[Prune]

I did the XY thing with the scope, feeding a triangular wave from my
soundcard into the amp. I reduced voltage and bias current about 7%
each, and the display remains a straight line, I can't see any
curvature, unless I push it close to clipping. I don't know to what
extent it's the feedback correcting this, since the amp doesn't work
with feedback disconnected, the DC offset gets huge (even though it's
pretty low feedback since it goes back to the JFET sources, not gates).
I also can't hear any difference. So I guess I didn't have anything to
worry about...

As for the circuit being PP, I don't see it. Each side drives a
separate electrode on the headphones, but each electrode is driven SE.
A differential output amplifier is not the same as PP since the side
pushing is on the other side of the load from the one pulling at that
time.

The one problem I still have is some drift of the DC offset as the amp
warms up, so I'm hoping that several DC volts offset wouldn't be a
problem since the load is a capacitor and shouldn't draw DC.


Patrick Turner wrote in
:



Prune wrote:

Thanks, I guess that answers that question. Now if I could only find
curves for the MOSFET. The datasheet only has current vs voltage
ones.

But what current and what voltage at what electrodes?

Mosfet data should appear like pentode data, ie with Id plotted on a
vertical
scale and Ed horizontally, and the curves you see are Rd curves at
various values
of Egate.

But you have a PP circuit, and wishing for complementary oddorder
distortion cancelling is rather like like
wishing santa make a special trip in July.

Just remember that a +1V applied at the cathode gives +20V at the
anode; the tube becomes a non-inverting amplifier.

One way is to fire it up and start fiddling with bias currents,
but without tempting smoke, and see what your THD meter tells you.

If I attempted direct drive of ESL panels I'd use a far simpler
amp circuit without any SS within, except perhaps for a CCS load in an
input LTP stage.

Good Luck,

Patrick Turner.


Patrick Turner wrote in
news:4608C3FE.2DCAD8A1 @turneraudio.com.au:



Prune wrote:

Hello gentlemen,

I've not been able to find EL34 curves for cathode-driven
operation anywhere. Help would be greatly appreciated!


The grounded grid curves are the same as for grounded cathode in
either pentode, UL or triode.

Voltage gains are indentical.

All that changes is the input impedance which is high for grounded
cathode, and very low for
grounded grid, and Rk in = RL / voltage gain.

There are no separtae groups of curves for grounded grid.

Musicman guitar amps used transistors to drive the cathodes of
output tubes,
and this could have been a reason why they went out of business.

Patrick Turner.