I note that with the chokes in circuit that they lengthen the capacitor power-on charge time and reduce ripple considerably.
I haven't personally spoken to Erno Borbely for many years, perhaps 20 years ago, when he was working with National Semi. We are both colleagues and competitors, and we design similar circuits. The last time that he contacted me was when I had criticized a design offered on the internet, that could be problematic, if care was not taken in device selection. He assured me that he selected the devices so that they would be in the right Idss range to stay out of trouble, and I apologized for impugning that he had done anything wrong, even though an amateur might make a serious mistake with the same circuit.
However, I am reminded of the Borbely designed Hafler power amp circuit that apparently did so well, decades ago, in a differential subtraction test, and how it actually performed, or sounded in my own hi fi system, decades ago. Not bad or anything, it was just that Dick Marsh was finding 'improvements' to it. I wonder if Erno would admit, let's say, on his 'deathbed' (hopefully far in the future) that it is all that was needed to make a 'perfect' power amplifier.
However, I am reminded of the Borbely designed Hafler power amp circuit that apparently did so well, decades ago, in a differential subtraction test, and how it actually performed, or sounded in my own hi fi system, decades ago. Not bad or anything, it was just that Dick Marsh was finding 'improvements' to it. I wonder if Erno would admit, let's say, on his 'deathbed' (hopefully far in the future) that it is all that was needed to make a 'perfect' power amplifier.
Glad I could help bring everyone together...
A consensus is always good!
So, why was it that we were testing resistor residuals a while back?
Anyone refresh me on that?
_-_-bear
Bear
Just because some people have claimed that some resistors sound better than others and certain measurements agree with what they hear, there are others who prefer the resistors that measure poorly. Although I find it interesting that there is a correlation with what some hear and the measurements, in my OPINION I do not think the difference is great enough to pass the types of scientific tests many here rely on, except for some obviously inferior parts.
For example I am assembling a large sound system. For the past week I have been soldering connectors to P.C. cards. On some connectors I am using lead free solder. By visual inspection I find that I need to use about 1/2 as much lead free solder per joint as I do with the older leaded variety. I find the fillets do not form well if there is too much solder. The joints are not as shiny as the leaded ones. I do know that 40/60 solder was used for body work because it flowed nicer than the 60/40 electrical solder, so I am not surprised that the lead free does not flow as nicely.
When I measured solder joint distortion, I could not find any significant difference between the various solders tested. However the lead free are not as shiny, so some folks think they are "cold" or bad solder joints.
So the lesson for me is to use a hotter iron and less solder to keep spectators happy!
Not surprising some people hear with their eyes, there is good research as to that. So I will try to please them, it helps get the bills paid. Shiny-Happy, pay bill.
So when you can measure one part is better than the other, use the better parts, it will make your ears happier, although it may really be your eyes doing the hearing. Of course after you use ten or twenty better parts it may actually be your ears not your eyes.
Ears, eyes, sounds better, not worse, go for it... It is a binary thing. That must be why digital audio is so good!
Have fun,
ES
There is no such thing as a totally transparent audio system – not in any existing system.
Thanks.
While these generalities look so ´innocent´ they are comprehensive and it is impossible to show the validity of a test or its results if you don´t have controls incorporated.
If an experimenter does not use a negative control, he does not know if a positive test result (means rejection of the null hypothesis) was due to the effect investigated or due to other effects.
If he does not use positive controls on a sufficient sensitivity level he does not know if a negative test result (means the null hypothesis could not be rejected) was due to the inaudibility of the effect investigated or due to other effects.
It is irrelevant if your are researching the same as jj or just a claim.
@Jakob, Frank, John,
I believe that it is, not sure how new it is. I went to a conference on hearing aid development at Eindhoven University, a presentation and discussion by a guy involved in algorithm development for hearing aid DSP's. These guys are very deep in the details of how hearing works on a physical and psychological level. It's a field where a lot of money goes around and with strong growth rates (after all, our average age rises), and fierce market competition.
I'd like to learn more about it when I get some time, it was totally new to me.
Another thing I learned: these guys produce 120dB loudness at your ear with a total power budget of 3mW....
jd
Joachim what you say is very important. I remember an interview with a guy, forgot his name, who designed lots of amps for Rotel. He said that he found a way to make an amp sound pleasant and enjoyable, but not transparent. The reviewers of his amps were often confused because it measured not very good but they just couldn't turn it off!
Unless you are a very naive listener, a good-sounding amp must meet some basic technical specs. Reasonable low distortions, no audible hum or noise, control over the speaker with a reasonable damping factor, enough output level so that it doesn't clip every few seconds, that sort of thing. We can fight over the exact figures, but I doubt whether 0.0001% THD and a damping factor of 500 are really necessary.
jd
I try to make my designs as technically well as i can and i think we had progress since the 50th. If it whould not be this way then an original Wiliamson amplifier whould still be world class in terms of sound that is indistinguishable from the source. All this tests have been made in the 50th and 60th and the conclusion was that well designed amplifiers provided they are not overdriven sound the same. All that contribute here know that of cause. Still i am sure we have to do this testing again because source material and speakers got better. A sound quality test is only as good as the source and the speakers ( plus setup ). I have attended double blind tests and in no case was the soundquality as high as i get it in my home. What i try to say is that there is nothing wrong with ABX double blind tests as long as very high standarts of source material and transducers are provided but that was not the case in tests i attended.
Also cosider that humans are not able to distinguish fine details if they did not have enough experience with this details. I know an anecdote where people listened to Edisons talking machines and could not distinguish the artificial voice from the original voice. When Columbus landed in the karibien, the native americans where thinking that gods are comming on white clouds. In fact i have seen a television documentary where
researchers found out that the native americans could not see Columbus ships at all because they had never seen something like that so could not make sense about what was happening.
Also cosider that humans are not able to distinguish fine details if they did not have enough experience with this details. I know an anecdote where people listened to Edisons talking machines and could not distinguish the artificial voice from the original voice. When Columbus landed in the karibien, the native americans where thinking that gods are comming on white clouds. In fact i have seen a television documentary where
researchers found out that the native americans could not see Columbus ships at all because they had never seen something like that so could not make sense about what was happening.
@ janneman,
quite new results in better modelling and deeper insight, for example:
PLoS Computational Biology: Power Efficiency of Outer Hair Cell Somatic Electromotility
with a lot more information in the references:
http://www.iib.uam.es/servicios/nin...ayo/Hair_Bundle_2006_NATREVNEU_Fettiplace.pdf
which covers more the way the system works.
That there is an interaction was already known earlier, but it was only speculated about any possible feedback mechanism, see for example:
hearing2
quite new results in better modelling and deeper insight, for example:
PLoS Computational Biology: Power Efficiency of Outer Hair Cell Somatic Electromotility
with a lot more information in the references:
http://www.iib.uam.es/servicios/nin...ayo/Hair_Bundle_2006_NATREVNEU_Fettiplace.pdf
which covers more the way the system works.
That there is an interaction was already known earlier, but it was only speculated about any possible feedback mechanism, see for example:
hearing2
Yes that seems logical. That is why it is so important to verify before making the test (d)blind that the subject actually can hear a difference.
But there are other sides to it.
If we agree that 'reasonable, but not top-notch' equipment can hide fine details and therefore 'sound the same', they should sound the same whether sighted or (d)blind. Logically, if they sound different sighted, but not (d) blind, on the same system, we can conclude that they really DO sound the same, irrespective of what we thought in the sighted test.
We can't really escape it, nor should we want to: if the sighted test says different, and the (d)blind says no difference, on the same system, it's no difference.
jd
- Status
- Not open for further replies.