John Curl's Blowtorch preamplifier

[syn08]

So?

Nothing really special. Just challenging Pavel's statements about op amps "dominant pole" "slew rate limited" etc... On a more general tone, challenging the "all discrete" approach in Anno Domini 2008. Which doesn't impact the merits and records set 20-30-40 years ago by John Curl et. al.

Edit: at the same cost and size, John? Ahem... And just in case you or anybody else is asking, yes, there's a lot to design in such a design. Starting with the compensation schema and ending up with the layout.

 

[john curl]

And this is what we do for a living.

 

[syn08]

And this is what we do for a living.

Which is perfectly fine John, it is truly amazing that your reputation still sells. Not many audio professionals ever had such a long career.

Unfortunately, the "doing for a living" sometimes conflicts with the technical arguments and reality. That's usually where the golden ear brigade comes to the rescue.

 

[john curl]

Specs are not really the same thing as good sound. I wish that it were so. You have much to learn.

 

[Steve Eddy]

Specs are not really the same thing as good sound.

An externally hosted image should be here but it was not working when we last tested it.

se

 

[PMA]

Ovidiu, which do you prefer, this:

(Dispre2)

 

[PMA]

or this?

(5532, same gain, same output)

/not a slew rate limitation for 5532/

OA behaviour is similar to CD jitter.

Edit: in fact conditions for 5532 are softer, input voltage stepped only to 2V, though for Dispre to 3V. Gain = 4.3x

 

[syn08]

Specs are not really the same thing as good sound. I wish that it were so. You have much to learn.

Q.E.D.

 

[syn08]

OA behaviour is similar to CD jitter.

Huh? You lost me here.

I have another one that slews 30V in 45nS but its not 100% opamp based and has only -110dB THD20 and noise 10nV/rtHz. Its built on microstrip I will post some pictures tomorrow.

Edit: do you really think I'm using 5532?

 

[john curl]

WELL DONE! PMA

 

[syn08]

WELL DONE! PMA

Q.E.D.

 

[scott wurcer]

Beat me to it.

 

[john curl]

That sure is helpful.

 

[john curl]

PMA, they are just frightened of a new measurement technique. Even the experienced designers, what a shame.

 

[scott wurcer]

You didn't ask for help. Here's an example of a discrete op-amp that you can't integrate. Note JE990 style inductors in input, but here circa 1966. The use of an explicit ground pin is now discouraged showing that even Dick Burwen didn't quite get it. Still can't find an earlier use of folded cascode as in that Harris op-amp.

 

[john curl]

Scott, please call Dick Sequerra, I think that he might 'talk' to you now. Let us hope.

 

[GK]

PMA, they are just frightened of a new measurement technique. Even the experienced designers, what a shame.

Yeah, nobody has ever measured PIM distortion before.

 

[scott wurcer]

Scott, please call Dick Sequerra, I think that he might 'talk' to you now. Let us hope.

What brings this on? Again.

 

[KSTR]

I for one can *not* exactly reproduce Pavel's "PIM"-graphs, but slightly different ones, still showing some zero crossing shift with increased amplitude.

Albeit with a different amp (F5-style, but with bjt diamondbuffer frontend, which doesn't matter much in this context). It is compensated for decreased open-loop gain above the audio band (pole/zero pair) to better handle cap loads (10nF used here). When I remove the compensation and cap load, I get the similar curve "distortion" as Pavel with his dispre vs. 5532, better correlated zero crossings and overall shifted in time.... until I zoom in further...

Test freq was 40kHz to get a clear picture , and drive from 10% to 100% (of just below clipping level) into 5.6R.

EDIT: I recognize my graph doesn't match Pavel's

- Klaus

 

[MikeBettinger]

PMA, they are just frightened of a new measurement technique. Even the experienced designers, what a shame.

For my primary measurement technique I'll rely on EAR sampling, it's old, but very high resolution. It actually aligns pretty well with the format adopted by the incoming QC of the end user.

Actually I'm the end user so this works out well.