TW circuit NFB problems

[guitardude57]

Howdy,

I have a chassis that I use for experimental circuits. The last was a Plexi type that worked well. Use a switched presence pot to turn off NFB to see/hear effects of in/out circuit. The Trannys are from old tube integrated amp with 8K OT and big PT with volt doubler out of neccessity from original design. 7189 output

B+ volts

1 455V
2 405V
3 336V
4 315V
5 303V

Screen resistors bring volt to acceptable ranges. Of course the OT sees the above voltages. No red plate....strong audio output, sounds good and fairly stable. Even when cranked.

I redid the circuit to Express type. Still sounds wunnerful, but different, do to what the circuit is.

I unhooked the NFB lead to switch/presence pot, and the presence pot sweeps (you can hear it with no high frequency sweep.....no nfb)
When NFB hooks up, 400-500 hz oscillation occurs, that varies with the turn of the pot.

Checked values and continuity, everything looks good.

I know the NFB has inherent instability issues in this design.

What can I try?

Thanks

[Zippy]

Move the NFB to the other side of the OT.

[guitardude57]

Move nfb to other side of the OT?

It worked fine in other circuit (Marshall 1986 circuit) which is identical to what is there now except rather than 47K NFB resistor there is a 100K on the TW circuit now.

All of the PI on back to the output, is the same, with the exception of 82K/100K of PI are the opposite ....... V3a and b are reversed on the TW circuit compared to the Marshall circuit.

I realize this after start up....heard the oscillation, then reversed the PI a/b wiring and components......oscillation still there.

That really should not have made any difference anyway, since in either case things were in correct phase....... the 82K on 3 a or b side of the triode. I reversed them anyway to be sure. No difference.

The NFB is wired to the secondary 8R tap like it is supposed to be. Also tried another secondary tap (16R), it changed the oscillation frequency up one octave.

What do ya think?

[mlp-mx6]

If the phase of the current amp circuit is 180 degrees off from the earlier circuit that worked, then you're now providing positive feedback.

[guitardude57]

Thanks,

In other builds if there was a reversed primary it would let you know on turn on. Since the presence pot is a turn switch-on, and I fired it up with it off.....no squeal until I turned on to adjust presence.

That must be it and an easy fix.
Warming up the iron now.............thanx again.

I will have to take pix of this thing. All point to point with no boards. On a steel chassis..........and the bugger is stable...........and should be more quiet of hum when I get the presence up to snuff.

With 2 X 7189 output is clear, loud and beautiful clean.........until you crank the guitar volume.........wow!

The next experiment will be with 2 X 6L6GC for output on this.

[skyboltone]

Generally I'd go for the positive feedback answer first thing too. But this is strange. Usually PF is not 500hz and then double at higher ohms right? The fact that it's semi stable is interesting too. Have you chopstuck your NFB wire. It's pretty sensitive to surrounding noise makers although where something could be feeding back in octaves beats me. There's a pretty stable tank circuit being created here somehow. Curious.
Dan

[guitardude57]

OK,

I have the NFB ironed out. Reversed primaries.

Pretty darn quiet now.

Just have some dimed hf oscillations to deal with now. The usual treble and presence whistles.

Tomorrow, (damn, it is tomorrow....) when I get up, will go at it again with the express build guide, and swap some wire locations and get those whistles nailed down.

Been a fun build. When I get a chance to take a pic or 2, will put them up. You won't believe how this was put together!

The next one I build will use an aluminum chassis, and the conventional board to mount the parts. If this steel box is this quiet....I will have no problems the next round!
Thanx for the help guys!