First cycle distortion - Graham, what is that?

'Linear distortion' is defined in some textbooks. For our purposes it could be seen as a stationary (unmoving over time) undesirable change of frequency response and or phase response of an audio signal as it passes through audio reproduction device. It is a reproduction error in that case because it is deemed to be unwanted. In other words, it is a linear distortion because it does not produce any new frequencies, but it may sound objectionable nonetheless.
 
Last edited:
I have already given an example from my practice when using an equalizer that introduces only linear distortion, they tried to achieve a good sound. With the help of the equalizer, only linear distortions were introduced, which they tried to correct other linear distortions and really got a linear frequency response, but the sound was disgusting. If, in your opinion, linear distortion is so harmless and does not introduce additional harmonics, then what is the matter? According to my concepts, linear distortions can only be talked about in a steady state, for example, at the 1000th period when all transient processes have ended, then when the concept of "linearity" comes into force - where output *continuously* varies in *direct proportion* to input.
 
In physical reality, it has been argued that nothing is completely linear nor completely stationary. But something can be so close to both of those things that for many purposes the assumption of elegant mathematical perfection works fine. Maybe not so much for other purposes. Of course, people can and do argue about when mathematical perfection should be the operating assumption.
 
Last edited:
Graham wrote:
“Of course many have professed that I lack an understanding of everything related to thd or a suddenly starting sinewave etc., also that no such wave exists, but all they actually cover are obvious aspects that were once my own stepping stones whilst studying the never ending stream of first cycles that make 'music'.
Amplifier that shows itself better capable of coping with a suddenly starting 10kHz sine that cannot exist in real life is also going to be more linear when coping with the dynamically changing cycles that do.
Many solid-state amplifier designers would get a surprise if they set thd performance aside for just a moment and studied the first cycle (dynamic linearity) behaviour and output characteristics of their own creations.
I suggest that denying the use of first cycle observation is to deny insight into amplifier circuit operation.”
Graham has repeatedly cited examples from his observations that amplifiers that introduce less distortion in the first period of a 10 kHz signal showed their best side in listening audio tests.
From what we call the distortions arising during the period of transient processes, they do not become more linear. Indeed, distortions introduced by linear reactive components do not depend on the signal amplitude, but this does not diminish their contribution to distortion in transient processes. And if meters of nonlinear distortion (Audio Precision) can not record the harmonic contribution of these distortions, then this does not mean that they do not exist.
 
petr_2009
What phase margins are you achieving in your designs and is an output inductor implemented?
The phase margin must be at least 30 degrees (optimally 60). I try to stick to 60 + -10 degrees.
As for the use of the output inductor, John Curl wrote about its use that it does more harm than good. Therefore, it is desirable to develop such amplifier designs that do not need it.
Graham wrote that current feedback amplifiers have a resistive output impedance and are therefore less likely to need an output inductor.
You can find designs with inductors 5 ... 15 μH. I believe that if the amplifier is not stable in tests for a reactive load of 0.1 μF; 0.5 μF and 2 μF, then it is necessary to select the inductance of the smallest possible value ensuring stability in operation. Often only 0.2 ... 0.5 μH is sufficient. Oddly enough, but often the most critical load is 0.1 μF capacitance.

Originally Posted by billshurv
No, you have given an anecdote
Judging by your posts, your lexicon is limited to just one word anecdote
 
Assertions are being made that a certain quality level can only be reached with extended power bandwidth. So which is the better of amplifiers A or B?

Amplifier A - Phase margin 120 degrees, gain margin 27db, no output inductor, FPBW 80KHz.

Amplfier B - Phase margin 60 degrees, gain margin 14dB, output inductor, FPBW 500KHz
 
AX tech editor
Joined 2002
Paid Member
It's not even a theory. He's just trolling. Remember he said there is no such thing as linear distortion? Then someone explains to him what linear distortion is. Then he goes, yeah, but in my concepts, linear distortion ...

He just makes up stuff as he goes along to keep the thread alive. Impossible to have an intelligent discussion this way.

Jan
 
I bring up the examples of amplifiers A and B.
Amplifier B with PM 60 degrees will exhibit overshoot with a step response which will be further distorted by the inductor. This overshoot will be proportional to the amount of negative feedback.

Amplifier A will have no such worries.

Graham made observations that were for him true at the time but as far as can be determined there was no engineering reference. Hence its anecdotal nature. Graham favored JLH 10W because of its FCD behavior. A reviewer built versions of the JLH and noted that the output capacitor affected sound quality and preferred a version without it. He commented that the sound quality was very good, but added, his valve amps were better. Valve amps are not noted for their wide power bandwidth.

We can add anecdote after anecdote but the behavior of amplifiers is described with phase and gain margin. Negative feedback is a servo system. Following servo system rules, there is no need for FCD or TIM.

Walt Jung and a colleague studied TIM in real equipment. Paraphrasing their conclusions:
You can drive a car with flat tires and it will have bad road handling. This is true. But why would you want to do that?
 
On the eve of the collapse of the Soviet Union (USSR), one of the research institutes conducted research on the effect of speed distortions on sound quality.
Was found:
“If the speed distortion is very small (less than 0.01%), there is no statistically discernible effect of the amplifier. Just neutral, detailed, but without traces of harshness, and, what is important, reproduced regardless of the topology and amplifier design. That is, amplifiers of different circuitry and ideology, which fit into a real signal of 0.01%, are almost impossible to distinguish, as well as to notice any defects that could reliably be attributed to the amplifier. "
After the collapse of the USSR, funding and research were stopped.
I have already shown that on a sinusoidal signal this kind of distortion appears at the beginning of the first period. Tests show a direct relationship between speed distortion and signal propagation delay time. To determine what level of delay is necessary to ensure speed distortion of less than 0.01%, I synthesized the signal using the sum of two ideal sinusoidal signals, one of which is delayed by 200 ns. The total signal is delayed by 100 ns, and at the beginning of the first period, signal distortion is formed, similar to the distortion that occurs in the signal comparison unit of a real amplifier. As shown by the result of measuring harmonic distortion, it is with a delay of 100 ns at a frequency of 20 kHz that the level of distortion is 0.01%. Considering that in real amplifiers, in addition to high-speed distortions, there are also nonlinear distortions, it makes sense to reduce the level of permissible high-speed distortions by 1.5 ... 2 times (up to 50 ... 70 ns).
unfortunately there is no one here to explain anything, the attitude to my posts is the same as to Graham's posts
 

Attachments

  • Test_20kHz_tPD-100n_speed-error.png
    Test_20kHz_tPD-100n_speed-error.png
    37.6 KB · Views: 146
  • Test_20kHz_tPD-100n_extended.png
    Test_20kHz_tPD-100n_extended.png
    28.3 KB · Views: 141
  • Test_20kHz_tPD-100n_THD.png
    Test_20kHz_tPD-100n_THD.png
    18.6 KB · Views: 136
Last edited: