I'm in the process of
Blackfacing a 70s Bassman
and really appreciate all the
help I've gotten here. So
naturally, I have another
question. 

I've reached the point of
working on the PI input circuit
(and the PI load resisters,
which are both 100K in my
amp), but I've never really
understood how this circuit
produces the inverted signal
(this is the double 12AT7
circuit: I understand the single
triode version clearly). I've
also noticed a few threads
here that have suggested
modifications to the feedback
circuit (including
disconnecting it) and I'm
wondering whether it affects
the inversion process at all, or
simply reduces distortion by
neg. feedback. 

If any of you knowledgeable
people have a moment to shine
a light over here, I'd really
appreciate it. 

-R L

moocow
2/24/98
6:41 PM

One thing to watch for is the
phasing of the output
transformer relative to the
phase splitter input. Some
Bassmans have the green wire
of the output transformer going
to ground instead of the black
one. When you rewire the amp
with the blackface circuit,
make sure you have the black
wire grounded and the green
going to the output, or you will
have positive feedback instead
of negative feedback !

The key to how the inverter
works is the fact that the two
triodes share a common
cathode resistance. When the
signal on the grid of the first
triode goes positive, the tube
begins conducting more
causing the cathode voltage to
increase. Since the other triode
has it's grid AC grounded
(through the .1uF) it doesn't
move. So the cathode voltage
becomes more positive in
relation to the grid voltage
which turns the second triode
off. 

The inverse is true, as the grid
of the first triode goes
negative, it conducts less
causing the cathode voltage to
drop. The cathode->grid
voltage of the second triode
becomes less, turning it on
more. 

The negative feedback tap
moves counter to the cathode
voltage reducing it's swing,
thus reducing the gain of both
sides. 

Hopefully I didn't oversimplify
too much.... 

--johng

John, 

Thanks a lot for this very clear
explanation. It all makes sense
as you explain it. I've been
doing a bit of blackfacing, and
studying some circuit
diagrams, I've noticed that
Fender seemed to prefer 1 meg
grid resisters in this circuit
during the tweed and blackface
era, and a combination of an
82K and a 100K plate resister,
but later, in the
"mastervolume" circuits,
switched to lower values (like
330k) for the grids, and
matched 47K for the plates.
Do you think the lower values
are mainly about reducing
gain, or are the grid ones
designed to compensate for the
1 meg mv (bracketed by
coupling caps)? Any idea why
they moved towards matched
values for the plates at this
point? 

It seems the general consensus
is that the BF circuits have a
better sound; do you think it's
worth the change? 

My particular amp also has
both the preamp stages routed
through an additional stage that
has a negative feedback from
plate to grid, to which the
feedback from the output is
also attached. I'm wondering if
anyone has an opinion on this
circuit. The BF bassman has
an additional stage, but only
for the bass channel, and w/o
feedback loop (the output
feedback goes into the PI
circuit, as usual). 

This is all pretty picky, I
know, but I'd appreciate any
thoughts you or anyone else
might have. 

thanks, 

R L

Do you think the lower values
are mainly about reducing
gain, or are the grid ones
designed to compensate for
the 1 meg mv (bracketed by
coupling caps)? Any idea why
they moved towards matched
values for the plates at this
point? 

My take on these changes are
this: 

The grid resistors and plate
resistors were changed for
different reasons. The grid
resistors were changed from
1Meg to 330K along with the
coupling capacitor, .001 ->
.01uF. This lowers the Bass
rolloff frequency into the
phase splitter by almost 2
octaves. I think this was done
to compensate for the speakers
Fender was using at this time.
From personal experience I
know that the Utah speakers
have much less bass response
than the Jensens do (at least
Naylors' reproduction of the
Jensen). When I blackfaced my
Deluxe Reverb without
changing the speaker I ended
up with a very thin sounding
amp. Changing the speaker
fixed that. 

I think the change from
100K/82K to 47K in the plates
was to compensate for the bad
things the capacitors on the
grids of the output tubes did to
the high-frequency response.
Hiss is always a problem with
these amps and I think Fender
put these caps on the output
tubes to knock down the hiss.
There are also reports that
these amps have a tendency to
oscillate because of lead dress
if these caps are removed.
However, they (maybe?) didn't
consider how much of a
loading effect this would have
on the plates of the phase
splitter. By reducing the plate
resistors, they take a slight hit
in gain but increase the high
frequency response by at least
an octave. Although better, it's
not quite enough. 

--johng

Thoughts on the multiple
feedback loops of the AA371
Bassman ? They're just plain
evil! 

Assuming the amp will be
used for bass, I use a variation
of the AA864 Bassman on
amps like this. As you noted,
the guitar channel does not go
through the extra stage, and the
feedback loop is the same as
other blackface amps. This
keeps the guitar channel from
being compromised. 

Use the stock volume and tone
control circuit and .01 uF
phase inverter input cap from
your AA371, but the rest of the
amp should be AA864,
including the bias circuit. 

If you are truly ambitious, try
the 6G6-B Bassman circuit. It
uses four triodes in the bass
pre-amp (before the phase
inverter) and it splits the treble
control from the volume and
bass. You'll have to replace
some pots, which may be hard
to find.

* * * * * *

I just had one of those in my
shop this week for a complete
rebuild and the owner wanted
the bass channel converted to a
Plexi Marshall style with a
cathode follower and tone
section etc etc. (I used the free
unused triode section for this.) 
I also redid the guitar channel
to sound more like a 
hot black face Fender with
overdrive.Very cool! 
On the phase inverter I lifted
the grounded end and installed
two 2K2 resistor in series
with a .1uF cap connected to
the center of them and then
grounded the cap and one end
of the reisistors then I
connected the other end
opposite end to the ungrounded
phase inverter section. 
Sorta like a Tweed
Bassman/Marshall but no
presence pot, just the two
resistors as a voltage divider
with the cap bypassing the
grounded one. 
Man was that wordy, 
Bruce