Because it's not enough "just" to be fast at all costs.
It has to be "clean" and after that fast as possible.
There are no such effects, name one, people make them up. Test signals are not test equipment like a microscope, we use oscilloscopes. what your saying is you need a 1ghz scope for audio or your missing something. Ridiculous.
If you all are saying that THD and FFT is ultimate and perfect then how it can be that it's still not describing the "perfect" sound..?
We all know it's not enough because there are other aspects which aren't covered (or at least not emphasized enough) by this method.
That's all and that's why there is (at least) a chance that maybe it's worth a try to search for new methods.
But why to attack it so actively I don't understand...
If you don't believe it's ok but why to argue it?
Why not to go along adding your brain capacities to this approach if you are intrested?
How could you tell so early that this is definietly a dead end..?!
If there were no such effect we would be at the "final stage" with all the power to know all background to all kind of distorion and amplifier "sound type".
But as far as THD doesn't describes this "domain" there should be some effects we don't consider right now.
Like the health example: if you know everything you cant get sick.
But if you go further you'll discover the hidden world of viruses and bacterias and then the mystic way of getting sick will be discovered.
We can handle the current state as closed and final but I guess it's more precise that it's probably not the case.
@Cortez, the only way to ensure good result for “first cycle” test is to make the BW as wide as possible. It is clear engineering problem and there is enough information on the net to address it, not a rocket science (especially in case of headphone amp). Anybody can try it and show the world his/her wunderamp.
What makes me skeptical is the lack of evidence a 1MHz amplifier is subjectively much better than a 200kHz amp. Actually it’s easy to demonstrate it. Take your 1MHz amp and put a switch which limits the BW to 200kHz. I'd bet there will be no difference (except in case of RFI pollution).
What makes me skeptical is the lack of evidence a 1MHz amplifier is subjectively much better than a 200kHz amp. Actually it’s easy to demonstrate it. Take your 1MHz amp and put a switch which limits the BW to 200kHz. I'd bet there will be no difference (except in case of RFI pollution).
You magnify them with sensitive equipment, like your microscope analogy. Driving the circuit with an unrealistic signal serves no purpose.
There are people who claim whey can hear all sorts of stuff - differences between top notch opamps, different upsamplers, mp3 320k vs uncompressed etc. I tried to challenge them to try different measurement technologies like comparing input and output differences while using noise or music signals. There are at least two programs out there that try to do just that. Nobody of the "golden ears" ever did anything like that, most of the time I got outright refusal to try, because nobody can be bothered "To do MY work".
Why don't I do it myself? Well, 1. I'm still decking out my system with decent DACs, amplifiers and am in the process of learning how to do proper speaker and room correction (which i believe to be paramount). And secondly up to today I have never experienced any difference between mp3 320k and WAV, not on a 5 way horn system, not on a headphone. never. So probably I have only bronze ears. if at all.
Why don't I do it myself? Well, 1. I'm still decking out my system with decent DACs, amplifiers and am in the process of learning how to do proper speaker and room correction (which i believe to be paramount). And secondly up to today I have never experienced any difference between mp3 320k and WAV, not on a 5 way horn system, not on a headphone. never. So probably I have only bronze ears. if at all.
I hope it's not necessary but it's doesn't hurt.
But getting back to the original idea of the topic:
A classical analog oscillopscope is still "just" showing us a continuous signal "live"
and so it's not good for analyzing a FCD. So the point is to the "store functionality".
I guess the main point with this method is not just the "first sine wave" but to have
one only one exact test signal and check the amplifier response to that in different points.
Maybe it's usefull to check it like this instead with a continuous and "steady" (AC) signal.
As I tried to illustrate earlier: what if an amplifier distortion is "dynamic"?
I mean if we start a 1kHz ideal sine wave gradually (if you hate starting it from 0...
)and as we reach the final desired amplitude the very first period will have 0.2% THD,
the 2nd 0.15%, the 3rd 0.1%, the 4th 0.08% ... and the 10th 0.00001%?
What if there was an amplifier which is "faster" and more stable..?
