If it were true, one need not bother with sota opamps either.
Secondly, there also a medium inbetween, whether LP or CD.
In my experience, opamps have become so good that an opamp design can outclass discrete ones.
Best sounding ones are still discrete builds though, but maybe not for long.
But Hey, the ones who can afford prime beef are usually of an age that it's irrelevant just the same for 9 out of 10, a lot of the real deal on the street is marketing related to cater the spend frenzy of the never happy few.
(frankly, if it wasn't so butt ugly, i'd fancy a Burmester 111 server in 5 years from now.
)
Last edited:
Back in 1978, I made a direct A-B comparison between a 5534 and a discrete module that I designed and built, 1 year earlier. IF the 5534 worked as hoped, I expected to not have to make any more discrete modules, except for special projects.
However, the discrete module won out, so I continued to make them, even though they were expensive to make, much larger in size and inconvenient to use. We also made 5534 replacement modules, composed of a 5534 with a discrete jfet front end. This worked pretty well, when we did not have the space for completely discrete modules. A hassle to make, but popular with some hi end manufacturers.
This is MY reality, not some ABX or other reality. However, every time I have ignored listening and depended on measurements, I have made designs that did not sound very remarkable. Modifying these 'failed' designs with 'subjective' techniques such as better passive parts, removed the sonic objections, and made them 'highly listenable'. Works for me, could work for many others.
However, the discrete module won out, so I continued to make them, even though they were expensive to make, much larger in size and inconvenient to use. We also made 5534 replacement modules, composed of a 5534 with a discrete jfet front end. This worked pretty well, when we did not have the space for completely discrete modules. A hassle to make, but popular with some hi end manufacturers.
This is MY reality, not some ABX or other reality. However, every time I have ignored listening and depended on measurements, I have made designs that did not sound very remarkable. Modifying these 'failed' designs with 'subjective' techniques such as better passive parts, removed the sonic objections, and made them 'highly listenable'. Works for me, could work for many others.
Bonsai:
What Borbely probebly refers to is microphony. Have a look at this post in the forum:
http://www.diyaudio.com/forums/analogue-source/158911-parasound-jc3-phono-66.html#post2554472
What Borbely probebly refers to is microphony. Have a look at this post in the forum:
http://www.diyaudio.com/forums/analogue-source/158911-parasound-jc3-phono-66.html#post2554472
Have to agree with John- ya gotta listen! If I hear something I don't like, I can usually find it on the test bench. OTOH, I'd never be sure that I'd covered all the bases with just measurements and could guarantee great sound with nothing more than that. Odds would be good, and there probably wouldn't be anything horribly wrong, but it's too easy to overlook something or, more likely, not know where to draw the line between good and evil 
Folks:
So we are getting some interesting opinions, and I hope the conversation remains cordial and fun.
John and Conrad:
I fully trust that you hear the sonic improvements with your discrete designs. This is great as it seems to roughly follow my experiences. And yes, my ears are the final judge, as they will be the end user!
From this point, what are you looking for test-wise for any form of correlation of your hearing perception to any test results? Independent reviews to confirm the sonics of your design are good. Over time I have been improving my ear training for the sonics of electronics. I am not a "golden ear", but know what I like.
Jacco:
At this point, I am thinking that I may be able to get very close to a discrete design by using the latest IC's. I'll get to building a preamp and listen to see if I can hear any differences. I'll then try to measure any differences.
To me, in the end, I feel sonic qualities should have some form of measured characteristics. Maybe we have not discovered them yet.
Thanks folks!
Dan
So we are getting some interesting opinions, and I hope the conversation remains cordial and fun.
John and Conrad:
I fully trust that you hear the sonic improvements with your discrete designs. This is great as it seems to roughly follow my experiences. And yes, my ears are the final judge, as they will be the end user!
From this point, what are you looking for test-wise for any form of correlation of your hearing perception to any test results? Independent reviews to confirm the sonics of your design are good. Over time I have been improving my ear training for the sonics of electronics. I am not a "golden ear", but know what I like.
