In fact, it turns out to be a "square circle". Variable CCS - variable constant current source))
That is, real measurements on professional equipment lie and are not reliable, and the results of simulations with simplified models of elements are more accurate and have priority. Do I understand your point of view correctly?
It seems you have learned something, to show your speed errors, you should limit the BW of your signals with just a few cycles to the BW where the GD is flat.
For the 250Khz BW amp shown here having a flat GD of 620nsec, this signal should be limited to 20Khz max.
The 100nsec that you mention is 5 to 6 times faster, only possible with a 1.5Mhz BW amp, completely insane and over the top.
I propose as an exercise to use this simple amplifier and show all your pictures with vector error, speed error etc. etc.
Then we can analyse, because a linear filter process has no distortion.
And please don't start with an another subject, just do this for a change.
Hans
.
Attachments
The THD meter shows nonsense. There is absolutely no correlation with sound quality. This was known more than 70 years ago. It's time to know. Hence, they began to invent new testing methods: IMD, DIM-100, multi-tone, etc.
In this regard, the respected Lynn Olson wrote the following in one of his works:
“In electronics, the correlation between distortion percentages and subjective sound ratings is almost zero. The lower order harmonics are almost inaudible compared to the higher ones, although they dominate the THD measurement figures! The meter arrow tells the developer complete nonsense.
It's time to take a break from the myth of "euphonious distortion" and find those real subtle sources that create distortion in the amplifier that the human ear detects. Once we find measurement methods that actually help, rather than hide the truth from us, it becomes easier to design listener-friendly electronics. "
That is why David Hafler tried to propose the differential method at a new level in the form of SWDT, while recognizing that this method was one of the first.
Hans, I've posted more than enough amp models to support my point. If you understand the test charts (not like Banksy), you should easily understand me. If you don't understand, then I will not help you with anything, like many others ...
Last edited:
I apologize in advance, but who is Lynn Olson and why is he respected? I have not met this name and I do not know about his works. So far, I see words that are usually written in advertising articles, and not in scientific publications, and this does not cause respect. It's like saying that temperature measurement is a useless activity and does not give us an idea of the patient's health status. But after all, no one expects an effect from measuring temperature, as from contrast tomography, right? This is just one of the methods used to assess the condition, and if there is a temperature, then this is already a reason to think about more serious studies.
However, the question I asked you was about something else. And I never received an answer, only a reference to a certain third person and his opinion concerning the issue very indirectly.
This is a false statement. You did not show models, but only the results in the form of pictures, which makes it difficult to repeat your results and makes their verification and refutation either difficult and time-consuming, or impossible. All requests to stop at a particular scheme and conduct a strictly scientific discussion are not accepted by you, which makes the dialogue useless.
I no longer expect that you will learn to be attentive to the discussion and the methods of its conduct, as well as to the quality of the arguments given, just stating the facts.
Well, current in Q5 and Q6 will change in the same way without Q2 and Q3 also.
I do not see the advantage in this context of a "diamond" input.
Sorry when this is common knowledge - but "variable constant" is a nice one ...
As for the compensation of linear distortions by adjusting the frequency response and phase response of the input signal to the frequency response and phase response of the output voltage of the amplifier under test, as M. Sapozhkov did in his dissertation in 1954, and then P. Baksandall in 1972, as well as in a number of articles by I. Akulinichev, this is a gross mistake. In this case, we artificially eliminate linear distortions, and in fact it is they that largely form the phase relationships between the higher harmonics of the signal responsible for the timbre. After all, the amplitude-phase conversion caused by this has not been canceled.
Back in the 50s of the last century, Theo Williamson wrote that for high-quality amplification of an audio signal, it is enough that at maximum power THD was no more than 0.1%. In this case, the content of harmonics is virtually undetectable in the most sophisticated listening tests (obviously he meant short-spectrum tube amplifiers, this does not apply to modern transistor amplifiers). He also noted that phase shifts between the harmonic components of a complex signal in dynamics, especially on attacks of sounds, have a significant impact on the naturalness of the sound.
This idea was proved by the well-known expert in the field of psychoacoustics D. Griesinger in his talk "Pitch, timbre, source separation and myths about sound reproduction through loudspeakers" at the AES-132 held in Budapest in 2012. In his report, he showed that the phase relationships of the upper harmonics in the spectrum are extremely important, and also that when playing through loudspeakers, the structure of the amplitude and phase spectrum is always disturbed to one degree or another, therefore, the sound is always poorer in timbre, clarity, clarity, etc. . than natural sound.
fagos? открой книгу Раковского и изучай от корки до корки, хотя с твоим подходом ты вряд ли что поймешь, если до сих пор ничего не понял и продолжаешь корчить из себя чуть ли не Гуру
Back in the 50s of the last century, Theo Williamson wrote that for high-quality amplification of an audio signal, it is enough that at maximum power THD was no more than 0.1%. In this case, the content of harmonics is virtually undetectable in the most sophisticated listening tests (obviously he meant short-spectrum tube amplifiers, this does not apply to modern transistor amplifiers). He also noted that phase shifts between the harmonic components of a complex signal in dynamics, especially on attacks of sounds, have a significant impact on the naturalness of the sound.
