OTs and Global NFBs

[defaced]

When cloning an amp, how important is the OT in the behavior and setup of the NFB? Meaning, say I take a salvaged OT and copy the NFB from another amp, should I expect the same behavior from the presence and depth controls?

I ask because of the parts I have control of when doing a clone, the OT is the one I have the least amount of control over, and to me, the NFB controls are typically very strong controls on the amps I use, so cloning their behavior properly is important to me.

[Chuck H]

All things being equal with the same primary impedance and secondary tap ohm connection, the NFB loop and any associated circuits, like prsence and resonance controls, should behave the same or at least predictably similar to an amp you copy.

Chuck

[Steve Conner]

The bandwidth of the OT has a big influence on the NFB, IMO. If you took a hi-fi tube amp and put a guitar OT in, it would go unstable and burn up from ultrasonic oscillations.

Swapping one guitar OT for another shouldn't have as much of an effect, as they're all more or less equally crappy, but I expect they will all add their own flavor to the mix. As long as it's stable with the presence and resonance turned to minimum, it should be OK at all other settings.

[Russ]

The bandwidth of the OT has a big influence on the NFB, IMO. If you took a hi-fi tube amp and put a guitar OT in, it would go unstable and burn up from ultrasonic oscillations.

At first I thought you had it backwards and you should've said: if you put a hifi OT in a guitar amp, it would go unstable and burn up from ultrasonic oscillations. But since it's NegativeFB, you're saying that the guitar OT's lack of bandwidth limits the frequency range of the NFB and therefore allows the hifi amp to go into oscillation?

If that's correct, then does that mean that a hifi OT in a guitar amp would tend to limit ultrasonic oscillations?

[defaced]

All things being equal with the same primary impedance and secondary tap ohm connection

That's kinda what I was thinking. Being that an OT has a fixed inpututput ratio, as long as it were maintained in the clone, most of the NFB properties would be maintained as well.

[Steve Conner]

At first I thought you had it backwards and you should've said: if you put a hifi OT in a guitar amp, it would go unstable and burn up from ultrasonic oscillations. But since it's NegativeFB, you're saying that the guitar OT's lack of bandwidth limits the frequency range of the NFB and therefore allows the hifi amp to go into oscillation?

If that's correct, then does that mean that a hifi OT in a guitar amp would tend to limit ultrasonic oscillations?

Yes. It's not really possible to discuss this in great detail without busting open nasty cans of mathematical worms like the Nyquist stability criterion:
http://en.wikipedia.org/wiki/Stability_criterion

If you don't want to get your hands dirty, just remember that when the total phase shift in the NFB loop adds up to 180 degrees, negative feedback becomes positive, and with sufficient NFB an amplifier becomes an oscillator. Guitar OTs have more phase shift at the high and low ends of the spectrum than hi-fi grade ones, so they can't tolerate as much NFB.

(This is in fact why hi-fi OTs are designed for lower phase shift: so that more NFB can be applied around them before the screaming and motorboating begins. The Williamson hi-fi amp has one more gain stage in the NFB loop than a typical guitar amp with a LTPI, and Williamson specified a special OT with lots of interleaving in his original article.)

A hi-fi OT in a guitar amp will limit oscillations sure enough, but the main audible effect will be deep, clean bass instead of dirty farting noises. The last amp I did an OT upgrade on got enough extra bass to bash the speaker cone off the grill, and I had to put the smaller OT back in.

When it comes to presence controls and the like, they work like tone controls, except being in the NFB loop their effect is mirrored. Turning the presence knob clockwise reduces the NFB loop gain at high frequencies, so the power amp's gain increases.