Re: 6SN7 amp: what am I seeing?


"RS" <RS@xxxxxxxxxx> wrote in message
news:2nfkp4lfp0e93h5jk5tlhu4tf4jspeihoj@xxxxxxxxxx
Consider what happens in a given circuit when a tube is full off vs
full on. A normal gain stage looks like a conventional voltage
divider, with the plate resistor forming one leg, and the tube forming
the other. When that tube turns off, the voltage swings close to B+,
as it's pulled up by the plate resistor. When it's full on, it goes
toward ground (it won't get there, but it tries). So assume B+ to
ground as a theoretical swing.

That will be typical of a regular gain stage, and of both the
Fender-type PI (long-tail) and the paraphase. (Note: The long-tail
does have a bit of overhead, but negligible for the conversation).

By contrast, when the concertina turns full on (again, assume zero
ohms), the plate and cathode resistors are both in play, forming their
own voltage divider. So both plate a cathode voltage will be at 1/2 B+
rather than ground. The plate can swing upward from that point. The
cathode can swing downward from there. But it's essentially limited to
1/2 of the voltage swing.

Now add in the fact that there's a 'saturation voltage' when a tube is
turned full on. It's not really zero. So that decreases voltage swing
even more.

This is different from absolute gain. In fact, if an output stage
requires high voltage swing, then a concertina could be a bottleneck,
and cause clipping. Not sure what your drive requirements are in this
case, but you'd at least have options for other output tubes.

Aha! I think this is exactly the critical question. Rather than design for
what might be, let's focus on what is. The finals are a pair of 6SN7GTB
cathode biased, wired so that each tube alone is push pull and each section
in tube #2 is parallel to a section in tube #1. Here's where the well goes
dry for me. How do I determine the drive requirement for a 6SN7?

I now wonder again, Gary G was in favor of the concertina. (Not ready to
change so quick, as it's a bit of work now.) Maybe he knows that it will
not be a bottleneck and better yet, a good match for this particular tube?

Approximate schematic here:
http://home.comcast.net/~psymonds/6SN7amp.htm
Cathode resistors on the finals are now 1K. The replacement PT is showing
460-0-460 in circuit. B+1=461v, Va=452v, Vk=~17.5v. Plate load resistors
on the PI are 100K, not 120K. B+2=396v (plates at 252 and 262), and
B+3=392v (plates at ~230v). Grids are all at or close to zero. Preamp/PI
cathodes are all over 2v and less than 3v.

I am about to insert a 1K dropping resistor just to see if I can get DCV out
at the rectifier below 450 (cap rating, spec *** max on 6SN7), and the PI
below 250 (spec *** max 6SL7). I will also be moving forward, one step at
a time, with lifting the cathode bypass caps on the preamp stages.


.

Loading