Grid stoppers causing HF oscillation?

[ECV02]

Hi all,
I recently soldered up a 10 watt push pull ECL82 amp, designed for guitar. When I cranked it up to 11, I would get the nastiest (and I mean NASTIEST) sound coming out of the speakers. The amp would motorboat at around 13Hz, but upon further inspection I noticed a roughly 40kHz sine-ish wave on the grid of the first gain stage. Since it happened when cranked, I tried changing the overall gain (there is no NFB!) to no avail. Out of desperation, I bypassed the 68k grid stoppers and it worked just fine. Now, however, the amp is extremely sensitive to RF interference, and if I bring my cell phone anywhere near the amp I can hear radio stations being demodulated  . Is there any way to circumvent the grid stopper oscillations? I thought those resistors were supposed to attenuate those HF oscillations to begin with. Any help would be greatly appreciated.

-E

[shooter]

make 'em 33k
add NFB
only go to 10.2

[sluckey]

Got schematic?

[ECV02]

Here it is… probably should have posted it in the first place. Please treat those final 2 triodes as pentodes with 2k2 screen stoppers, fed from the 220v B+. Sim won't let me draw pentodes for some reason.

[ECV02]

As a side note, the motorboating only happens when the cathode caps are connected.

[Willabe]

That's only part of the schematic, need to see the power supply too.

Where were the 68K grid stops mounted? On the input jack or tight to the grid pin on the socket?

[ECV02]

Stoppers were mounted on the perfboard. The power supply is from an SMPS with multiple outputs; not quite sure how to draw that one up. Noise frequency on it is well above 250kHz, so I'm fairly convinced that the power supply is not to blame. There is about 1 volt peak of noise, but like I said, it is well out of the audible range.

[Willabe]

Stoppers were mounted on the perfboard.

For grid stoppers to work (best), they need to be mounted directly to the tube sockets grid pin, with as little lead as possible between the resistor and grid pin.   

[pdf64]

Yes, a proper grid stopper is indicated.
Bear in mind that that anything in the grid circuit will act a receiving antennae.
A low value grid stopper for the 2nd stage, eg 5k6 - 10k, will also help to bring the HF response into line.