This idea was proved by the well-known expert in the field of psychoacoustics D. Griesinger in his talk "Pitch, timbre, source separation and myths about sound reproduction through loudspeakers" at the AES-132 held in Budapest in 2012. In his report, he showed that the phase relationships of the upper harmonics in the spectrum are extremely important, and also that when playing through loudspeakers, the structure of the amplitude and phase spectrum is always disturbed to one degree or another, therefore, the sound is always poorer in timbre, clarity, clarity, etc. . than natural sound.
fagos? открой книгу Раковского и изучай от корки до корки, хотя с твоим подходом ты вряд ли что поймешь, если до сих пор ничего не понял и продолжаешь корчить из себя чуть ли не Гуру
Last edited:
The Church of Holly First Cycle
In the name of holly Graham, Baxandall and Hafler let your amplifier be blessed with the picosecond Time Propagation Delay.
Bandwidth is unlimited in our beliefs. Universe and its laws are not worth to be mentioned. Believe in unlimited bandwidth and you will be rewarded.
Distortion is evil. Let it never pass binding posts of your amplifier.
Believe in the Holly First Cycle and your ears will never be tortured by horrible sin of 0.000000000001 % distortion. Let the unworthy suffer.
Everyone who doesn’t believe is stupid.
Brothers, join us and bask in glory while nonbelievers will be damned.
Everyone with little or no understanding of audio is a welcome candidate.
Just send PM to the Church First Acolyte. You will know his name.
Yup, I’m irritated.
My post has all the quality of peter’s ‘arguments’.
In the name of holly Graham, Baxandall and Hafler let your amplifier be blessed with the picosecond Time Propagation Delay.
Bandwidth is unlimited in our beliefs. Universe and its laws are not worth to be mentioned. Believe in unlimited bandwidth and you will be rewarded.
Distortion is evil. Let it never pass binding posts of your amplifier.
Believe in the Holly First Cycle and your ears will never be tortured by horrible sin of 0.000000000001 % distortion. Let the unworthy suffer.
Everyone who doesn’t believe is stupid.
Brothers, join us and bask in glory while nonbelievers will be damned.
Everyone with little or no understanding of audio is a welcome candidate.
Just send PM to the Church First Acolyte. You will know his name.
Yup, I’m irritated.
My post has all the quality of peter’s ‘arguments’.
The difference between amplifier outputs might be of commercial interest for salesmen. The difference between input and output signal of an amplifier is far more interesting. Contraptions able to show such differences (adjusting output phase and amplitude to enable nulling of in- and output) are easy to make. As a rule the device's output is watched on a scope but listening to it might be more useful than measurement because of the psychological "annoyance multiply" factor.
Theo Williamson drew conclusions about the level of technology, amplifiers and acoustic systems available at that time. This gives us reason to doubt the accuracy of the figures, but it still does not prove your position in any way. You are criticizing the Ian Didden amplifier, which, according to your own measurements, does not exceed 0.05% distortion over the entire frequency range. If you believe Theo Williamson, then the difference between your amp and Ian Didden's amp should not be heard. Do you agree with this thesis?
I have never pretended to be a guru, but you are not able to separate personal grievances and dislike from a technical dispute, which is clearly expressed in your speech and reflected in your judgments. Besides, it's you who are playing the prophet of the audio world right now, not me. Critical thinking is not your strong suit.
He said:
Back in the 50s of the last century, Theo Williamson wrote that for high-quality amplification of an audio signal, it is enough that at maximum power THD was no more than 0.1%. In this case, the content of harmonics is virtually undetectable in the most sophisticated listening tests (obviously he meant short-spectrum tube amplifiers, this does not apply to modern transistor amplifiers)." (bold mine)
So he brings up *something*, then says, it does not apply to our discussion. Huh? Then why bring it up? Aren't we not enough confused already??
Jan
Back in the 50s of the last century, Theo Williamson wrote that for high-quality amplification of an audio signal, it is enough that at maximum power THD was no more than 0.1%. In this case, the content of harmonics is virtually undetectable in the most sophisticated listening tests (obviously he meant short-spectrum tube amplifiers, this does not apply to modern transistor amplifiers)." (bold mine)
So he brings up *something*, then says, it does not apply to our discussion. Huh? Then why bring it up? Aren't we not enough confused already??
Jan
Last edited:
Tombo, you understand nothing. You are expounding outdated technology useless in sonic reproduction quality. Nowadays we use diamond based transistors for atto second propagation delay measured utilizing non local entanglement principles distortion analyzer. Throw your misguided THD and FCD out the window, instead trust the yet undefined SOA Zero Point Fluctuation (ZPF) index for your simulation.
Jan, I brought Williamson's point just to emphasize once again that THD less than 0.1% does not mean an improvement in sound quality. The number of zeros after the decimal point does not mean absolutely anything. As the author of FCD Graham tried to convince the forum participants many times, THD is measured on sinusoidal signals in a steady state, that is, after the end of transient processes. But in music signals there are no pure sinusoids, there are chaotically changing impulse signals, so measuring THD is absolutely useless. Audioprecision (and similar devices) can only be used in the technological process to check amplifiers for compliance with the rating data (more precisely: good / bad) and nothing more.
The sound quality largely depends on the Group Delay, which, until recently, was not measured and was not entered in the amplifier passports.
The sound quality largely depends on the Group Delay, which, until recently, was not measured and was not entered in the amplifier passports.