(If you know what I mean...
)Not a bad idea at all..!! (Of course as a transient load to check the responses.)
Maybe we would get some surprising and unexpected results at the different stages...
Not necessarily (See above.)
No one said that the (at all costs) fastest amplifier is the holy grail.
The topic is about a possible distortion detection method which
is built up on a unique test signal and measuring method.
-> Checking a surface with UV light serves no purpose...
You don't have to believe but then please dont argue or spam the topic with your scepticism just because.
Not everyone and all the time needs a UV lamp but it can be usefull to go on from a specific case.
Regarding the analogy level "debate":
Of course we have to be abstract on this level as we are trying to discover something new
we don't already know. We have to have some blurry image at least to start with.
It was the same when the greeks said that there should be even more tiny things that
make up our whole world not just sand (but atoms) even if they didn't see it directy
just in their imagination. You have to first "create" the devil to defeat it later.
I was referring to the specific requirement for good response to sudden 180 deg phase flip central to this thread. If one wants it then wide BW is the only way.
I'm open to new ideas and theories. If anybody demonstrates audible difference between straight 1MHz amp and the same amp with RC lowpass to 200kHz on the input, I'll give credit for important finding/invention. Till then any further discussion is just empty theorizing.
Please have the decency to stop posting for your own sake - you are embarrassing yourself.
//
Attachments
One suggestion to first cycle distortion exterminators. I’m serious. If wide BW is difficult you can try with nonlinear frequency response above audible range. Some rise in 50-300kHz range might give results you are looking for. Yet I remain skeptical about positive impact to subjective perception.
When trying something new you can't be 100% sure where you'll end up.
It's just an idea, an experiment, a hope.
There is an idea that testing an amplifier with only 1 period can be usefull and that's all.
Of course it's related to speed but not just that.
Yes it is... that's how brainstorming works... At least this is how I see it.
No one said that this is a fully developed and proved method...
It's an idea and a good topic to talk about, maybe it has some potentials...
Thanks for your personal reaction.
I'm trying to "fight" for the original topic and for the idea but you just sush me per se...
But be calm I'll leave soon.
Till then some uninteresting thoutghs reagarding speed and distortion detectability:
The "new" things should be fast and tiny by nature (as otherwise we'd already caugth them).
It's relative easy to build a precise and stable DC amplifier, rigth?
The goal is them same with AC, but going up with the speed the things start to mess up.
Some (maybe) good questions could be:
How many speed (or rather time constant) an amplifier has?
How many input (control) point an amplifier has? (Don't get lost based on the schematic.)
As we start to "pull" the system more rapidly it starts to "fall apart".
The amplifier stages/parts will be modulated (all in a different way)
and so it even can interfere with itself.
(And I'm not talking about the conventional unstability oscillations.)
It's similar when driving a car: as the speed goes up (even on a completely
straight and perfect road!) new weird effects and strange "noises"
(you could call them harmonics if you like) are arising from nothing.
And after all an ideal amplifier (of course):
- is "fast" (not GHz) but -> reacts quickly (and "cleanly") to the input
- doesn't start to live it's own life in it's parts -> doens't generate even more trouble
- handles the road good -> doens't start to shake when the road is not 100% perfect
These are the "last" technical challenges we can solve.
All other remaining things are then "just" fine tunings like the color profile on your TV.
In listenable sound, the audio level at 20 kHz is ten times lower than at 2 kHz.
So the frequency of constant magnitude of group delay you advise can be divided by 10.
The process is different about passive low pass filter and slew limiting amplifier. Jan Didden can explain better, because he has better English than me
look in the mirror and recognize yourself
From the description of the Soulution 711 amplifier design:
“From the input the soulution 711 first buffers the music signal. A high performance correction amplifier is then used to capture and correct any deviations in the signal, very quickly and very precisely. Ultra wideband amplification using no feedback loops follows, a less than 10 nanosecond transit time and level accuracy to within 0.1 dB ensuring signal purity. The whole is housed within a sealed module, maintaining a common temperature across all stages to further guarantee precision. The output stage comprises 14 power transistors per channel, all mounted onto a massive copper rail which is temperature controlled for a constant quiescent current. The soulution 711’s impressive low level signal handling is reflected in its raw power delivery. Current levels of 120 A and more are readily attainable.»
