tube amplifier

Negative feedback

Negative feedback

No, I’m not talking about feedback (not even positive) feedback on the articles on this site. These are always welcome, both positive and negative.

This is about feeding back a small part of the amplified signal to the input of the amplifier.

Types of feedback

positive feedback
positive feedback without phase shift


There are two ways of feedback, the positive and the negative feedback. In the positive, the part of the signal that we feedback is in phase with the original signal.

The output signal is thereby amplified. A negative consequence is that the distortion increases and the amplifier can become unstable. That is why this form of feedback is almost never used in audio amplifiers.

negative feedback
negative feedback without phase shift

With negative feedback, the returned signal is in opposite phase. This will reduce the gain but also ensures that the amplifier becomes more stable and the input impedance increases and output impedance decreases.

So with negative feedback we can adapt the input impedance to the device that delivers the signal (DAC, CD or record player).

All this may sound a bit difficult for non-technical people, but in fact we personally are constantly giving feedback or receiving feedback. You take action, see what the result is and adjusts. Most of the arrangements in a factory also work in that way for the simple reason it is the only way a process can be controlled.

Negative feedback also ensures a reduction of the “noise”. Reading the different articles concerning feedback, only will increase your uncertainty about it. The distortion (distortion) becomes lower, but equally the amount of distortion on the higher harmonics increases.

We could therefore say that as a result of negative feedback the distortion decreases, but at the same time the complexity of the distortion increases.

 

What is the source of this distortion?

Distortion occurs when the gain is not constant under all conditions. If a device had a perfectly constant gain under all circumstances, there would be no distortion.

With an amplifier that has a varying gain depending on the signal, the signal that comes out will have a different shape than the input signal.

The amplification of an amplifier stage is usually not straight but curved or nonlinear. This nonlinearity is caused by several factors:

  • The changing current through the amplifier.
  • The changing voltage across the amplifier
  • The changing temperature of the amplifier.

In all cases in this document, when I speak about amplifier I mean the element that amplifies: a transistor, fet, op-amp or tube.

It is important to realize at every amplifier stage that the signal always changes with respect to the original if this stage is somewhat non-linear.

This signal is distorted forever and even though you use different techniques to reduce the distortion after the creation, you will never get the original signal back. So best is to avoid distortion from start on. Reducing the amount of amplifier stages is a means of avoiding excessive distortion.

If a lot of negative feedback is applied, there is often a need for more reinforcement and often one or more stages are added. The result is even more distortion of the signal because no single stage is fully linearly amplified.

Because this concerns the distortion of the higher harmonic frequencies, you can speak of a non-linear distortion. If you try to measure this with a distortion analyzer, you will not even measure them in most cases. That is why most manufacturers can also submit the most beautiful figures.

To make matters worse, we will use this non-linear distorted signal to return to the input of the amplifier.

There are people who swear by negative feedback and on the other hand there are people who will avoid it at all times because it takes all “life” out of the music.

I have opted to use no negative feedback in the amplifier stages of the Flea-power. I say this because feedback is used in the tubes regulated power supply to stabilize the output voltage.

Not using negative feedback was a decision based on listening and not on measurements. I am aware that I am against some people’s opinion, but as much as I respect someone’s preference for feedback, I hope to get that respect from the proponents of feedback.

In this week I will try to place some pictures for clarification in the article.

Leave a Comment

Het e-mailadres wordt niet gepubliceerd. Verplichte velden zijn gemarkeerd met *

12 + twenty =

This site uses Akismet to reduce spam. Learn how your comment data is processed.