MOSFET Vs. ZERO FEEDBACK

At mid seventies was developed in Europe a concept about the advantage of low global feedback amplification. This concept said that feedback amplifier has problems because the feedback signal has a delay to return to input and then some phenomena are present in detriment of sound quality and naturally.

Today the discussion continues but restricted to the feedback amount. 

Excellent results can achieved with a synergic relationship between local feedback and global feedback.

As explanation of that let us explore the concept "delay" between output and input. Electromagnetic energy flows inside a conductor at 230.000 km/s. This mean that a cable of 23cm long (distance comparable to amplifiers dimensions) is a magnetic wave in 1 nanosecond , this is a period of a 1GHz signal. A 20.000 Hz signal has a length inside a conductor of 11.5 Km or 11.500 metros. The concept delay is not applicable to an audio amplifier, try to put a 11.500 m cable inside your amplifier.

When in an audio paper about feedback or damping factor you read "delay" you really must read "phase". Phase is inherent to self reactance of input, drivers and output devices and the transfer characteristics of local and global feedback circuits.

You can't evaluate seriously how many feedback is appropriate if you do not analyze the topology and devices involved in the design.

 

History:

The problems with amplifiers in the early '70 was: the design criteria and the semiconductors devices.

In those years an audio amplifier was like operational amplifier, this mean: high gain and a dominant pole, generally in the input stage, always stable in any load condition in the close loop situation. We see in many designs of that era an integrated operational amplifier, like UA741, in the input stage, in order to use the differential pair and the dominant pole.

We know that a dominant pole at low frequency, generally at 10Hz, has two problems: one the slew-rate and TIM and second, the degradation of damping factor by the fall of loop gain at mid frequencys.

To correct understanding of dynamics process like Slew-Rate and TIM you must know that these phenomena are independent of feedback because the feedback are cut under these situations (see*)

In order to have a solution to these problems is modified the design criteria. Today the audio amplifier has no dominant pole and the power bandwidth is greater than the small signal bandwidth.

To acquire these criteria you need two elements: first differential input stage pair with discrete components, emitter degenerator resistors and appropriate bias current, and second very fast drivers and output devices, like MOSFET's, to move the non dominant input stage pole to approximately 15000 Hz..

With these two design criteria, an audio amplifier with multiple feedback, discrete components, drivers and output MOSFET devices, is the best option as output power audio amplifier today.

*A complete analysis you can see in Gray & Meyer "Analysis and Design of Analog integrated Circuits" John Wiley and Sons. 1977 pag. 541 y siguientes.

 

Contact Us  info@vn-amps.com.ar

BACK