Come on we're in a world where better specs does not mean better performance...so in many of these respects, actually building the stuff for a long period gives you insight and knowledge about things that works and don't work...One have to listen to the experienced...There is much learning here and also lot's of not so much learning.. task is to find the gold in all the jibberish...
Specs are often important in the commercial world. That is MY world of design. I don't have to agree with the specs. but I have to meet them, if they are imposed by other 'standards' committees. THX is one example of a set of 'specs' that I must meet for Parasound products.
Do these 'specs' accurately reflect what some audio product will sound like? Not necessarily, and that is why many Hi End designers IGNORE THX and other standards like that. Sometimes, they make designs with embarrassingly high distortion when reviewed by 'Stereophile' for example. This is NOT necessarily because they are sloppy designers, or deliberately introduce distortion into their designs, but because they refuse to compromise by using global negative feedback.
Sometimes, especially with single ended tube designs, the RESULT is amazing! It sounds REALLY good! Better than I have been able to do, without joining them, and NOT using global negative feedback in some of my better designs. How I envy them, and their sound quality, and that is ONE reason why I try to understand what is 'wrong' with 'negative feedback' and how best to minimize any problems.
Do these 'specs' accurately reflect what some audio product will sound like? Not necessarily, and that is why many Hi End designers IGNORE THX and other standards like that. Sometimes, they make designs with embarrassingly high distortion when reviewed by 'Stereophile' for example. This is NOT necessarily because they are sloppy designers, or deliberately introduce distortion into their designs, but because they refuse to compromise by using global negative feedback.
Sometimes, especially with single ended tube designs, the RESULT is amazing! It sounds REALLY good! Better than I have been able to do, without joining them, and NOT using global negative feedback in some of my better designs. How I envy them, and their sound quality, and that is ONE reason why I try to understand what is 'wrong' with 'negative feedback' and how best to minimize any problems.
Come on we're in a world where better specs does not mean better performance...so in many of these respects, actually building the stuff for a long period gives you insight and knowledge about things that works and don't work...One have to listen to the experienced...There is much learning here and also lot's of not so much learning.. task is to find the gold in all the jibberish...
Actually better specs do mean better performance if you are looking at the right stuff to spec.
But there still is the matter of taste as to what you actually want.
Mr Børresen,
reminds me, your loudspeakers were demo'd with amps by Burn-Meister couple of years ago (032 integrated, afair).
Recently, it's shifted towards using Rowland amps.
Apparently, if you have an influence in these choices, it seems you do monitor specs.
Last edited:
Martin, the circuit we simulated has as the "second" stage double folded cascodes supplied by
contant current sources. This gives a current output that drives an All In One Go transimpedance RIAA and then comes a buffer. We did not try to unbalance the drain loads. In the simple case of drain resistors you could be right with both aproaches but we did not try that. I know that you are a capable simulator so you could try yourself what happens. It could affect output impedance and clipping but i am not sure about that. in the circuit we tested it was mainly the total Gm that counted. We did not try to grossly unbalance though.
John, concerning the audibility of distortion i really think that a nice exponentially falling distortion profile can sound better then no distortion at all. That is one of the "unmasking" effects i talked about earlier. It is simply not true that we have to do with perfect recordings in most cases and the "nice" distortion introduced may help to restore some missing overtones. I know, this view is provocative and the added distortion is artificial in a philosophical sense. Stiil the best amp i ever heard in my system is the Vacuum State KT88 amp that the late Allen Wright designed. It has no global feedback and Allen did sugest right away that it could have large amounts of distortion because he did not make any distortion measurements in the design phase. Looking at the circuit i think it may not be that bad. I asume that it has around 0.1% at cruising level. A level that whould put any competent solid state designer to shame. Ironically it sounded very clear and open. I could really forget the system and listen to the music. More in the latest Vacuum State News Letter.
contant current sources. This gives a current output that drives an All In One Go transimpedance RIAA and then comes a buffer. We did not try to unbalance the drain loads. In the simple case of drain resistors you could be right with both aproaches but we did not try that. I know that you are a capable simulator so you could try yourself what happens. It could affect output impedance and clipping but i am not sure about that. in the circuit we tested it was mainly the total Gm that counted. We did not try to grossly unbalance though.
John, concerning the audibility of distortion i really think that a nice exponentially falling distortion profile can sound better then no distortion at all. That is one of the "unmasking" effects i talked about earlier. It is simply not true that we have to do with perfect recordings in most cases and the "nice" distortion introduced may help to restore some missing overtones. I know, this view is provocative and the added distortion is artificial in a philosophical sense. Stiil the best amp i ever heard in my system is the Vacuum State KT88 amp that the late Allen Wright designed. It has no global feedback and Allen did sugest right away that it could have large amounts of distortion because he did not make any distortion measurements in the design phase. Looking at the circuit i think it may not be that bad. I asume that it has around 0.1% at cruising level. A level that whould put any competent solid state designer to shame. Ironically it sounded very clear and open. I could really forget the system and listen to the music. More in the latest Vacuum State News Letter.
