The nice thing about RC time constants is the R and C don't seem to know the
difference if they are in series or parallel.
The same formula I gave you works and I think of it in terms of 3db too.
So I would look at the RC as a circuit not particulary in parallel with the plate||plate
load... etc.
68pF||68K = abt 34KHz to 35KHz.
The 100K pot with max R would make the circuit act like the capacitor was bigger in
value and lower the frequency.
68pF||100K = abt 23KHz to 24Khz
But the resistance to ground is now higher so it will have less shunting effect.
So if you really wanted the 23KHz to 24Khz shunt with about the same attenuation
efffect, I'd just use a bigger cap.
100pF||68K = abt 23KHz to 24KHz.
22m||.05uF = abt .1Hz!!
So ALL the audio frequencies are shunted back to the grid.
That's negative feedback from almost DC to all freqs!
But then there's that darn 22M series resistor sitting on top of the grid load resistor...
hmmm, a voltage divider.
Get it?
Bruce

This 68pF/68K combo is not some kind of genius secret tone thing, it just shunts
extremely high freqs to ground and for the most part, in a non evasive way. 
But, I've found the 68pF/68K thing works really good in amps when hitting EL84s
too hard and it seems to get rid of some of the harshness in hard driving Fender
amps too.

I started using it on BF Fender Deluxes when the cheap Chinese 6V6 tubes I was
getting sounded harsh and fizzy on the extreme top end.
I couldn't think of any other way to slope that off without killing the livelyness of the
amp.
It can go on any plate load stage I'm sure, but I like it a little better when it is right
before the PI driver cap. 
In a Fender with AB673 circuit with reverb, it would be on the triode section right
after the tone stack and before the 3m3/10pF combo.
So yeah, on pin 6 right to the cathode or ground is OK.
On my custom amps, I have an eyelet on the board to run them right after the
coupling cap directly back to the ground of it's cathode resistor/cap...if I need to
use it, I can.
Boy, it sounds like you totaly get it when it comes to the small plate cap to ground,
the cap across the plate load resistor or the small plate cap to the cathode.
The main thing is to get the cap AC grounded on one end.
Applying it to the cathode probably injects a little more negative feedback at that
freq though.
I bet the cause and effects are nearly the same.
Another interesting thing to try is splitting the anode resistor into two different equal
values and bypass one of them with a 500pF cap or split the anode load and use a
small value resistor on the plate (20K)+ (82K) and connect the coupling cap
between the two plate load resistors and then bypass the second one... that is
feeding B+.
I guess there would be a number of different things to try...if you even need to do it
at all.

    Question: is there any noticable audible difference between adding in a
    100pF cap:
    1) from the plate to ground
    2) from the plate to the cathode
    3) across the plate resistor

Hmmm. I'm not sure. I 'll try it tommorrow and see.
I think I prefer using the resistor and cap in series because I can build a thru RC
circuit and limit how much of the unwanted audio is passed to ground.

A few months ago when you asked me if I had seen the 22m/.05uF NFB thing used
before, I told you about old ham radio tricks like the "gimmick" and plate
neutralizing.
It's all part of the same thing only at audio and sometimes one can extract great
sounds from basic/simple ideas.

Bruce 

I didn't think you were still using the ceramics in the reverb.
Yes, the old time domino silver mica caps!
At one time I had a large cigar box filled with them from previous past lives! 
I think almost every tweed Fender amp I've worked on had at least one in them.
I think it's pretty cool you could duplicate a bad sound!
Personally I had been thinking it is harder to make a great clean sounding amp then a
"distortion" amp.
But , yes Gil, a great distortion amp is probably just as hard anyhow. 
Heck they're all hard!
I'd like a little more analysis on the effect of the reverb mixing section in the Fender
amp and the distortion problem.
I wonder if you are hearing the preamp's odd upper harmonic distortion products
leaking right past the 10pF SM cap unattenutated.
The cap and resistor form an RC 5KHz.
My little 68pF/68K shunt is about 34KHz. 
Those small SM caps with short leads can probably pass RF with relative low ESR
and all that.
I wonder what would happen if you put a small inductance in series with one and
possibly a few pF of a gimmick to ground before it.
You remember the gimmick conversation Steve?

Well, sometimes I feel my tongue has been biten off a few times too many as it is
and I still get just as tone deaf as the rest you guys playing too long with the same
amp and circuitry over and over.
If I spend too much time on any tone circuit, my ears get fatiqued and I end up
making them too bright. 

Yes, I have my share of parts drawers full of 27pF to 680pF 500V/1Kv ceramic
caps and I do use them too. 
But, I find that more often then not, a SM cap is either transparent or adds shimmer
in a nice way to those amps.
But remember what I said about fat Willie here and the harp players, they hated the
SM caps in their Fender amps.
Now I mentioned to Steve a year or two ago that I also have found that I can use
less top end or reduce the effect of a presence control with the amps that have SM
caps in the tone circuit. 
Bob Swanson, DynaQ Amps, has a theory about spreading high and low freqs that
is very interesting and has proven to me a few times that some amps sound great
when tweaked to freqs I know I can't even hear!!
BTW, I also frequently use a SM 47pF to 110pF cap (68pF being the cap used
more often) in series with a 68K resistor shunted to ground, right after the plate load
resistor and before the coupling cap from the tone recovery to the next stage, as a
way to cut some of the same shimmer I guess I just like it. Sorta like Dolby?

Not that I think I am too good to try any.... :>) No, I have not tried to mock up any
of the new Dumble tone tweaks yet, except a loose adapatation of a modified treble
pot to bass pot deal that worked very nicely and I really liked the added flex it gave
my amp.

I'm afraid the thread original Dumble Amps ran so long, I became Numble. pun
intended

Bruce