-low plate voltage
applications
 was with a code-practice oscillator built from a circuit in the ARRL
 Radio Amateur's Handbook.  ISTR it used a 6SN7 that used the
 6.3v from the heater winding as plate voltage.  At any rate, it would
 drive a pair of earphones.  Perhaps someone with a RAH collection
 from the 1950's might locate the circuit. 
Thank you!  That's the perfect tube multivibrator lead which allowed me to find these
references using "code oscillator" and "6SN7" as search terms:
p29, Dec '55: "A Heterodyne Crystal Calibrator AND CODE PRACTICE
OSCILLATOR" uses both halves of a 7N7, 6SN7, or 12AU7 (or pairs of many triodes) and
two crystals to give marker signals at the DIFFERENCE of the two crystals
p41, Jun '52: "...Frequency Standard...Oscillator" uses a 6F6
oscillator, both halves of a 6SN7 as a multivibrator
p30, Mar '62: Uses a 6AU6 oscillator at 100 kc and a 12AU7
multivibrator to also give 50 kc markers
p33, Jan '65: Uses a 6AQ5 oscillator at 100 kc and a 6AS6 10-kc
sub-harmonic generator (no diode and NO inductors!)
p34, May '65: Uses a 6AU6 oscillator at 100 kc, a 12AU7 multi-
vibrator for 10 kc points, and another 12AU7
All of these and more were found here:
http://newsgroups.derkeiler.com/Archive/Rec/rec.radio.amateur.homebrew/2007…
Do you recall if the p29, Dec '55 project was the 6.3V plate voltage project you
mentioned?
The successful use of 6.3V as the plate voltage in a multivibrator using a standard HV
tube is extremely encouraging.  The lowest voltage used for the audio circuits I've
built was 24V which was most likely used to put the tubes in a more linear operating
region which wouldn't be necessary with a multivibrator.
  While dual (identical) pentodes aren't too common
for
 you to use the first grid as a space-charge element,
 I wonder if one of the sync-separator tubes, such as the
 6BU8 might not do the trick.  Single cathode and first grid,
 independent 3rd grids and plates. 
Thanks, that will be an interesting thing to try.
  In the 20's and 30's before small-geometry
power
 tubes became available, "microwave" experimenters would take a
 hefty triode and put a high positive voltage on the grid and a
 slightly negative voltage on the plate to induce Barkhausen
 oscillation.  But that has nothing to do with your problem, so
 forget that I mentioned 
But thanks for mentioning it anyway since I'm interested in reading about any unusual
operating modes for tubes.
Bill