Jacco:
At this point, I am thinking that I may be able to get very close to a discrete design by using the latest IC's. I'll get to building a preamp and listen to see if I can hear any differences. I'll then try to measure any differences.
To me, in the end, I feel sonic qualities should have some form of measured characteristics. Maybe we have not discovered them yet.
Thanks folks!
Dan
OK John, I'll tell you...
wait for it...
successful audio products have a very high margin! All the rest is just having the right mystique, plus some luck and guesswork.
Definitely OT but here are the RIAA preamps I've owned since the beginning of time, plus some comments.
Dynaco PAS-3X- Decent but limited by the rest of my crummy system at the time.
Dynaco PAS-3X with the published mods of the time- Not convincingly different though I was probably lured in by the descriptions in the mod articles.
Dynaco PAT-4- It was new and solid state- Harsh, I should have kept the PAS.
Ace Audio Zero Distortion (no buffer)- quite listenable 2-transistor circuit for the college student with no money. I still have the schematic- it would be interesting to hear one today.
Crown IC-150A- the absolute worst sounding preamp I've every had. Built nice but made my ears hurt.
--up to this point I was in my 20s and my hearing was good to about 20K- that's no longer the case.
National Semi LM394+LM318 circuit from the data book- Probably the best I'd had to that point. Gave it to my dad, who still uses it.
Audio Research SP-4- Solid state using potted mystery modules. Good, but I rarely use it with vinyl so can't say much with confidence. Does anybody know what's in those modules?
Crown IC-150- Yes, I bought another one. I couldn't help myself. It was just as bad as the first and I sold it too.
National Semi LME opamp design, similar to application notes but with traditional unsplit equalization. Of everything, this is the most transparent, natural and invisible preamp I've had to date.
In between these things I built and compared various circuits. I cloned several copies of the Ace Audio design, all of which were natural and smooth. My guess is the RIAA wasn't terribly accurate. I also listened to quite a lot, including the SP3A-1, so I did have some decent stuff to compare to, just no money to buy high end. In fooling with the LME design I tried just about every popular opamp from AD, Burr Brown, LT and National, and found the LME audibly superior to all of 'em. I have not built discrete RIAA circuits other than the Ace, only power stuff, so I'm no source of wisdom there. In general I go for minimum noise, minimum distortion and as accurate RIAA as possible. Simplicity, good quality components, clean power and decent layout attract me more than circuit complexity. The final arbitrator is an emotional response to certain LP tracks. FWIW
wait for it...
successful audio products have a very high margin! All the rest is just having the right mystique, plus some luck and guesswork.
Definitely OT but here are the RIAA preamps I've owned since the beginning of time, plus some comments.
Dynaco PAS-3X- Decent but limited by the rest of my crummy system at the time.
Dynaco PAS-3X with the published mods of the time- Not convincingly different though I was probably lured in by the descriptions in the mod articles.
Dynaco PAT-4- It was new and solid state- Harsh, I should have kept the PAS.
Ace Audio Zero Distortion (no buffer)- quite listenable 2-transistor circuit for the college student with no money. I still have the schematic- it would be interesting to hear one today.
Crown IC-150A- the absolute worst sounding preamp I've every had. Built nice but made my ears hurt.
--up to this point I was in my 20s and my hearing was good to about 20K- that's no longer the case.
National Semi LM394+LM318 circuit from the data book- Probably the best I'd had to that point. Gave it to my dad, who still uses it.
Audio Research SP-4- Solid state using potted mystery modules. Good, but I rarely use it with vinyl so can't say much with confidence. Does anybody know what's in those modules?
Crown IC-150- Yes, I bought another one. I couldn't help myself. It was just as bad as the first and I sold it too.
National Semi LME opamp design, similar to application notes but with traditional unsplit equalization. Of everything, this is the most transparent, natural and invisible preamp I've had to date.
