What goes to design of great amps?
The same that goes to design of any great thing: understanding of design criteria, and understanding of how to achieve their optimum.
D. Self did not write his book about how to do it. He wrote his book about how to optimize an amp built on the certain topology. And of course, when you are limited by the single topology you have one, and only one, way to improve it: to select more and more optimal component sets for it.
But when your goal is to build a great amp, you are free to use different parts and different topologies.
The same that goes to design of any great thing: understanding of design criteria, and understanding of how to achieve their optimum.
D. Self did not write his book about how to do it. He wrote his book about how to optimize an amp built on the certain topology. And of course, when you are limited by the single topology you have one, and only one, way to improve it: to select more and more optimal component sets for it.
But when your goal is to build a great amp, you are free to use different parts and different topologies.
That is very true. Unless you enjoy sitting around listening to sine waves, specs do not tell the whole story. How an amp performs with music or movie tracks is entirely different.john curl said:
The best amp that I ever owned was a BEL 1001, after listening to it at the home of Dick Brown, the company owner. The BEL was rated at only 50w/channel but outperformed amps with much higher power ratings.
I agree with what you say (the last time I bought speakers, I didnt lok at any print, just compared them side by side) but I dont (and most would agree) believe its something magical. I believe our meausurements (objective standards) are not yet up to the task, that there is some physical explanation, we just dont understand it. The subjectivists dont even want us to look for it!
This is not necessarily true. Many 'subjectivists' may ignore measurements, but they don't generally condemn them. OF COURSE, there is some physical and objective reason why we can hear differences, even when virtually all the generally available 'objective' criteria have been met. We just don't measure it with the conventional test equipment available today. This should be of little surprise, because our test equipment still does not measure FM modulation of audio tones to any extent, Hirata, or other factors. They just remain buried in the literature.
I shoudnt generalize, I probably havent met enuogh subjectivists and am going on some of the extreme examples Ive seen, like the $$$$ power cords some swear by. My apoligies to the resonable ones. In general I think we agree.
Good discusion!
John, you mention other factors, do you know if people are working on measurement equipment for these. I suppose the complexities of such requires large investment, and the poeple who can afford it (SONY etal.) might not see any practical benefit. (increase there bottom line).
Member
Joined 2009
Paid Member
The deeper and more perfect the null you can obtain between the input and output, the more perfectly you've reproduced the electrical signal. We can reproduce electrical signals with remarkable precision. We can reproduce them so no human being can tell the difference in a properly conducted test, though I won't side track the discussion by naming that test. What seems to be difficult for people to accept is that perfect reproduction of the electrical signal often doesn't sound as good as imperfect reproduction. As a gross example I like LPs a lot, but by any measure they're a lousy and contaminated version of the original signal. As long as people try to connect electrical perfection with sonic perfection the arguments will go on, keeping the hobby alive and interesting.
It seems to be what is 'between the notes' is the most important. That is, jitter, IM products that usually look on instruments like hash, etc. We WANT to find the answer AND be done with it. Just try to make something that is really special? I do it all the time, and it isn't easy or cheap. TRY to make something cheap and easy that sounds just as good. I dare you!
Member
Joined 2009
Paid Member
I think it has been done , within the idiot's realm. for $50 US dollars and ON semi's 0281/0302's , one can have 8000$ bliss.
The "supersym"- AKA AMPSLAB BI120 are equal to a class A genesis stealth. I have both before me now.
So It is has already been done , as many have built the symasym , the "supersym' shall redefine what a DIY amp can do!!
OS
Measurements vs Hearing impressions...
This is messy subject, mainly because science is still ragged behind on understanding of how we perceive sound. There is an interesting paper published on Nature in 2007, finding that each of our neurons is tuned to a single note (<3% difference in frequency, roughly 1/4 note). The paper is a free access so anyone can download and read it (Nature vol 451, pp 197-201, "Ultra-fine frequency tuning revealed in single neurons of human auditory cortex. Y. Bitterman, R. Mukamel, R. Malach, I. Fried & I. Nelken). The paper states that bandwidth of response of a single neuron roughly matches to the capability of frequency discrimination by untrained person. What does this mean? An untrained person can hear the difference of two notes because different neurons fire up when he listens if the difference is larger than 1/4 notes. Somehow a trained person can discriminate finer differences, maybe by using the difference in the combination of neurons get fired up. I believe a lot of audiophiles and music lovers fall into a category of "trained person".
We are not born equal in terms of auditory neurons. If the density of auditory neuron follows the pattern of other sensory neurons, there could be 10 fold difference from person to person on the density of them. Each auditory neuron is tuned to a narrow bandwidth while we have a limited number of them... Some have more while some have less. Also, as we age, we loose many of them. The rate of loss is not equal, either. I have a colleague (28 yr) who totally lost hearing above 16kHz while I (48 yr) still maintain some hearing there (I'm lucky). I will lose mine soon or later for sure but just not yet.
Why am I writing this? I'm trying to say that:
1. Measurement of electrical characteristics of audio equipments may not reflect how we like (of dislike) the sound generated by them. We are far away from understanding of what is "fine" or "good" sound, as we don't fully understand how we hear. Of course, an amplifier with way more distortion than how it was designed, should have problems (oscillation, poor construction like cold solder joint, interference, etc) and measurements can reveal them. In near future, new measurement methods and equipments may come up which can predict how the audio system will "sound" by incorporating new discoveries on research of mechanism of our hearing and perception.
