[AMRadio] 807's triode connection question...

[Bob Bruhns]

Vince,

Since you're experimenting with 807s, here's something else
to try.  It's not triode connection, but it does reduce
distortion considerably, and it retains most of the high
tetrode sensitivity of the tubes.  I have tried this with
6L6GCs and 8417s, so it should also apply to 807s.

Connect the tube as usual for tetrode operation.  Supply
screen voltage from a regulated supply.  But insert a
resistor of about 5.6K in series with each screen grid, and
do not bypass the screen.  That's two 5.6K resistors, one in
series with each screen grid - NOT one resistor, then
fanning out to both grids.  This is important for good
results.

When you do this, the screen voltage will be a little lower
at high audio levels, so you might raise screen voltage 50
volts or so, and adjust control grid bias accordingly.

What I found was that resistance in series with the screens
generally causes the transfer characteristic to curve
downward.  The right amount of series resistance causes this
downward curve to balance out the upward curve usually seen
in tetrode and pentode amplifiers, resulting in much better
linearity.  I found 5.6K to be optimum with 6L6GCs operating
at 350V plate and screen voltage with a typical plate to
plate impedance.  Other values may be optimum at different
operating voltages and impedances.

See also my articles on this subject in AM Press/Exchange,
Issue 100, May, 1992 (Screen Impedance Negative Feedback),
and Issue 110, March, 1994 (Improving the Linearity of
Screen Grid Tubes), at www dot amfone dot net (not sure if
url addresses get through here).

OK, now it's time to go cook some hot dogs on the grill...
happy Fourth!!!

  Bacon, WA3WDR