How important are the small capacitor values in DOA design, example: 2520

I'm building a number of different opamps and despite my super careful planning I missed a few values on my mouser order.  I've reordered the correct parts, but I was wondering how important certain parts are.  For this example I'll use this schematic of a 2520 that I found in a different thread, that happens to be very similar to the GAR2520. 

1.  C4 is 39p.  Would a 47p work there just fine?  What about a 100p?

2.  Output resistors. Each build seems to have a different output resistor value, and they are all with in 5r of each other.  This one is 5.6r, another is 8.2r, several are 10r.  Does this small amount really make a difference?  If I used 10r in this 2520 would I blow something up, or would it function just fine?



http://web.qx.net/jgreenlee/API/2520bc.JPG

every thing matters...

larger cap would slow down/reduce HF gain

10 ohms in emitter degeneration would reduce the class A current so increase crossover distortion

whatever... it will still make sound and shouldn't release smoke...  :

JR

JohnRoberts said:

every thing matters...

larger cap would slow down/reduce HF gain

10 ohms in emitter degeneration would reduce the class A current so increase crossover distortion

whatever... it will still make sound and shouldn't release smoke...  :

JR

Click to expand...

So, let me get this straight.... the values in there are all well thought out and by design???  ;D
Sounds like I'll just stick to the schematic!  As a follow up question: What if, in the DOAs that have 10R, it was replaced with 5.6R?

Do you know of any guides or literature that explains the basics of DOA circuitry, so I could avoid asking silly questions on the internet?

dysenterygary said:

So, let me get this straight.... the values in there are all well thought out and by design???  ;D
Sounds like I'll just stick to the schematic!  As a follow up question: What if, in the DOAs that have 10R, it was replaced with 5.6R?

Do you know of any guides or literature that explains the basics of DOA circuitry, so I could avoid asking silly questions on the internet?

Click to expand...

There was a white paper in the IEEE journal about op amp design back in the early 70s (?)  when they designed the first unity gain stable IC op amp (741 a big deal back then).  Perhaps a little beyond what you are expecting.

JR

dysenterygary said:

What if, in the DOAs that have 10R, it was replaced with 5.6R?
Do you know of any guides or literature that explains the basics of DOA circuitry, so I could avoid asking silly questions on the internet?

Click to expand...

Douglas Self discusses the effect of emitter resistors in his power amplifier book.

Regarding 2520's specifically, some of the cloners pay  little or no attention to bias thermal tracking;
point being larger value emitter resistors provide a measure of stability despite increased distortion.

JohnRoberts said:

There was a white paper in the IEEE journal about op amp design back in the early 70s (?)  when they designed the first unity gain stable IC op amp (741 a big deal back then).  Perhaps a little beyond what you are expecting.

JR

Click to expand...

https://ieeexplore.ieee.org/document/1051851

Is this the paper you're referring to?  If so I'll see if I can find the full copy.  No guarantees I'll understand it, but I'll probably know some of the words. 

Speaking of words, as a self taught hobbyist, sometimes just learning the correct word or phrase to search is all the help I need.  For instance, I've never heard the term "emitter resistor" before you mentioned it.  I've searched and searched for information on "output resistors" and could never find what I was looking for.  One quick search of "emitter resistor" and I have my whole weekend of reading planned out!  So, referring back to my first post/question about the 39p capacitor; is there a specific term for that circuit with the 39p and 20k resistor in parallel? 

Thanks again!

dysenterygary said:

https://ieeexplore.ieee.org/document/1051851

Is this the paper you're referring to?  If so I'll see if I can find the full copy.  No guarantees I'll understand it, but I'll probably know some of the words. 

Click to expand...

no... maybe 10 years before that

2. J. E. Solomon, "The monolithic op amp A tutorial study", IEEE J. Solid-State Circuits, vol. SC-9, pp. 314-332, Dec. 1974.

This will still not teach you about discrete design,,, it was about overcoming limitations in early IC design.

Do not waste you time with an IEEE paper, there are other more accessible texts about general discrete design.

Speaking of words, as a self taught hobbyist, sometimes just learning the correct word or phrase to search is all the help I need.  For instance, I've never heard the term "emitter resistor" before you mentioned it.  I've searched and searched for information on "output resistors" and could never find what I was looking for.  One quick search of "emitter resistor" and I have my whole weekend of reading planned out!

Click to expand...

Perhaps research "emitter degeneration".. Transistors have high transconductance (very small Vbe changes result in large current changes). Adding a resistor in series with the emitter mitigates the high transconductance as the voltage drop from the current in the series resistor reduces the voltage change at the base-emitter junction.  This technique is used in output stages to reduce runaway current (Transistor Vbe drops with increased temperature, increasing the current draw with fixed voltage drive. Emitter degeneration resistors can offset that. ). Of course its more complicated than that.   

So, referring back to my first post/question about the 39p capacitor; is there a specific term for that circuit with the 39p and 20k resistor in parallel? 

Click to expand...

The general concept you need to look at is "dominant pole compensation" for op amps.

I gave some general advice about op amps in this almost 40 yo article. http://www.johnhroberts.com/des_art_1.pdf

Thanks again!

Click to expand...

