Op-amps vs Discrete for preamps/DACs

[kop89]

With the availability of good op-amps like LM4562 today, would it still make sense to design using discrete circuits for preamp and DAC output stage?

[Conrad Hoffman]

There are those who think discrete circuits will always be better, and you can certainly build some darn impressive discrete circuits. Then there are those who think today's best op-amps leave little to be desired, and that discrete versions add nothing but cost and complexity. Like so many things in audio, no resolution is expected in the near term.

[Panelhead]

Used to think that all solid state sounded terrible. Then decided that the problem was IC's and A/B output stages. Today I use a couple discrete phono stages and disc players.
My favorite amps use a bipolar opamp and power chips. I know discrete is better, but no one who has ever listened guessed the pedigree. I fact at a audio get together several years ago, many asked how I squeezed tubes into small aluminum boxes.
So I guess what this says is it is the circuits, layout, and power supply that has the major impact on sound quality. The parts and devices used are audible, and have an effect. But good sound is available from almost all devices if the first three are right.

George

[nigelwright7557]

Originally posted by Panelhead
Used to think that all solid state sounded terrible. Then decided that the problem was IC's and A/B output stages. Today I use a couple discrete phono stages and disc players.
My favorite amps use a bipolar opamp and power chips. I know discrete is better, but no one who has ever listened guessed the pedigree. I fact at a audio get together several years ago, many asked how I squeezed tubes into small aluminum boxes.
So I guess what this says is it is the circuits, layout, and power supply that has the major impact on sound quality. The parts and devices used are audible, and have an effect. But good sound is available from almost all devices if the first three are right.

George

After growing up in an environment where only valves were around I too didnt think much of semiconductors when they arrived. Valves just sound so much warmer and better.

However once I got into building my own disco gear in the 1980's I found valve amplifiers to be very expensive so had to look at a semiconductor amplifier.

Today I have compromised a bit and use a valve pre amp/mixer with a 1000WRMS MOSFET amplifier. The valve adds some warmth to the sound and the big amp really puts out a lot of power into my big speakers.

[audioman55]

National new line of LME current feedback, LME49713, and voltage feedback LME49710 opamps at around .00005% THD+N are better than just about any discrete design could be. The matched NPN and PNP transistors in this high voltage process have identical physical dimensions and carrier flow. Get some and try them out. Also the metal cans measure the same as the dips but exhibit better sound quality in our blind tests.

They also make great power supply regulators. The LME49713 has 100 ma of output current so it will work directly for most preamp level circuits. The LME49710 has 26 ma so using it with the LME49600 250 ma buffer (built on the same process) Turns it into a 250ma opamp which makes it a great part for an analog power supply design.

I recently built a demo D/A preamp using all these parts (all metal cans for the signal path, AND 7 power supply regulators, in the best one) and it is quite remarkable! (Others manu's who are now using the part in new designs and heard the demo unit said that...but I think so too.)

Best Regards Everyone,

Audioman55 / Mark