Discrete Opamp Open Design

[jan.didden]

Originally Posted by scott wurcer

PMA's circuit looks like two CFA inputs looking at eachother, it lends itself very well to full differential connection.

For that matter that simplified 844 made with discrete duals could be tuned up nicely for several applications. I think there are cheap SMT dual's in the BXXXX series. That's 8 SOT6's and an output pair and you have access to all nodes so you play with Charles Hansen's ideas.

Some 8 years ago I had a serious project with a retired IC designer to get a couple of AD844 bare dies, interconnected in a smart way, packaged in a 20-pin DIL.
Total bill was supposed to be some 12k$ for few 1000 chips.
Backed out at the last instant because, no matter how good it would be, it would still be shot down if someone would thoughtlessly use the word 'IC' somewhere

"I'm following this thread with interest".

jan

[jneutron]

Originally Posted by bcarso

One such part: DMMT3906W, in SOT-363: http://www.diodes.com/datasheets/ds30312.pdf

I see there are some matching specs.

Brad

Nice that they are adjacent die.

jn

[RNMarsh]

Originally Posted by Wavebourn

I have to admit, I like Scott's exercise. Quite elegant.

A beautifully done high gain circuit..... with elegance and simplicity.

[ BTW -- th jfets are still available on internet. At least 6 months ago when i bought some]

[Wavebourn]

Originally Posted by john curl

Really Wavebourn, are you sure?

Absolutely.

[john curl]

I'm not so sure, but I could be convinced with a simulation that showed lots and lots of 2'nd harmonic with matched devices.

[RNMarsh]

Originally Posted by Kindhornman

john curl,
Your so practical, and this discussion doesn't seem to have anything to do with practicality. I guess we can just reinvent the wheel until we get a square one and then decide that it won't go around any better than what exist, but that seems to be the point of this project. Just to try something and see what happens.

You missed a few theads leading up to this --> Actually, it is also about discrete low gnfb vs. IC or high gnfb. Most IC are high gain circuits which depends on high nfb for low distortion. In some audio circles high gnfb has been out of favor for many years. Scott's discrete but high gain design is to be compared and listened to IC and other circuits.
-RNM

[Wavebourn]

Originally Posted by john curl

I'm not so sure, but I could be convinced with a simulation that showed lots and lots of 2'nd harmonic with matched devices.

John, no need for simulation. Use your common sense. It is trivial. I can understand if such solution was proposed as basic block for some digital device, but not for analog amp, especially when we speak about high end.

[scott wurcer]

Originally Posted by john curl

I'm not so sure, but I could be convinced with a simulation that showed lots and lots of 2'nd harmonic with matched devices.

EVUL suggested we stick with cheap readily available parts. I think I could certainly drive 100 Ohms at 20V p-p and still be under 110dB seconds at 20kHz.

BTW that other circuit is free of seconds either.

[Wavebourn]

Originally Posted by RNMarsh

You missed a few theads leading up to this --> Actually, it is also about discrete low gnfb vs. IC or high gnfb. Most IC are high gain circuits which depends on high nfb for low distortion. In some audio circles high gnfb has been out of favor for many years. Scott's discrete but high gain design is to be compared and listened to IC and other circuits.

Can you translate it into English please, regarding Scott's design and NFB?

[scott wurcer]

Originally Posted by RNMarsh

You missed a few theads leading up to this --> Actually, it is also about discrete low gnfb vs. IC or high gnfb. Most IC are high gain circuits which depends on high nfb for low distortion. In some audio circles high gnfb has been out of favor for many years. Scott's discrete but high gain design is to be compared and listened to IC and other circuits.
-RNM

Yes I forgot to mention the circuit can be easily tuned with loading or degeneration for a large range of gnfb. As first presented it is as Wavebourn notices very much in the op-amp vein, very large Aol and very low frequency breakpoint. You'll have to leave me out of the listening though I tend not to hear any differences until they're gross.