So, it is difficult to find amplifier with slew limiting more than 1MHz. Because, many people still believe that amplifier with 100kHz slew limiting is good enough for them. Except Jand Didden, of course, who built 10MHz slew limiting amplifier.
Slew limiting is a combination of voltage level differences and frequency, you can't just discuss it in terms of freq alone. That is why it is specified in volts per second.
An NE5534 will not slew limit at 100kHz - as long as the level is 100mV.
Winfield Hill's 10MHz amp is not meant to drive audio into speakers, that's ridiculous. The amp has other uses.
Jan
An NE5534 will not slew limit at 100kHz - as long as the level is 100mV.
Winfield Hill's 10MHz amp is not meant to drive audio into speakers, that's ridiculous. The amp has other uses.
Jan
Many amplifiers often drive until clipping sometime, so it is relevance to discuss it in the term of frequency limiting.
Going back to the original topic issue:
While falling asleep/waking up today I tried to come up with some other metaphores and
this image came into my mind: lets take a conventional factory with the conveyor belt.
It represents the process of the production (i.e. the amplification / signal build-up).
When there is a well practiced collaboration the end result and the quality will be good.
But when the production line has to change and do something else, the whole system will
be a bit "confused" for the first time because it has some inertia for changes like this.
During this phase the production velocity will decrease and even the end result and quality
will be worse and the first few product will be probably "scrap" generating waste.
But after a certain settling time the process will be better and better and it will
reach an almost perfect result and quality level.
I like this example because it's very similar (in a lot of ways) with an amplfier.
It has a very specific and clear main target and main process.
It has a lot of internal and intermediate stages working together (both serial and paralell).
All these stages can have their own sensitivities and requirements to do the job right.
Also the system as a whole can have even local and global feedback loops, crosstalk, etc.
But the main difference between a real life factory and and amplifier is that when producting
a big quantity, a factory can afford to have some waste which it can throw to the garbage
(so the customer doesn't get it at all), but in an amplifier we can't have that.
It has to be "perfect" in every moment, it's not enough to work good after a time or in average.
And an ideal factory has 2 important speeds regarding this model:
1) the conveyor belt production speed (when doing the exact same thing continuously)
2) the speed (and a clean process) when adapting to new and changing requirements at the input
I guess FFT/THD is mainly for case #1 and the FCD is rather targeting case #2.
Other aspects:
- workers sensitivity to some conditions (air pressure, temperature, food quality, etc)
- efficiency and speed of the "quality assurence" processes
- factory reactions to strange (and sometimes reactive) customer behaviour
- overall performance regarding external conditions (factory building vs storms, wind, rain, etc)
While falling asleep/waking up today I tried to come up with some other metaphores and
this image came into my mind: lets take a conventional factory with the conveyor belt.
It represents the process of the production (i.e. the amplification / signal build-up).
When there is a well practiced collaboration the end result and the quality will be good.
But when the production line has to change and do something else, the whole system will
be a bit "confused" for the first time because it has some inertia for changes like this.
During this phase the production velocity will decrease and even the end result and quality
will be worse and the first few product will be probably "scrap" generating waste.
But after a certain settling time the process will be better and better and it will
reach an almost perfect result and quality level.
I like this example because it's very similar (in a lot of ways) with an amplfier.
It has a very specific and clear main target and main process.
It has a lot of internal and intermediate stages working together (both serial and paralell).
All these stages can have their own sensitivities and requirements to do the job right.
Also the system as a whole can have even local and global feedback loops, crosstalk, etc.
But the main difference between a real life factory and and amplifier is that when producting
a big quantity, a factory can afford to have some waste which it can throw to the garbage
(so the customer doesn't get it at all), but in an amplifier we can't have that.
It has to be "perfect" in every moment, it's not enough to work good after a time or in average.
And an ideal factory has 2 important speeds regarding this model:
1) the conveyor belt production speed (when doing the exact same thing continuously)
2) the speed (and a clean process) when adapting to new and changing requirements at the input
I guess FFT/THD is mainly for case #1 and the FCD is rather targeting case #2.
Other aspects:
- workers sensitivity to some conditions (air pressure, temperature, food quality, etc)
- efficiency and speed of the "quality assurence" processes
- factory reactions to strange (and sometimes reactive) customer behaviour
- overall performance regarding external conditions (factory building vs storms, wind, rain, etc)
There is no relevance from clipping to frequency or slew rate limiting. Clipping has nothing to do with slew rate limiting or with frequency limit.
Stay focused ;-)
Jan
If you are so superior to us simple inferior populace who wonder and ponder the discrepency between big number techonoligal marvel amps with THD in the ppm range (as an example) and the disappointing listening experience many people have with those "perfect" amps, then why don´t you post encuraging and helpful information to help us mere mortals in our quest to understand this subject better.
You mention FFT like it was a religious sacrament that is not to be used as a proper tool in the right way - but to be revered like the end all tool that explains everything in infinity - amen!. While gladly turning a blind eye to the circular reasoning that the FFT can never show anything else then it is made to do. Everything is a nail for the hammer and everything is sine waves for FFT. And taking it to the extreme with infinit bandwidth it is true. But in our part of the universe where infinity is a pretty theoretical concept, this does not make sense. There must be some reason that people and companies pour time, money and effort into developing nonlinear versions of the fourier transform (as I have shown previously in this thread).
If you don´t understand why and how a series of nonlinearly coupled nonlinear resonators create nonlinear and nonsinusoidal signals then you might want refrain from writing these condecending "answers".
