Can someone explain "negative feedback" in amps?

[glassjaw7]

I've been reading a lot of different info on the topic, but don't know what's legit or what it all means/how it works...

I've read that an AC30 gets its unique tone from lack of negative feedback. I also know that Dave Friedman sweetened up my recto by changing the negative feedback. What does it all mean??? Thanks.

Also, if anyone can recommend any good reading on the topic, or just on amplifier electronics in general I'd appreciate it. Not trying to be the next amp builder, just want a better understanding.

 

[wolfeman28]

Merlin Blencowe explains it in an easy to understand way in his book, and also a short snippet here.
http://www.freewebs.com/valvewizard/localfeedback.html

This is a good read as well.
http://www.aikenamps.com/NegativeFeedback.htm

Kevin O'Connor's books go into the subject in depth as well, with graphs that show what it does to the signal/frequency response.

 

[jerrydyer]

its when that bitch doesnt talk back

 

[JakeAC5253]

You blend two in phase signals and they sum each other, you blend two out of phase signals and they cancel each other. Negative feedback in amp design has to do with selecting certain frequency ranges to send back (or forward) to cancel with input or output signals so as to help carve the final output tone. It is called negative because you are utilizing out of phase signals. Positive feedback is the combining of in phase signals to carve tone.

 

[ttosh]

I thought negative feedback had a lot more to do with feel than tone?

 

[glassjaw7]

Thanks for the replies!

 

[wolfeman28]

I thought negative feedback had a lot more to do with feel than tone?

It can do both. You get a dampening effect, and frequency shaping. Think 'depth' and 'presence' controls in the 5150's/SLO/etc power section. These are actually controlling the negative feedback circuit of the power amp.

 

[steve_k]

Think of it as your amp's own recycling program, dumping unused signal at the output level back into the mix.

Steve

 

[glassjaw7]

Wow, I need to read a LOT more to even begin to understand all this...appreciate the help though guys.

 

[glpg80]

You blend two in phase signals and they sum each other, you blend two out of phase signals and they cancel each other. Negative feedback in amp design has to do with selecting certain frequency ranges to send back (or forward) to cancel with input or output signals so as to help carve the final output tone. It is called negative because you are utilizing out of phase signals. Positive feedback is the combining of in phase signals to carve tone.

note entirely. positive feedback is used in tank circuits with a unity gain of positive 1. negative feedback with stages enclosed in 180* in-phase signals will do nothing but free run oscillate and do so uncontrollably.

 

[petejt]

You blend two in phase signals and they sum each other, you blend two out of phase signals and they cancel each other. Negative feedback in amp design has to do with selecting certain frequency ranges to send back (or forward) to cancel with input or output signals so as to help carve the final output tone.

Which frequency ranges in particular would be selected? Does the Recto have a unique frequency that is selected?

And are there any amps that have a variable control to change which frequency ranges are selected?

 

[rockstah]

in Marshall amps signal is taken from the output transformer and injected back into the circuit. whats it all mean in application? with more neg feedback one gets a smoothed tone. with less one gets a more hairy, aggressive tone as well as more and perceived gain.

this is what we see as the presence circuit. we can effect the freqs with the cap on the presence pot. by changing which tap we are taking neg feedback from the OT as well as the resistor used that feeds the presence pot we effect how much or how little NFB we want.

Mark

 

[ConcreteVampire]

in Marshall amps signal is taken from the output transformer and injected back into the circuit. whats it all mean in application? with more neg feedback one gets a smoothed tone. with less one gets a more hairy, aggressive tone as well as more and perceived gain.

this is what we see as the presence circuit. we can effect the freqs with the cap on the presence pot. by changing which tap we are taking neg feedback from the OT as well as the resistor used that feeds the presence pot we effect how much or how little NFB we want.

Mark

Excellent answer! For Marshalls, I concur, but what's going on in Soldano's? I "hear" what you say with a Marshall, It's that sizzle/aggro thing...
in Soldano's it's even gronkier! more burly an effect if you will.... (Since we won't find Gronky in the dictionary)

 

[JerryP]

There are many ways to use negative feedback, each stage can have it's own loop. The very basic explanation is taking a signal that's out of phase with the original signal and injecting it back into the signal path. Opposites cancel each other out. By using filters to control which frequencies are sent back you can control what gets cancelled and what doesn't.
On the Mesa Recitifiers when you select the modern mode it disconnects the feedback circuit which is why if you add a common depth mod control the the Mesa it doesn't work when in the modern mode.
Jerry

 

[JakeAC5253]

You blend two in phase signals and they sum each other, you blend two out of phase signals and they cancel each other. Negative feedback in amp design has to do with selecting certain frequency ranges to send back (or forward) to cancel with input or output signals so as to help carve the final output tone.

Which frequency ranges in particular would be selected? Does the Recto have a unique frequency that is selected?

And are there any amps that have a variable control to change which frequency ranges are selected?

For the Rectifiers, the voicing has to do with the preamp. From what I remember, the third stage bleeds a significant amount of high end, and afterwards a smaller boost to upper highs. Thats how you get that sound with less high mids and more sizzle.

Resonance and presence controls act as controllable negative feedback loops. The presence controls what highs are sent back, and the resonance controls what lows are sent back. The 10pF 5150 cap mod is an example of negative feedback as used in a preamp voicing. Basically some of the very high end content is being shunted back instead of being allowed to pass to output.

 

[petejt]

Thanks Jake.

 

[srinivassa]

Man, it's a mouthful. The thing about phase is definitely thinking on the right track. Positive feedback adds a signal to itself, so that it grows and grows and grows in a loop, like feeding speaker output back into a microphone. Negative feedback does the opposite. It subtracts a signal from itself, making the signal more stable and reducing certain characteristics of the original. Note that in both cases, the signal being added or subtracted is a much scaled down version of the original. That's why you don't blow your speaker every time you get positive feedback. That's also the reason negative feedback doesn't completely kill a signal.

As far as in amps? I don't see much use of either. Negative feedback would be possibly helpful. A lot of amplifier applications use it. Typically guitar amps do not, at least not in the signal path. Buffering up an effects loop is something that typically uses negative feedback in order to promote stability and create a high impedance.

I can't think of an example I've seen of negative feedback being used in a tube amp except buffers for a spring reverb or an effects loop. It could be done though. You would sacrifice a bit of overall gain, and you could use the feedback to control the overall sound. Feeding back high treble (negatively), for example, would keep the attack sounds reigned in. There is no wrong or right way to do it. I'm pretty sure most amps don't use it though.

You blend two in phase signals and they sum each other, you blend two out of phase signals and they cancel each other. Negative feedback in amp design has to do with selecting certain frequency ranges to send back (or forward) to cancel with input or output signals so as to help carve the final output tone.

Which frequency ranges in particular would be selected? Does the Recto have a unique frequency that is selected?

And are there any amps that have a variable control to change which frequency ranges are selected?

For the Rectifiers, the voicing has to do with the preamp. From what I remember, the third stage bleeds a significant amount of high end, and afterwards a smaller boost to upper highs. Thats how you get that sound with less high mids and more sizzle.

Resonance and presence controls act as controllable negative feedback loops. The presence controls what highs are sent back, and the resonance controls what lows are sent back. The 10pF 5150 cap mod is an example of negative feedback as used in a preamp voicing. Basically some of the very high end content is being shunted back instead of being allowed to pass to output.

Ooopps, sorry Jake. Didn't read your whole post. I guess that is probably true about the presence and resonance. That probably does end up feeding back into the phase inverter tube on a lot of amps. My bad.