OT Primaries "backward" in a Push-pull amp...

[Paul-in-KC]

All,

I'm sure this must be a stupid question - so, just to prove I stupid, I'll ask.

I recently completed my first amp build. It's a Marshall 2204 HW clone from the Ceriatone "kit" - with Mercury Magnetics xformers and choke.

It sounds fantastic - but when I first started testing with AC signal input, I got the dreaded "honky squeal" that one gets if the OT primaries are reversed (at least on a 2204).

I was relieved to find out that there was a "wrong" way to connect the OT primaries, because I knew that I didn't think it would matter when I was making the original connections - and I got it backward.

My question is, why is there a wrong way?

My conceptual understanding is that you have your "finished" preamp signal ready to be amplified - that is input in to the PI circuit which creates a second signal that is a mirror image of the original and those two signals are fed into the power tubes and the output from there is fed in to the two OT primaries. One side "pushing" and the other side "pulling".

I obviously have some details wrong/missing, because there it matters which input is on which side of the OT.

Can someone explain what I am missing?

Thanks.
-Paul
(sorry for the dumb question)

[harleyboy2112]

What happens in a push pull amp with negative feedback is you turn that negative feedback into positive feedback if you hook up you OT wires wrong. In amps with no negative feedback (tweed deluxe for example) it doesn't matter.

[rdjones]

Your negative feedback became positive feedback when the signal is "180 out".
The same remedy can be accomplished by swapping the leads to the output tube's grids.
If you had a single impedance secondary on your OT, (no multiple taps) you could also flip the secondary connections.

reddog

[Paul-in-KC]

Harley Boy and Red Dog - thanks fellas!

Now I need to absorb the "negative feedback" concept.

I knew there was something that I was missing - and it makes sense now that you say that - but the reason that I missed that point is that I'm not solid on the NFB thing.

I'm a newbie - but I'm a quick study.

Thanks again.

Prepare yourselves for more "newbie" questions.

-Paul

[diagrammatiks]

It's not so much as a wrong way and a right way...
It's just the nature of the transformer that one lead is in phase and the other lead is out of phase.

in actuality you don't need to switch the transformer windings.

You can switch the coupling cap input or output leads so that they are going to a the other set of tubes.

[DonMoose]

Now I need to absorb the "negative feedback" concept.

It may help to seek out an animation of an op-amp circuit (too tired right now).

Negative feedback in an amplifier helps to linearize its response to an input and, sort of, tell it it has swung far enough in that direction (push or pull) to make the voltage-dividered output voltage exactly cancel the input voltage. That's why the NFB goes to the 'other' PI input (you can think of the 2 grids as + and - inputs to an op-amp).

So if you swap the 'push' and 'pull' wires on the OT primary, you have positive feedback, which has your output slamming from rail to rail as hard as it can.

I'm sure it can be explained more gracefully, and certainly more exactly, but this is at least a workable model.

[Paul-in-KC]

Don (and everyone),

Thanks again for taking the time to share your knowledge.

It is appreciated.

-Paul

[rdjones]

Tidbits of NFB :

The stages of a typical push-pull tube amp that have the most nonlinearity are the phase inverter, the output tubes and the output transformer.
This is why a small amount of signal applied in reverse polarity from the secondary to the "bottom" of the phase inverter tail will tend to flatten frequency response and reduce distortion as it reduces the overall gain of the stage(s).
The level fed back is small compared to the main signal, usually -15dB or less.

The Marshall and other amps that utilize a presence control will vary the amount of high frequency content that's fed back to change the character of the high end response.

rd

[VacuumVoodoo]

You drive a car so that you keep constant distance to the car in front of you.
You get too close - ease on the accelerator
You stay too far back - step on the accelerator
That's negative feedback
Now rewire your brain so that you step on accelerator when you get too close, the closer you get the harder you step on it.
That's positive feedback.
It always leads to disastrous results, unless you're designing an oscillator.

[rsalinger]

You drive a car so that you keep constant distance to the car in front of you.
You get too close - ease on the accelerator
You stay too far back - step on the accelerator
That's negative feedback
Now rewire your brain so that you step on accelerator when you get too close, the closer you get the harder you step on it.
That's positive feedback.
It always leads to disastrous results, unless you're designing an oscillator.

I had never really heard it explained like that. Thank you VacuumVoodoo. That made sense.