G-String Insanity (not sex-related)

[clyde1]

Oh, if only it were sex-related I could probably handle it. Ha! Enough whining and on with the description. I built a 2xEL84 head, 1st channel is a triode half (or paralleled halves) to a cathode follower, tone stack and PI.
Second channel cascades 2 triodes into the same cathode follower and tone stack, channels separated at the CF by resistors. Ch2 obviously has much more gain than 1. When plucking the G string, esp. around the 4th and 10th fret, I get a buzz on top of the note which increases with gain and treble, both channels.
I've subbed in different PS caps, changed my grounding scheme as this is what I suspected but have done several builds following the same template with no trouble whatsoever. The TX's are Mojo 18W Heyboers, also subbed in another OPT, no difference. I've sent signal from V1 and V2 directly to the PI, omitting the CF and tone stack, same problem.
I just now hooked up a signal generator putting 1V to the first triode through the amp and measure 14V on the grid of the first EL84 and 6V to the grid of EL84 2. Shouldn't they be equal? I'm pretty much out of options here, the amp seems wired correctly, lead dress has never been an issue
with me. The PI has std. 18W configuration, 100K plates, 470K grids,
820R cathode and 56K tail resistor, .02uf cap from 2nd grid to top of 56K.
I'm thinking it's the PI but I've hit the ceiling here.

[Enzo]

Just because a resistor is MARKED 100k doesn't mean that is actually its resistance. If teh PI is not balanced, find out why. Start by verifying the parts.

[clyde1]

Thanks for the reply Enzo, I just wish it were that easy, both resistors measure 99K. Supply voltage is 269V, input plate is 204V, 209V input 2,
input grid 47V, input grid 2 49V, top of cathode R is 69V, top of tail resistor is
68V. After posting my query I did a little more digging and found that pin 1 and pin 2 of my EL84's were internally connected, thus shorting out my 8.2K grid stoppers mounted on these pins-Aha, a sure cause of this parasitic. I re-wired them properly, same problem. These were EI EL84's, the JJ's are not internally connected between pins 1 and 2, a word to the wise if you're mounting grid stoppers this way, although if EI's go the way of the dinosaurs it won't be an issue-hope not.
Returning to topic, would unbalance cause this problem and only on one string?

[Ray Ivers]

Clyde1,

channels separated at the CF by resistors.

Do you have a schematic?

Ray

[clyde1]

I sure do Ray, but only on paper; how best to post it or get it to you? I could
fax it if you would be willing to look at it? Thanks.

[Ray Ivers]

OK, got the print... a few things you might want to look at:

1) The .02uF cap on the 2nd PI grid really should be grounded. Perhaps some amp(s) out there hooks it up this way, but I'd try connecting it to the 56K tail resistor's ground point instead.

2) Six 12AX7 stages load the 258V B+ node, with its tiny 16uF cap - this is about three more than I'm comfortable with, but YMMV. I'd place a 4.7K series B+ resistor followed by a 50uF cap to the first two tubes' ground point(s) in the B+ line, directly above the .001uF cap feeding the 500K volume control.

3) The 2K shared screen resistance (following the 1st 32uF filter cap) will throw some serious screen modulation on the B+ bus; one reason why many amp makers use a choke here even when their accountants tell them not to , and also why small-value resistances are usually chosen for this point in lieu of a choke.

These things may or may not help your G-string problem (FWIW, the first thing I thought of was fret buzz), but I would give them a try anyway just to see.

Ray

[clyde1]

Thanks Ray, lots to chew on and I'll start sussing this out tonight.
Fret-buzz it weren't, two different guitars, both Strats, particularly bad in the out-of-phase position. FWIW I tuned the amp with my Hollow T, proving that a good guitar even makes a bad-sounding amp behave. Ha! Too bad Tom's
shutting it down. I'll be back.