Nothing wrong with being self taught... I've learned a lot that way.  8)

JR

JohnRoberts said:

Do not waste you time with an IEEE paper, there are other more accessible texts about general discrete design.

Click to expand...

Can you recommend some?

As a fellow diy-er with no formal education in electronics, my approach has been to hold to well known opa designs as gospel for now, and try to get at what schematics do around them that makes for optimal circuits.  Which is to say, watcha making with your inspired 2520's DG?

user 37518 said:

Can you recommend some?

Click to expand...

Sorry, I am sure other forum members here will have many good suggestions.

JR

PS: while not a 2520  Deane Jensen RIP  wrote an AES paper about his DOA back in the '80s. https://www.technicalaudio.com/pdf/Jensen_Transformers/Jensen_OpAmps_990_and_related/Jensen_JE-990_opamp_JAES_reprint_1980.pdf

IMO this 990 may be the better op amp but who's comparing?  These days you can buy IC op amps off the shelf that spank them both.  Some of the books listed in Deane's footnotes are on my own bookshelf, but not exactly DIY how to books.

JohnRoberts said:

  These days you can buy IC op amps off the shelf that spank them both. 

Click to expand...

Which op amp can output +25dBu into 75 ohms?

Cheers

Ian

dysenterygary said:

1.  C4 is ....2.  Output resistors. ...

Click to expand...

Some answers you can get here:
http://www.nanovolt.ch/resources/discrete_opamps/pdf/SGA-SOA-1_documentation.pdf

For more in depth information, I suggest this book:
Bob Cordell - Designing Audio Power Amplifiers

DOA is just a small power amp.

ruffrecords said:

Which op amp can output +25dBu into 75 ohms?

Click to expand...

In which audio application do you need that?

moamps said:

Some answers you can get here:
http://www.nanovolt.ch/resources/discrete_opamps/pdf/SGA-SOA-1_documentation.pdf

Click to expand...

+1 I forgot about Sam's work on that

For more in depth information, I suggest this book:
Bob Cordell - Designing Audio Power Amplifiers

DOA is just a small power amp.

Click to expand...

Smaller and far easier, but yes the same principles apply.

JR

ruffrecords said:

Which op amp can output +25dBu into 75 ohms?

Cheers

Ian

Click to expand...

http://www.ti.com/lit/ds/symlink/lm675.pdf

3A and 60V rail

More like for driving 4-8 ohm loudspeakers.

This suggests it could be done if there was a large enough market. 

Back when you and I were both young, elevated voltage swing was used to get above marginal noise floors. These days things have gotten much quieter, negating the need for running hot signals.

Most commonly used A/D convertors run from 5V rails (some 3.3V) and have modest input impedance.

JR

PS: I haven't rolled my own DOA since about the 80s, and only did it once for a console summing amp. Since it was dedicated for use as a virtual earth summer, it was probably horrible in non-inverting mode.

JohnRoberts said:

elevated voltage swing was used to get above marginal noise floors. These days things have gotten much quieter, negating the need for running hot signals.

Click to expand...

and your evidence is ?

no surprise to some folks on this forum, high headroom sells.
a few  actually listen to the signal chain before printing to that 5 volt A2D.

gridcurrent said:

and your evidence is ?

Click to expand...

decades of experience...

API was known for high voltage discrete design and marketed their "high headroom" as a benefit. Back in the 80s I hired a marketing consultant with a recording  background who swore by that bus headroom benefit...( the API marketing voodoo was strong). It doesn't take advanced math to vet that (everything is relative).

Back in the 80's I wrote a column (for my "audio mythology" series in RE/P magazine) comparing +4dBu vs -10dBV platforms. The -10dBV gear did surprisingly well for headroom, while most -10dBV manufacturers cut corners with lower rails, and slower chips (the slower chips were easier to tolerate with lower nominal path levels requiring less slew rate all else equal).  Without modern (for  30 years ago) low noise ICs the -10dBV category wouldn't be practical.

no surprise to some folks on this forum, high headroom sells.

Click to expand...

indeed...  so do other popular buzz words like tubes, transformers, DOA, yadda yadda.

Curiously Mackie resurrected the "high headroom" screed to sell mixers a decade or two later (they did run some mix buses -6dB cooler, and accidentally  : avoided rigorous clip detection on their input strips). They had many drinking that marketing kool-aid, until made in USA lost cost competitiveness with similar Chinese made SKUs. 

a few  actually listen to the signal chain before printing to that 5 volt A2D.

Click to expand...

Huh?

Hopefully they listen to the final result, but I trust bench measurements more than my ears.

JR

PS sorry about the veer....

moamps said:

Some answers you can get here:
http://www.nanovolt.ch/resources/discrete_opamps/pdf/SGA-SOA-1_documentation.pdf

For more in depth information, I suggest this book:
Bob Cordell - Designing Audio Power Amplifiers

DOA is just a small power amp.

Click to expand...

Moamps, thank you!  This is very informative!

dysenterygary said:

So, let me get this straight.... the values in there are all well thought out and by design???  ;D

Click to expand...

Yes. But it doesn't mean that a different value will not work. It may be a different optimization. The lowest the emitter resistors, the higher the output capability...up to a point, where another component gives up and probably releases its magic smoke.
The small cap that defines the roll-off may be increased; it may result in increased HF distortion, but will be potentially more stable.