In between these things I built and compared various circuits. I cloned several copies of the Ace Audio design, all of which were natural and smooth. My guess is the RIAA wasn't terribly accurate. I also listened to quite a lot, including the SP3A-1, so I did have some decent stuff to compare to, just no money to buy high end. In fooling with the LME design I tried just about every popular opamp from AD, Burr Brown, LT and National, and found the LME audibly superior to all of 'em. I have not built discrete RIAA circuits other than the Ace, only power stuff, so I'm no source of wisdom there. In general I go for minimum noise, minimum distortion and as accurate RIAA as possible. Simplicity, good quality components, clean power and decent layout attract me more than circuit complexity. The final arbitrator is an emotional response to certain LP tracks. FWIW
Conrad, your subjective experiences are similar to mine. I owned a PAS-3(X) for 10 years before replacing it with the Levinson JC-2. The PAT 4 sucked! As did the Stereo 120. I have had (and still have some) Dyna MK2,3,4 and Stereo 70. All good,but not 'great' products.
Others of my acquaintance called it the ICK 150, and you know why. But, could never figure out why everybody thought it sounded so lousy by just looking at the schematics.
You are right, EMOTIONAL response is the key. Either you get it seriously or you don't. That is why I like many competitor's vacuum tubes. The AR D-150 tube amp was wonderful with STAX headphones, for example.
Others of my acquaintance called it the ICK 150, and you know why. But, could never figure out why everybody thought it sounded so lousy by just looking at the schematics.
You are right, EMOTIONAL response is the key. Either you get it seriously or you don't. That is why I like many competitor's vacuum tubes. The AR D-150 tube amp was wonderful with STAX headphones, for example.
Have to agree with John- ya gotta listen! If I hear something I don't like, I can usually find it on the test bench. OTOH, I'd never be sure that I'd covered all the bases with just measurements and could guarantee great sound with nothing more than that. Odds would be good, and there probably wouldn't be anything horribly wrong, but it's too easy to overlook something or, more likely, not know where to draw the line between good and evil
I agree!
Hi,
my experience seems to be similar to Conrads and Johns.
I prefer rather simple JFET based circuits running in pure class-A and if possible without global feedback but just local degenerative feedback.
While IC-based circuits present music in a way that I never forget listening to a technical system playback, the discrete circuits offer a quality of authenticity and Life-likelyness. I also made the experience that many dynamic speakers are really incapable to show the difference. With very high resolving systems like ESLs the differences become obvious within seconds.
I use ICs where high gain and an amplitude response similar to the ICs openloop response is needed, such as DC-servos and vinyl-playback and where compactness is an issue.
To a degree I disagree with Jacco about the quality of modern ICs.
Crossover distortion is the one that can´t be corrected for using a feedback loop. Now the aim of the chip manufacturers seems to be to win a ´efficiency race´. Modern ICs consume less current than many discrete circuits input-stages alone, let alone the Vas-stages. While this is generally a most honourable aim, its not the paradigm in audio playback, at least not for home hifi.
jauu
Calvin
my experience seems to be similar to Conrads and Johns.
I prefer rather simple JFET based circuits running in pure class-A and if possible without global feedback but just local degenerative feedback.
While IC-based circuits present music in a way that I never forget listening to a technical system playback, the discrete circuits offer a quality of authenticity and Life-likelyness. I also made the experience that many dynamic speakers are really incapable to show the difference. With very high resolving systems like ESLs the differences become obvious within seconds.
I use ICs where high gain and an amplitude response similar to the ICs openloop response is needed, such as DC-servos and vinyl-playback and where compactness is an issue.
To a degree I disagree with Jacco about the quality of modern ICs.
Crossover distortion is the one that can´t be corrected for using a feedback loop. Now the aim of the chip manufacturers seems to be to win a ´efficiency race´. Modern ICs consume less current than many discrete circuits input-stages alone, let alone the Vas-stages. While this is generally a most honourable aim, its not the paradigm in audio playback, at least not for home hifi.
jauu
Calvin
"Modern" is a relative word, first time i bought BB627/637 was 1990, at $30/pc (5 up order), similar story with familiar numbers by AD and LT, high gain parts as LT1028 around '87.
Current specs of an LME49710 are not much different than the ones of an NE5534.
The balanced opamp preamp circuit i botched together with LM4562 samples on copper clad didn't sound all that inferior to my +25W total power consumption all-discrete balanced preamp with 4-layer amp boards (both <10ppm thd+n, me ESL hugger too).