2. We have large variation on what we can hear and how we hear. This is based on genetical background and personal history. If person A cannot hear a difference, it doesn't mean person B should not hear a difference. In a case you hear a difference, it doesn't mean that other people can hear the same difference.
In addition to these variations, there is a variation on the "taste" of sound.
3. This is not scientific remark, but I believe there are some systems that most of the people will agree that they sound "good". Advancement in physics, physiology, and cognitive science may give us some clue on the direction for such system, but meanwhile, we all can scratch our heads and have fun on endless discussion/argument on what is a "good" sound!
Thank you very much for reading to this point. All comments and critics are welcome!
Regards,
Satoru
This is messy subject, mainly because science is still ragged behind on understanding of how we perceive sound. There is an interesting paper published on Nature in 2007, finding that each of our neurons is tuned to a single note (<3% difference in frequency, roughly 1/4 note). The paper is a free access so anyone can download and read it (Nature vol 451, pp 197-201, "Ultra-fine frequency tuning revealed in single neurons of human auditory cortex. Y. Bitterman, R. Mukamel, R. Malach, I. Fried & I. Nelken). The paper states that bandwidth of response of a single neuron roughly matches to the capability of frequency discrimination by untrained person. What does this mean? An untrained person can hear the difference of two notes because different neurons fire up when he listens if the difference is larger than 1/4 notes. Somehow a trained person can discriminate finer differences, maybe by using the difference in the combination of neurons get fired up. I believe a lot of audiophiles and music lovers fall into a category of "trained person".
We are not born equal in terms of auditory neurons. If the density of auditory neuron follows the pattern of other sensory neurons, there could be 10 fold difference from person to person on the density of them. Each auditory neuron is tuned to a narrow bandwidth while we have a limited number of them... Some have more while some have less. Also, as we age, we loose many of them. The rate of loss is not equal, either. I have a colleague (28 yr) who totally lost hearing above 16kHz while I (48 yr) still maintain some hearing there (I'm lucky). I will lose mine soon or later for sure but just not yet.
Why am I writing this? I'm trying to say that:
1. Measurement of electrical characteristics of audio equipments may not reflect how we like (of dislike) the sound generated by them. We are far away from understanding of what is "fine" or "good" sound, as we don't fully understand how we hear. Of course, an amplifier with way more distortion than how it was designed, should have problems (oscillation, poor construction like cold solder joint, interference, etc) and measurements can reveal them. In near future, new measurement methods and equipments may come up which can predict how the audio system will "sound" by incorporating new discoveries on research of mechanism of our hearing and perception.
2. We have large variation on what we can hear and how we hear. This is based on genetical background and personal history. If person A cannot hear a difference, it doesn't mean person B should not hear a difference. In a case you hear a difference, it doesn't mean that other people can hear the same difference.
In addition to these variations, there is a variation on the "taste" of sound.
3. This is not scientific remark, but I believe there are some systems that most of the people will agree that they sound "good". Advancement in physics, physiology, and cognitive science may give us some clue on the direction for such system, but meanwhile, we all can scratch our heads and have fun on endless discussion/argument on what is a "good" sound!
Thank you very much for reading to this point. All comments and critics are welcome!
Regards,
Satoru
Believe me , I am not being subjective , I don't dislike class a or any other topology , but a/ B before me (AB symasym) "blows away" the A class amp , no contest!!! AT 250 watts , maybe I am biased by the lack of heat
but it is an experience to be remembered. FLAC's and 320KBS MP3's , If I only had a CD player to really see the true headroom of this 77-0-77Vdc rail amp. 
I would die...
OS
But I can find no error.
listening to the genesis , which I know Is "burning" 1 A per amp (per 8) It has a sound that "melts into The walls".
This is subjective , but at obscene levels , the AB exceeds the A, no doubt. It retains the same soundstage (absolutely) and transientwise , lay's waste to it's inferior topology.
I know the A should be better , but at all levels , the majority chooses the "SYM" (not my fault)
I have repaired OEM topologies all my life to make a $$ , and the
symasym at 77 volt rails has "expanded" my expectations for what a DIY amp can deliver..
I also have "bat" ears .. so no $$hitty amps can get by me..
(this is a n. american "carlos" response , I know)
OS
listening to the genesis , which I know Is "burning" 1 A per amp (per 8) It has a sound that "melts into The walls".
This is subjective , but at obscene levels , the AB exceeds the A, no doubt. It retains the same soundstage (absolutely) and transientwise , lay's waste to it's inferior topology.
I know the A should be better , but at all levels , the majority chooses the "SYM" (not my fault)
I have repaired OEM topologies all my life to make a $$ , and the
symasym at 77 volt rails has "expanded" my expectations for what a DIY amp can deliver..
An externally hosted image should be here but it was not working when we last tested it.
I also have "bat" ears .. so no $$hitty amps can get by me..
(this is a n. american "carlos" response , I know)
OS
It isn't the same amp that you are comparing, is it? A true, powerful, class A amp seems to beyond most people here, even me. Therefore, most true class A amps are 100W or less. I build an 800W amp (4 ohms, rated) with lots of current to spare, and it does sound pretty good, loud, BUT it is not a class B amp, or a class A amp, just something, in between. However, IF I COULD turn up the bias even further by adding fans or water cooling, it would sound better, nonetheless. Measure better, too!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.