Tron-like sound when using EL34 instead of 6L6 in "Two-Stroke" amp

[MagnaClone108]

Hello all,

I built my first amp, Dave Hunter's Two-Stroke, about 10 years ago using a 5f2a chassis and the "traditional" Fender layout. Inspired by the simplicity of my MagnaTone Varsity Deluxe 108 and its point-to-point wiring, I decided to recreate the Two-Stroke using P2P wiring. The Two-Stroke is an SE amp based on the 5f2a circuit, with only two resistors different - 1k voltage dropping after the first filter cap, and 68k instead of 22k for negative feedback - and it adds a cathode voicing switch for the preamp. Otherwise it's a Tweed Princeton.
I successfully cloned the Varsity Deluxe and made a Tweed Princeton using P2P, so I decided I would convert the Tweed Princeton into a Two-Stroke so I could use an EL34, which I prefer. I swapped out the OT for a larger Heyboer, changed the resistors, added the 3 way switch. I fired up the amp using sacrificial tubes - JJ 5Y3, TungSol 5881 and 12AX7 - in case of phase mismatch between the OT and NFB. All was good. So, much to my surprise, when I swapped the 5881 for an EL34 I heard, instead of the sweet hum of a poorly grounded amp :) , a synthesizer-like sounding bwoo-eeeeep! (or Bweee-ooo) after about 18 seconds. This didn't at all sound like what I've heard with a 6v6 with an OT-NFB mismatch, but rather a synthesizer sweeping from high to low frequencies.
So I compared this P2P amp with my other Two-Stroke, the one using the Princeton chassis and Fender layout. The readings were nearly identical. And, that amp happily functions using an EL34. I used a different EL34 and disconnected the NFB - same result.
Why would the amp work perfectly fine using a 5881 and 6L6 but not an EL34, when the other layout functions with either? The components are identical so I am stumped. I power off the amp when I hear that hideous sound to avoid blowing a tube so I haven't attempted any measurements with the EL34. Anything else I should be looking at?
Measurements I took:
                   5f2a               P2p
AC supply      351      347
B+                390      396
B+2 (screen) 350               350
Plate             384      382
Bias                56mA            51mA
V1A plate      154      156
V1B plate      154      156

Thanks for any advice you can offer.

[HotBluePlates]

... 68k instead of 22k for negative feedback ...

Unsolder that feedback resistor, and see if your sonic problems remain.

[tubeswell]

Yes it’s probably feedback loop.


But you also jumpered Pins 1 and 8 on the EL34 socket?

[MagnaClone108]

Tubeswell,

This P2P layout makes use of pin 1 as an empty connector. Pin 6 of the preamp connects to pin 1 of the power tube, with a 22mfd coupling cap then going from pin 1 to pin 5 of the power tube. So that may be the issue that I will need to work around if I want to use an EL34 in this amp. FWIW pin 1 and 8 are not jumpered on the other layout I used and the EL34 works just fine. But I think I will add a terminal strip to free up pin 1 and jumper 1 & 8 and I'll post if that solved the problem.
 
HotBluePlates,

I disconnected the wire from the output jack to the 68k resistor with no effect. I thought NFB would affect all tubes I put in there not just EL34s. Am I correct?

[tubeswell]

Pin 1 is the suppressor grid on an EL34. Whereas in a 6L6, the suppressor is internally connected to the cathode (Pin 8), if you want the suppressor grid to work in an EL34, you need to do the connection externally.


If you’re using socket Pin 1 as a mounting point for something else, all bets are off when you plug an EL34 in.

[HotBluePlates]

... I thought NFB would affect all tubes I put in there not just EL34s. Am I correct?

You mentioned swapping the output transformer, so getting the phasing correct was a 50/50 proposition.  And positive-feedback doesn't create the same symptoms in all cases so the surest test is simply disconnecting feedback to find whether the problems stop.

However, Tubeswell made the correct first observation:

This P2P layout makes use of pin 1 as an empty connector. ... FWIW pin 1 and 8 are not jumpered on the other layout I used and the EL34 works just fine.

If you’re using socket Pin 1 as a mounting point for something else, all bets are off when you plug an EL34 in.

As Tubeswell said, you must connect the suppressor (Pin 1) to the Cathode (or to a negative bias voltage, but not the thing Pin 5 is connected to) for proper operation.

It's a complicated series of interactions to determine whether the "sorta-output" at the Suppressor will generate positive of negative feedback at the Input that is the control grid at Pin 5.  But it's always a bad idea.

It's a lot like saying someone got in a car blind-drunk and not wearing a seatbelt, but they made it home safely.  Occasionally getting away with it doesn't mean it's a Good Idea.

[MagnaClone108]

Happy to report that I modified the layout so I could jumper pins 1 and 8 and now the amp is working just fine. Thanks to everyone for their input!