Well in my limited controlled experience the OP-AMP including AD797 never proved to sound better than a discrete stage.
But to answer to this question, Joachim will definitely be more qulified as he worked extesively with both technology.
Designing discrete is more fun and i can get a better dynamic range and less noise.
That applies primary to the input stage. Where opamps major is low distortion, high gain and good DC precision. When i use opamps i bias then into class A and use a discrete output buffer. i also make sure that they do not slew. I also found that diffent opamps sound different. Just last year i build a phono stage for LÁrt du Son and Martina Schöner made a very detailed listening test to opamps. We made sure that non of them oscillated. That can really screw up the sound. Funny enough we had the same priorities and identified the same opamp as best sounding out of at least 10 different types.
Why that is is not explainable from distortion only. A well designed NE5534 stage has so low distortion that you can chalange even a good distortion analyzer. Especially with the class A trick and shunt feedback. Douglas Self has shown that. The only explanation i have is settling time and PSU rejection. The ear may also be able to differentiate materials and construction techniques. That has to be put in perspective though. The differences we heard may not be night and day and it is a bit like wine tasting. Quite small differences can come out stronger with more experience and listening time.
That applies primary to the input stage. Where opamps major is low distortion, high gain and good DC precision. When i use opamps i bias then into class A and use a discrete output buffer. i also make sure that they do not slew. I also found that diffent opamps sound different. Just last year i build a phono stage for LÁrt du Son and Martina Schöner made a very detailed listening test to opamps. We made sure that non of them oscillated. That can really screw up the sound. Funny enough we had the same priorities and identified the same opamp as best sounding out of at least 10 different types.
Why that is is not explainable from distortion only. A well designed NE5534 stage has so low distortion that you can chalange even a good distortion analyzer. Especially with the class A trick and shunt feedback. Douglas Self has shown that. The only explanation i have is settling time and PSU rejection. The ear may also be able to differentiate materials and construction techniques. That has to be put in perspective though. The differences we heard may not be night and day and it is a bit like wine tasting. Quite small differences can come out stronger with more experience and listening time.
Are you saying in other words that a more dynamic and quiter sound can be obtained by using discrete thus achieving an overall better sound?
That would match my little experience too and that might be the reason why lately I have audityioned some serious different very expensive phono stage.
None of them really blown me away.
Maybe the reason has to be researched on the technology used, maybe they likely use OP-AMPs inside, just guessing.
Regarding different op-amps sounding differently, I have noticed that and it's just like other things such as resistors and capacitors: 99.9% of the time the difference is unmeasurable but the ear can easily detect it.
Some time difference is not day and night other times it is. Just thinking of comparing an expensive good signal cable vs. a "black/red" type of standard chord: difference there is really day and night.
That is so interesting to me and IMHO it is part of the game to dial everything right in to get best sound out of it.
That would match my little experience too and that might be the reason why lately I have audityioned some serious different very expensive phono stage.
None of them really blown me away.
Maybe the reason has to be researched on the technology used, maybe they likely use OP-AMPs inside, just guessing.
Regarding different op-amps sounding differently, I have noticed that and it's just like other things such as resistors and capacitors: 99.9% of the time the difference is unmeasurable but the ear can easily detect it.
Some time difference is not day and night other times it is. Just thinking of comparing an expensive good signal cable vs. a "black/red" type of standard chord: difference there is really day and night.
That is so interesting to me and IMHO it is part of the game to dial everything right in to get best sound out of it.
Yes, i think a phono with less noise and hum sounds better.
I even would say that noise, hum and distortion is not allowed. The opamps get the distortion part right so i use them for comparison.
Also some of the never opamps have very good transistors in there that you do not get as discrete parts. For example the germanium-silizium process in the new Burr-Brown opamps like OPA1641 and OPA1611. That kind of transistors allow simpler topology structures to be made. Ohh, i forgot, i think simpler sounds better usually WHEN the circuit does the job, so no compromise in noise, hum and distortion. Concerning precision of the RIAA i think that matching between channels is more important then say a plus-minus 0.1dB result although that results in good matching too by default.
