[Alex Pogossov]

"John L Stewart" wrote in message
...

Here is a different kind of test on the PP 6V6 AB2 amp that I posted a
while ago. And an updated schematic. The PS is a Hammond 271X into a FW
CT, 1N4007s into a PI filter of 8 microF, 5H & 100 microF for the +ve
side. There is also a -ve supply using 1N4007s on the same HV winding
into another PI filter of 10 microF, 1.5K & 10 microF. The lid is held
on for the -ve side by VR150 gas regulator. The OPT is a Hammond 125E
Universal.

The 6V6 screen leads go to the +ve supply. You can connect the dots.

From that one might think not to much is possible but they would be
wrong. My objective with this exercise as I had previously stated is to
try some different things. That is, discover what is possible in a
vacuum tube amplifier with low cost parts. It is well known that by
throwing unlimited resources (money & time) at a project that most
anything is possible. That is not my intention here!

There are many tradeoffs in any system. It is the designer's task to
separate what is important from what is not. In the real World market
the end product must be competitive.

The trace shows what happens to the +ve supply under pulsed conditions.
The top marker indicates 10 watts. The red trace is the one KHz test
signal, pulsed. The blue trace is the B+ supply at the 6V6 plates. The
pulse rep rate is about 14 Hz.

How did I do this? I will post that info later.

Cheer to all, John


+-------------------------------------------------------------------+
|Filename: Schema And Burst Example 10W 20C.jpg |
|Download: http://www.audiobanter.com/attachment.php?attachmentid=343|
+-------------------------------------------------------------------+

Apparently it is useless to allow for class AB2 output stage push-pull
operation while class AB1 with elevated Vg2=320V is quite capable of working
the output tubes 6V6GT to thier limits. It would seem that class AB2 can
only ruin the screen grids when the amplifier is overdriven into clipping.

However, upon reflection I can see one advantage of class AB2 in these
conditions.

Usually, in class AB1 control grids are driven via decoupling capacitors.
When an amp is overdriven, these caps can charge by the grid current into
class C1 and once overdriving burst is gone, get the amp virtually muted for
a while, until they discharge to the nominal bias level.

To avoid this coupling caps charging effect it is better to use cathode
follower drivers in front of the output tubes, but ( IMPORTANT! ) insert
rather large resistors (say 47K...100K) between the cathode folowers and the
control grids of the output tubes. In this case, if the amp is overdriven,
the coupling caps in front of the cathode followers will not charge as the
cathode follower will probably never run into grid current. At the same
time, large grid resistors will clamp positive excursions of Vg1 on the
output tubes. So 6V6GTs will work in sort of benign class AB2, not much
harder than AB1. Thus it will be safer fo the output tubes and there will be
no recovery problem after a burst. (These series grid resistros shall be
bypassed by small capacitors, say 100pF.)