Negative feedback - How much?

[diagrammatiks]

Also the amount of actual phase shift is frequency dependent. That's why Merlin suggests slugging dominant poles and grid to ground caps on the power tubes to minimize the highest frequencies.

[rdjones]

Also the amount of actual phase shift is frequency dependent. That's why Merlin suggests slugging dominant poles and grid to ground caps on the power tubes to minimize the highest frequencies.

There seems to be some ambiguity between phase/phase shift and polarity. These are two different things.
Phase and shift thereof is dependent on time and frequency.
Polarity does not depend on time or frequency and is either inverted or not.
Quite often an inverted polarity signal is said to be "out of phase", and although the effect is similar, the statement is not accurate.
An inverted polarity signal is so-called "out of phase" at all frequencies equally.
A signal that is delayed in time so as to cause a true phase shift will be differing amounts (degrees) of phase shift, ie if 1kHz signal is out 180, it will be 360 out at 2kHz. This is what causes "comb filtering".

rd

[Cliff Schecht]

Great point RD, I've even been corrected on this before during a big presentation and try to not mix up the two anymore. It's important to understand the difference too because both methods can be used to achieve an inverted version of an input signal but only one is time dependent.

Also from my experience it seems a lot of people who haven't worked with phase a lot don't really have a good understanding of it. Group delay is an even more foreign concept to many but is becoming more and more applicable in modern sampling systems. It's one of those things that people have heard before but don't really know how to describe it (or explain it) usually..

[rdjones]

To add to the confusion between phase and polarity is the fact that many British consoles refer to the polarity invert function with the phase symbol. Ø
And the phase inverter is really a differential output with opposite polarities rather than any real phase shift (hopefully).
There is probably some small amount of actual phase shift coming back on the NFB loop but for the most part it's negligible at guitar frequencies.

Some of the audio pros get a bit picky about using the correct expression, but the pervasive use of the terminology has become widely accepted for better or worse.
I suppose it's easier to say "out of phase" than "inverted polarity" and I'm probably as guilty as the next guy.

reddog

[martin manning]

And what to most of us call that thing that creates the two signals for the PP stage? A phase inverter, evidently a mixture of the two definitions.

[tubeswell]

After reading Merlin's chapter on power transformers in the 2nd book, I can see why there is a phase lag between the primary and the core and another one between the core and the secondary (so that the secondary is either 180 'out-of-phase' or in 'phase' with the primary - depending on which pole on the secondary that you take the ground return from - at least that is clearer to me now). I guess in an OT, the amount of phase lag at any given frequency will depend on the amount of the iron in the core, and the length and diameter of the wire in the windings? (This must be simpler to work out in a PT because the input frequency would be more or less constant.)

[Ian444]

Have any amp builders here actually found something positive or beneficial to the sound of the amp by using negative feedback?

You need more gain in the preamp and then lose it again with the negative feedback, in order to make the amp more linear, reduce output impedance, and um, something else...hmmm, sounds like a hifi thing. How unlinear is any well-designed amp? Very linear IME. I just don't get it. In hifi, every amp I built sounded more sterile, constricted, soul-less, colder, and solid-state with negative feedback.

The only benefit I found with negative feedback is some guitar amps break up in a way that is not so smooth as they get driven into distortion. With negative feedback, that extra gain that was swallowed up is regurgitated as the outputs start to saturate, launching the amp through that rough break-up zone. Some amps just break up smoothly and don't need it.

Just my 2c real world as I see it.

Zippy I know you did this just to stir us all up!

[Jana]

"Have any amp builders here actually found something positive or beneficial to the sound of the amp by using negative feedback? "

Yes. There is so much that can be done with the NFB loop beyond the cold, sterile and soul-less. NFB doesn't have to be about making the amp flat and HiFi.

[rdjones]

Have any amp builders here actually found something positive or beneficial to the sound of the amp by using negative feedback?

I like having a presence control ...

reddog

[angelodp]

Hi, Just finished the rj Supre amp and its terrific SE tone. I tried this idea for the NFB and it is subtle and seem to act as a presence control more than anything.

Best Ange

[IMG1642]http://i260.photobucket.com/albums/ii9/ ... es-1-1.jpg[/img]

[Zippy]

Zippy I know you did this just to stir us all up!

Maybe so, Ian, but I'm feeling a lil' like the Sorcerer's Apprentice as things started to get out of control!

Thanks, Ange, for giving my idea of the variable NFB control a whirl - I'll be using it on my Supre Tweed SE.

[angelodp]

I tried a 1m pot instead and it does work pretty well. It does bring on some more Drive and raunch. I have to dial in about 50% of the pot to get things going. It seems a to be more tweedy to my ears. I am running a 6L6 on this and have a 600v/stndby/480v switch. 600v is very nice indeed.

This version also shows a possible idea with the DPDT switch and an added fixed 1M

[IMG1645]http://i260.photobucket.com/albums/ii9/ ... es-3-2.jpg[/img]

[Zippy]

Thanks for running with this idea, Ange.

What is the minimum value of resistance that you find yourself using?

Ever turn that pot to 0? 3?

[angelodp]

The pot has little effect until its at about 50% ( audio 1M ). Fully cranked ( no NFB ) is my favorite.

[Zippy]

The pot has little effect until its at about 50% ( audio 1M ). Fully cranked ( no NFB ) is my favorite.

NFB is the way I tend to go with Fender blackface as long as they stay stable.

Your comment regarding ZZ Top got me thinking this morning of my favorite cranked tones - Buddy Guy through a tweed Bassman works for me!