Jacco... we have used so many different products to demo our speakers that it's unbelievable, from the rather fantastic VTL siegfreds over Rowland transformer equipped products and Ypsilon to Sim audio and also Burmester...in Munich Last year we used Goldmund...I do believe that the speakers must work with many different types and brands of amplifiers,,,,most often we work the shows in cooperation with the distributers and that will most often set the gear....
When it comes to speakers....all is in the compromises...I mean theres plenty of good reasons not to aim for the best of specifications and ruler flat response curves there...as you literary work in a minefield of psychoacoustic hearing elements that correlates really badly with what you can measure...On single drive units its differnt as you can't get good enough spechs...not in terms of liniarty but in terms of the lowest possible breakups and total absence of mechanical resonances...(different topic though and for me this is hobby and I don't contribute to the lsp forum)
When it comes to speakers....all is in the compromises...I mean theres plenty of good reasons not to aim for the best of specifications and ruler flat response curves there...as you literary work in a minefield of psychoacoustic hearing elements that correlates really badly with what you can measure...On single drive units its differnt as you can't get good enough spechs...not in terms of liniarty but in terms of the lowest possible breakups and total absence of mechanical resonances...(different topic though and for me this is hobby and I don't contribute to the lsp forum)
Last edited:
Joachim, you are getting closer. What I want to talk about is an IDEAL first gain block of a 2 gain block phono stage, that can have IDEAL characteristics. I, personally know how to make comp-diff jfet input stages, I do it every day. Here, I am TRYING to just use a single pair of parts for the input, with a FOLDED CASCODE AND a CURRENT MIRROR to create a push-pull drive.
Joachim, your last input was very close to a Vendetta Research SCP-1 input. Very well done, but perhaps too sophisticated for consideration here.
I was hoping to simplify a single differential N channel jfet, folded cascoded design that could be made to become an OPEN LOOP gain block for a phono stage. Of course, IF I were making it for myself, then I would use the complementary approach, but many, including Parasound, one of my clients, has found getting p channel low noise fets almost impossible to obtain, yet we can still buy 1000's of n channel fets with the same general specification.
The alternative to this sort of design, is to use an existing IC op amp. This is what I do in the JC-3. This approach works, but better can be done.
First of all, MC raw output is one of the most difficult and fastest sources for audio. It is NOT the music, but the occasional mistracking that is by far, the hardest to handle. An IC must pass this nearly full power 4 us pulse perhaps, and it is hard to measure any distortion contribution, but I am sure that if our test equipment was sophisticated enough, we would find lots of problems.
One of the best ways to avoid 'lots of problems' is to run the first gain stage OPEN LOOP. Then you can elegantly control the risetime, reducing it to a 10us rise time square wave, and handle it from there with whatever you want.
This is the 'secret' of the Vendetta Research SCP-2, and any other hi end phono design that I make.
However, not ANY gain stage configuration will work optimally, any more than just any IC will work in the phono input, even usefully. If you want to hear differences in IC's, I recommend buying a JC-3 phono stage and trying different op amps, and listening to each of them. Take notes! Measuring the differences is whole other ball game, but trust me, the JC-3 works pretty well, because that is what I had to do, last year, just before release of the product. I am still looking for the world's best IC to put in that place (from a listening point of view). New ones come out every month or so, but alas! They do not show the schematic, and there could be all kinds of 'compromise' that would ultimately show up in actual listening tests.
I was hoping to simplify a single differential N channel jfet, folded cascoded design that could be made to become an OPEN LOOP gain block for a phono stage. Of course, IF I were making it for myself, then I would use the complementary approach, but many, including Parasound, one of my clients, has found getting p channel low noise fets almost impossible to obtain, yet we can still buy 1000's of n channel fets with the same general specification.
The alternative to this sort of design, is to use an existing IC op amp. This is what I do in the JC-3. This approach works, but better can be done.
First of all, MC raw output is one of the most difficult and fastest sources for audio. It is NOT the music, but the occasional mistracking that is by far, the hardest to handle. An IC must pass this nearly full power 4 us pulse perhaps, and it is hard to measure any distortion contribution, but I am sure that if our test equipment was sophisticated enough, we would find lots of problems.
One of the best ways to avoid 'lots of problems' is to run the first gain stage OPEN LOOP. Then you can elegantly control the risetime, reducing it to a 10us rise time square wave, and handle it from there with whatever you want.
This is the 'secret' of the Vendetta Research SCP-2, and any other hi end phono design that I make.
However, not ANY gain stage configuration will work optimally, any more than just any IC will work in the phono input, even usefully. If you want to hear differences in IC's, I recommend buying a JC-3 phono stage and trying different op amps, and listening to each of them. Take notes! Measuring the differences is whole other ball game, but trust me, the JC-3 works pretty well, because that is what I had to do, last year, just before release of the product. I am still looking for the world's best IC to put in that place (from a listening point of view). New ones come out every month or so, but alas! They do not show the schematic, and there could be all kinds of 'compromise' that would ultimately show up in actual listening tests.
Scott, your AD797 sounded OK. It was just twice the price of something else that sounded as good (although slightly different) Another op amp failed to sound good at all. I hope to try others in future, including some AD parts. IF the AD825 were significantly quieter, I would use it over all others. Thanks Scott, for providing me with these parts in mini-dip form, so many years ago.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.