Philosophy
The ideal amplifier is perfectly load stable, 100% level and phase accurate, and has no long signal paths. It should not use tricks like over amplification and very high negative feedback to achieve good results in the lab but poor sonic per*for*mance in the listening room. The soulution 711 stereo*amplifier combines extreme speed with an ultra-wide band*width (1 MHz/– 3 dB) and a high current capability far beyond that of valves and output trans*formers. A unique and sophisticated circuit design opens up sonic qualities once considered incompatible: Precision, Velocity, Stability and Power. In the soulution 711 they are brought together for the first time – all in the pursuit of musical perfection. The soulution 711 is closest of all to this ideal. It can provide up to 300 Watt continuous power into 4 Ω and more than 6,000 Watt transient power per channel. More than enough to breathe new sonic life into even the most demanding, low-efficiency loudspeakers.
Last edited:
Cortez, colleagues at the top of the Dannig-Kruger chart at the beginning of the journey and imagining themselves as Gurus, look at the book and see a fig. Musical signals are not continuous harmonic, but rather impulsive random signals. They do not know that the term "linear distortion" applies only to the steady-state mode of a harmonic sinusoidal signal. And for musical signals, these are continuous transient processes with specific distortions associated with the characteristics of the group delay, which are not obvious from the frequency response and phase response graphs. The theory is that to transmit random signals without distortion, the bandwidth must be infinite and the phase response must also be linear to infinity. But in real life there are no such amplifiers. Here's what Cyrill Hammer said about this (I repeat for those who have forgotten or do not know):
“Impeccable performance is equally important. This is especially true for amplifiers with general feedback. The theoretical concept of negative feedback is very powerful, and the simplified mathematical equations that describe this concept always hold true. But they are only valid if the design takes into account the constraints of the concept. Delay from entry to exit must be zero! Obviously, this is not possible in real life.
There are two ways to solve this problem:
1. You simply do not apply any negative FB in your project at all (giving up the advantages of the concept).
2. 2. You speed up the amplifier to the level of a few nanoseconds of delay from input to output (respectively 200 MHz bandwidth in the case of Soulution 700 and 710), then the compensation errors are so small that they have no noticeable effect on the sound. As soon as you decide to go the second way, suddenly many new problems arise: thermal conditions, stability of the supply voltage, design corresponding to high-frequency devices, the occurrence of noise, interference, etc. "
Let me remind you that most tube amplifiers are just without general NFB.
The sound quality is affected by all the parameters of the amplifier in the complex, starting from the virtual parameters of the circuit, the quality of the PCB layout, the quality of the components used, the quality of the power supply, etc. and ending with the quality of installation in the case.
“Impeccable performance is equally important. This is especially true for amplifiers with general feedback. The theoretical concept of negative feedback is very powerful, and the simplified mathematical equations that describe this concept always hold true. But they are only valid if the design takes into account the constraints of the concept. Delay from entry to exit must be zero! Obviously, this is not possible in real life.
There are two ways to solve this problem:
1. You simply do not apply any negative FB in your project at all (giving up the advantages of the concept).
2. 2. You speed up the amplifier to the level of a few nanoseconds of delay from input to output (respectively 200 MHz bandwidth in the case of Soulution 700 and 710), then the compensation errors are so small that they have no noticeable effect on the sound. As soon as you decide to go the second way, suddenly many new problems arise: thermal conditions, stability of the supply voltage, design corresponding to high-frequency devices, the occurrence of noise, interference, etc. "
Let me remind you that most tube amplifiers are just without general NFB.
The sound quality is affected by all the parameters of the amplifier in the complex, starting from the virtual parameters of the circuit, the quality of the PCB layout, the quality of the components used, the quality of the power supply, etc. and ending with the quality of installation in the case.