This is great answer. This might lead nowhere and be totally worthless (don´t know), but it is a positive and encuraging post leading the discussion forward.
If all you are capable of writing is condecending "soft" (so as to avoid moderation) attacks on people who have a desire to discuss their ideas, then please do so some where else.
I might be naive but I thought this was a forum for people who want to discuss their ideas.
If people want to discuss why filling their loudspeakers stands with styrofoam and why this lowers distortion to negative values - then let them do that. I believe enough in people to let them make up their own mind regarding the validity in what is written on a forum.
This thread has been a great thread for me. I have learned a lot. Maybe not so much from the discussion we are having here, but from the countless hours of frenetic LTSpice usage trying to prove my ideas. Since I can´t prove them (yet) I keep on thinking, reading, searching for information and testing my ideas in LTSpice and in real life.
And in that I consider this thread a great success. It has sparked a lot of ideas and a lot of inspiration. And that is all I want coming to this forum. All the condecending and boring personal semi-attacks does nothing but detract from this (regardless if they are aimed at me or not).
Can we all help out in lifting this part of the forum to be a little more tolerant to differnt points of view or should we follow the dark forces of the US politics down into the abyss of polirization and intolarance for each others ideas?!?
Your ideas was challenged and you turned to spice for confirmation - and as you say, you learned something. A mature move as opposed to post stories of how cars that goes fast tend to wobble, trying to make an amplifier analogy. I hope you can see also then that the challenge itself had a value? If all here was just - ye ye I'm sure its fine - whatever... you maybe wouldn't have done the simulations... I regret somewhat the tone used - but when the arguments are so overwhelming and the criticising expertise so strong (not me!) - it gets embarrassing.
//
//
and First cycle distortion - Graham, what is that?
Are you involved in Very High End Hifi business to so often refer to amplifiers of stratospheric price ? Here it is a forum called DiyAudio where DIY means Do It Yourself. Your statements could more easily verified and checked with amplifiers made by members of this forum, they are not under niche commercial contraints.
You are definitely above me on this chart. Hats off)
I see what you're trying to do. Instead of a competent and well-reasoned argument, you are looking for the development of Yan to discredit him in the eyes of others and tip the scales in your direction. But everyone here is aware of these methods and they don't work. We are not discussing Ian Didden and his amps, but your methods of analysis and approval.
Just a great result. The perfect macros of the simulator you have produced distortions. Keep it up, this is definitely a breakthrough. Cyril Hammmer would be proud of you. You showed your level of understanding.
(Not really)
Kirill hammer is engaged in marketing and management activities, and does not design amplifiers. This is the responsibility of Christoph Schürmann.
So why do you continue to consider marketing publications to be scientifically reliable?
P.S. People are divided into two categories - some are able to admit their mistakes and develop further, while others are not. What category do you belong to? What kind of person will you be remembered for?
If you missed it, I expressed a more than reasoned opinion above in the topic, and in response I received only a Dunning-Kruger diagram. What kind of technical dispute can you have with a person in this case? Maybe you can do better, who knows... I can't do it yet.That's all he could answer to my arguments:
I did an analysis of several modifications of the popular amplifier by Hiraga, some of them have already been tested in hardware and received a good rating. Here is one of the options
Hiraga 20W class A
here's another example
Apex A40 fundamental improvement. (Sandy)
fagos, you are constantly looking for the beam in my eye. Recently, you recommended to colleagues to stay away from me, Whatever you do not stick to your recommendation, so no, you scour all my posts and, like a maniac, throw at the fan ...
Last edited:
@fagos
Understood. Looks to be frustrating on both sides.
I still suspect that the underlying problem is that when people hear something, they want an explanation for the underlying physical cause. Research shows the human default tends toward accepting as true something that superficially appears plausible. Once a very firm belief has been formed (perhaps even elevated to psychologically sacred status, something that must not be questioned), it can be difficult or impossible to change it. Debate and rational argument are only helpful to influence those undecided or only weakly decided, rarely effective on those already firmly decided.
IME the best results tend to come from showing someone the true causal explanation for their observation. Daniel Kahneman thought collaborative research into a problem with an opponent was best way to find mutual resolution between both people. Both researchers work together to dig deeper into the problem. Hard to do anything like that on a forum, and it takes time to bear fruit.
Understood. Looks to be frustrating on both sides.
I still suspect that the underlying problem is that when people hear something, they want an explanation for the underlying physical cause. Research shows the human default tends toward accepting as true something that superficially appears plausible. Once a very firm belief has been formed (perhaps even elevated to psychologically sacred status, something that must not be questioned), it can be difficult or impossible to change it. Debate and rational argument are only helpful to influence those undecided or only weakly decided, rarely effective on those already firmly decided.
IME the best results tend to come from showing someone the true causal explanation for their observation. Daniel Kahneman thought collaborative research into a problem with an opponent was best way to find mutual resolution between both people. Both researchers work together to dig deeper into the problem. Hard to do anything like that on a forum, and it takes time to bear fruit.
Last edited:
What kind of listening was it? Please share the setup details.
Feverish persistence is one of the telltale signs of those carrying business interest.
Peer support.
In that context, the right word is "perceive". Whether they actually hear or not would come after the verification.
Why did you not want to talk about Full power bandwidth? Full power bandwidth is slew rate / (2 x pi x maximum output voltage). I know you know. I think you have your opinion about it, but you do not want to share it.