Choosing of best sounding OP AMPs for the lowest possible THD+N -really the best Way?

When I first listened to my ES9038Q2M with a new (real I/V) output stage I was surprised at being able to distinguish a part of a multipart harmony that I had not noticed before. Before it was lost in the mix. In that case the stock board is pretty bad compared with the modified board.
 
What, precisely, is "half a voice", and how exactly did you distinguish it from one full voice?:D

It sounds like a musical interval, very slightly distorted in this case. An interval is two notes rather than one, thus there had to be two voices. The 'half' term was because the vocal notes are supposed to sound like two voices, not one blurred together interval texture, and perceived location in the sound stage was supposed to be different for each voice.
 
Last edited:
Well, what was the modification?

The stock board does not have a true I/V output stage. (And the analog supplies are noisy.) The stock board schematic (and the terrible implementation) is attached.

So I made a true I/V stage as discussed in the ES9038Q2M board thread. See this post and the "no hump" values: ES9038Q2M Board
 

Attachments

  • My Board Schematic.png
    My Board Schematic.png
    217.4 KB · Views: 450
Very many interesting notes.
The biggest disadvantage of all these considerations is that all exist distortions (THD+N) are lumped together in the diagrams and no distinction is made between THD components e. g. of a low order (inaudible up to around 1-2%) and those of a very high order, which always occur even with extremely low THD+N levels are still audible - go to fig. 3 under
http://www-f9.ijs.si/~margan/Articles/Class_B_Dist.pdf
(The fact that these diagrams refer to power amplifiers in non-Class A operation and not to small signal stages with OP amps doesn't matter for these fundamental considerations - so I think).
This was pointed out in detail in the already very old PDF papers in post #82 under
https://www.diyaudio.com/forums/sol...ooks-overview-google-books-9.html#post6707846
 
Nonsense, you just told that spent some tousands in tubes, and sound was poor, as the guys who know how to evalue the sound does put 10times less money end get remarkable results.
Btw, new opa's is not bad of course. :)
Forget AD797, OPA627, 275, 1028 and all the other OLD opamps. I used them for a long long time but they are old and we have new opamps with MUCH higher SQ.

There is NO WAY that one of these opamps sound better then the mentioned OPA1612 and other new ones like OPA2192 etc. and if they do it is the fault of the circuit or whatever. As Mark pointed out OPA1612 is a terrific sounding opamp and if you have problems with its bold mighty extreme low bass performance or with the extreme high resolution and unfiltered highs you can use other new opamps with a more gentle sound like OPA2209, OPA2189 etc.

BTW, last year I crippled my €4.000 (DIY!) tube preamp with tube regulators and ELCAP free external PSU weighing over 20kg full of Lundahl transformers & chokes. Cos the OPA1612 do sound better with normal PSU regulators not to mention Jung regs...
 
True, is that human have the habit to trust what he/she understand him/her self. It means we bind to todays known meaurement equipment, thd imd sn, whatever. But the listening is more important, as all these thd imd sn is just an esiest way to measure what our science succed now. While people binded to that - if not they would force to realise they worth nothing/hang up or something.
Science is depending of objectives and the scope of research, it depends of the open mind of researcher and knowlege of these days. If they found some rules it doesn't mean its absolute. As more we discover as much more left undiscovered. One more habit - leave in comfort zone 🤣.
Another way listen and musical precepcion that is not the science and is subjective for sure, but the subjects with good listenging memory and experience can reach better result than science does.
Also would big confuse base listening experience on professional musicians as they focus is different often. In listening sessions with professional musicians some evaluations was something like "yes this sysyem hits to beat, this also, etc. ". While lot of tallant music bands who we listen to have poor records(sure there goes talk about f.e. 70-90y of past century), also we have some few examples what did well, all of them was professionals.
What if you need someone to listen with their ears so that they can recommend a more involved/non-standard measurement to "detect" or quantity an undesired artifact (using their experience to make the recommendation)? (What if that someone is a customer who is using your parts in their product?)

What if your standard test rack and standard specifications can not do that measurement? And you need to make a new/modified test setup?


Telling them it is irrelevant rarely works out well.
 
I've come to ask for wisdom! I'm designing a little integrated board for my soon to be cheap DSP project (as a full time hobby, that is.) I'd like to experiment with op amp based differential signal generation. These are the op amps I have:

LM741 (CA741CE):
CA741CE Intersil - Distributors and Price Comparison | Octopart component search

LS141CB:
LS141CB STMicroelectronics - Distributors and Price Comparison | Octopart component search

NE5532P

Texas Instruments MC1458P:
MC1458P Texas Instruments - Amplifiers - Op Amps, Buffer, Instrumentation - Distributors, Price Comparison, and Datasheets | Octopart component search

Which one of these would be my best bet as a starting point? I will be comparing measurements with the classical transformer based differential.

Cheers!
 
Of the four you listed the NE5532 is by far the best op-amp for audio.

There are much better audio op-amps now but the NE5532 was used almost everywhere in pro audio for a very long time.

If you can you might want to try LM4562 or if you don't mind the cost perhaps consider OPA1612 and OPA2228.