i think those people are suggesting you change the 22nf's to 100nf's., but I don't know if that'll really do the trick.
I'd start with the single triode channel. Turn the ef86 channel to zero and forget it for now.
disconnect the 1nf on the tone control and see how it sounds. If that doesn't do it, then Id try changing the 250K volume pot for a 500K or a 1M.
I should add my advice is based on looking at low-frequency roll off of an inline cap followed by a resistance to ground. If the resistance is too low -or the cap is too high- low frequencies will be attenuated, highs will not be.
The 22nf coupling caps followed by a 220K resistance to ground (via the volume pot) sets up a 33Hz low freq. roll off.
The 47nf caps followed by the 200K grid leak resistors set up a 17Hz low freq. roll off.
these seem reasonable to me, at least compared to the RC circuits that the vol/tone controls form:
if the volume is at 50%:
-if tone control is at the upper wiper, it looks like everything below 1447Hz is att. 50%, and everything above goes though the 10nf cap 100% (C=1nf R=110K)
-if tone control is at bottom wiper, it looks like everything below 842Hz is att. 50%, and everything above goes though the 10nf cap 100% (C=1nf R=110K+220K) *except* what high freq's see the 10nf as ground.
No matter what, it looks (to me) that the tone control and the volume control can interact to heavily favor high freq's over lows.
Of course, that's if volume control is at 50%. if you run it wide open, the tone control loses some of its impact.
you can go after coupling caps, but don't look at a value of .047 in a vacuum, analyze the cap value AND the resistance to ground following the cap to determine if its a 'bright' setup or not.
Also, regarding cathode by-pass caps, either the 22uf or the 470uf: lowering the values will promote higher gain for high frequencies, and lower gain for lower frequencies. 22uf on the first gain stage gives the stage maximum gain for all audible frequencies. As you lower it to 10uf or 1uf, higher frequencies will still get maximum gain, but lower frequencies will operate at minimum gain. The 470uf value is more common in hi-fi circuits, and a 22uf is more common in cathode biased guitar amps. The 470uf probably makes the amp overdrive more like fixed bias amp than a typical cathode biased amp, but that's a lot of speculation.
a lot of factors can make an amp bright. speaker (The_Gaz) is huge, as is the type and size of wood, and the design of the speaker cab. If you can, like The_Gaz suggested, plug the amp into a cabinet that you know to have good frequency response on the low end.
Jumpering the channels is not recommended.
+1.
BTW:, anyone know whats up with the .68 cap to ground on the 4ohm tap? seems like that could be the culprit, too.