MiiB and i even decided on contouring the RIIA of the Paradise with a small lift in the deep bass, say plus 0.3 to 0.5db under 200Hz. We found that this way more records started to sound acceptable. What i really do not like is a rising response towards the treble.
I even would say that noise, hum and distortion is not allowed. The opamps get the distortion part right so i use them for comparison.
Also some of the never opamps have very good transistors in there that you do not get as discrete parts. For example the germanium-silizium process in the new Burr-Brown opamps like OPA1641 and OPA1611. That kind of transistors allow simpler topology structures to be made. Ohh, i forgot, i think simpler sounds better usually WHEN the circuit does the job, so no compromise in noise, hum and distortion. Concerning precision of the RIAA i think that matching between channels is more important then say a plus-minus 0.1dB result although that results in good matching too by default.
MiiB and i even decided on contouring the RIIA of the Paradise with a small lift in the deep bass, say plus 0.3 to 0.5db under 200Hz. We found that this way more records started to sound acceptable. What i really do not like is a rising response towards the treble.
By the way you can make an ultra low noise MC stage with opamps too by just paralleling
many like we do with discrete. I have shown stages like that with 8 in parallel and called that "Dolphin". I have shown that on the MPP thread.
For me an opamp is just a building block, like a resistor, a cap or a discrete transistor.
Progress comes in my book if you understand what the part does. i call that decoding.
many like we do with discrete. I have shown stages like that with 8 in parallel and called that "Dolphin". I have shown that on the MPP thread.
For me an opamp is just a building block, like a resistor, a cap or a discrete transistor.
Progress comes in my book if you understand what the part does. i call that decoding.
Bob Cordell has put so some additional information about his phono stage on his website:
CordellAudio.com - VinylTrak™ Preamp
CordellAudio.com - VinylTrak™ Preamp
This is really interesting stuff.
So is the phono completely discrete? and has a SE input and then transforms it inot balance?
I have looked for the Issue. The number 0 can be purchased on Amazon while they don't have number 4.
The website doesn't sell the issue.
You know where to purchase those?
Hi Everybody!
MasterPiece is back.
I have been working a little bit today on it and after some thoughts I have come out with a modified topology N-CH only.
To my eyes it looks pretty good and slick both absolute and relative offset are taken in care of by the servos.
The only thing I am not sure at this point is if we can reach out the needed gain with IF3602.
Technically they have more than 20 times the gain of 2SK170.
Unfortunately I don't have a spice model for it at the moment (if anybody has a clue on how to obtain them please let me know).
Meanwhile, please feel free to comment on it. This is a start up schematic and I am sure a lot of adjustments will need to be made.
MasterPiece is back.
I have been working a little bit today on it and after some thoughts I have come out with a modified topology N-CH only.
To my eyes it looks pretty good and slick both absolute and relative offset are taken in care of by the servos.
The only thing I am not sure at this point is if we can reach out the needed gain with IF3602.
Technically they have more than 20 times the gain of 2SK170.
Unfortunately I don't have a spice model for it at the moment (if anybody has a clue on how to obtain them please let me know).
Meanwhile, please feel free to comment on it. This is a start up schematic and I am sure a lot of adjustments will need to be made.
In the simulation it looks like I can't get more than 50 dB or the gain will collapse.
I am wondering if that is a glitch of the simulator or what.
I mean if you think that with 2 pairs of interfet we can reach out 63dB-ish that would be great.
Noise-wise they are significanlty quiter than any other fet I know of, so that should produce a very quite phono stage hopefully.
You said it is a starting point, if you have any recommendation at this point please let me know.
Speaking of which I have checked a little bit online and it seems that the interfet is like a part that you get with a high minimum order.
If that is true that would make impossible for me to try this part out.
Hopefully samples can be purchased at least 10 of them.
I left a voicemail to the sales agent and see what they say.
If that is true that would make impossible for me to try this part out.
Hopefully samples can be purchased at least 10 of them.
I left a voicemail to the sales agent and see what they say.
I am playing with paralleling more BFs which should simulate the behaviour of the IF part.
It looks like the DC offset goes out of control if I try to raise gain and I can't give an explanation to it either.
If anybody has a clue please let me know!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.