Discrete is bound to loose in the end, as there has been little to none development of discrete components in Japan for the last 20 years, and parts continue to become obsolete.
Most folks don't realise how long components as J74/J109/K170/K389 have been manufactured.
Sanken A1216/A1295/C2922/C3264 for 30 years on end, name me a single thing that has been around un-altered for such a long time.
Ummm . . . uA741?? (Do you suppose that's because Fairchild got it right the first time, and it doesn't need any improvement? )
Dale
Hi,
300B Tube. Introduced in 1937 following the earlier 300A model, manufacturerd unchanged by Westrex/AT&T untill 1988 (51Years) and remaining in production in 2011 from a number of second sources and resurrected WE, for a total of 74 years and still going strong.
Among discerning people who don't give it hoot what it measures like, Amplifiers build around the 300B, including ones based on circuitry as old as the 300A Tue are still considered "State of the Art".
Ciao T
Ummm, . . . i can also think of bricks being manufactured longer than 40 years.
uA741 isn't exactly state of the art, name me something that is and for so long.
300B Tube. Introduced in 1937 following the earlier 300A model, manufacturerd unchanged by Westrex/AT&T untill 1988 (51Years) and remaining in production in 2011 from a number of second sources and resurrected WE, for a total of 74 years and still going strong.
Among discerning people who don't give it hoot what it measures like, Amplifiers build around the 300B, including ones based on circuitry as old as the 300A Tue are still considered "State of the Art".
Ciao T
possibly the small signal fets are "the same" - but most "classic" semis, power bjt, op amps have been redesigned several times as process lines become obsolete, are replaced by newer lines having better controls, with die shrinks where practical for greater economy - you can see "process change notices" at some manufacturer's sites - they try to maintain "critical" spec "equivalence" but power Q die shrinks can be seen in smaller capacitances, presumably power dissipation limits in the smaller die are compensated by thinner Si, better thermal path design
I have heard people complaining that they can't get true equivalents to replace "period" steel case 2N3055 for Neve mixing desk restorations
current tube production generally isn't believed to match NOS from the day - are there any tube lines with continuous production history and personel continuity to assure transfer of unwritten production "know how"
I have heard people complaining that they can't get true equivalents to replace "period" steel case 2N3055 for Neve mixing desk restorations
current tube production generally isn't believed to match NOS from the day - are there any tube lines with continuous production history and personel continuity to assure transfer of unwritten production "know how"
Back to Dan...
Dan, I suspect that you are hearing the noise in your line stage and other amplification stages, not the Borbely. The noise on the Borbely ought to, or is most likely to be subatantially below the surface noise of your LP given the fact that you are using a MM highish output cartridge.
The difference in noise between one pair and two pairs (complementary) jFETs is measurable, but very small, and is mostly of interest when one is using a MC cartridge.
The extra noise in one channel needs to be addressed, but given your description, something else seems like it is the source of your noise...
Now, DAN, you really need to get something better than what ur listening to and on... maybe the best reason is one that you yourself have shown yourself! That is that you have heard the difference between the Borbely and the opamp phono stages even on what you have to listen to!
May I be so bold as to suggest some ESL (Stax?) headphones and a small pure class A amp to drive them?? I know your discretionary budget can handle the hit... go for it!
_-_-bear
Dan, I suspect that you are hearing the noise in your line stage and other amplification stages, not the Borbely. The noise on the Borbely ought to, or is most likely to be subatantially below the surface noise of your LP given the fact that you are using a MM highish output cartridge.
The difference in noise between one pair and two pairs (complementary) jFETs is measurable, but very small, and is mostly of interest when one is using a MC cartridge.
The extra noise in one channel needs to be addressed, but given your description, something else seems like it is the source of your noise...
Now, DAN, you really need to get something better than what ur listening to and on... maybe the best reason is one that you yourself have shown yourself! That is that you have heard the difference between the Borbely and the opamp phono stages even on what you have to listen to!
May I be so bold as to suggest some ESL (Stax?) headphones and a small pure class A amp to drive them??
I know your discretionary budget can handle the hit... go for it!_-_-bear
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.