Frustrated with ASR's view on op-amps

[mhardy6647]

Let the record show (and I am going to drift, briefly, on topic) -- I don't even roll tubes... and heaven knows I could.
As has been pointed out here, most explicitly (in my memory) by @SIY, it is perfectly possible to design and construct a vacuum tube audio amplifier circuit that is very sensitive to minor variations of the electrical characteristics of an individual vacuum tube. Such a circuit may well perform demonstrably differently with individual vacuum tubes.

The industry that designed and manufactured vacuum tube devices for a living worked diligently and with outstanding skill to minimize or eliminate the sensitivity of such circuits to minor tube-to-tube variation -- not to mention age-related drift in vacuum tube performance.

ASIDE: Consider the requirement of very broad-band RF tuners, oscillators, detectors, clippers, demodulators, sync, chroma, sweep, deflection, high voltage, and other circuitry for the amazing (early 1950s!) all-analog NTSC monochrome-compatible color over-the-air broadcast television system. Early color TVs had on the order of two dozen tubes in therm (some of which contained multiple active elements). The circuits (many of them tuned) in them had to be tolerant of significant variation and aging of the active componenets.

I reckon that it is perfectly possible to design a touchy enough audio circuit around an IC op amp (or a discrete one!) such that different op amps would perform markedly differently. I believe the goal of good engineering is to minimize or eliminate such idiocyncratic performance in an electronic appliance.

PS ... and never forget that vacuum tube operational amplifiers are always an option! OK, now I am being snarky again -- and I shall stop.

 

[Robin L]

For me, beats Taylor Swift, any day of the week!

Funny, got my first Taylor Swift CD just yesterday (Folklore). Seems as though my op-amps were designed around her arrangements, but not her voice. Odd.

 

[Joachim Herbert]

ASIDE: Consider the requirement of very broad-band RF tuners, oscillators, detectors, clippers, demodulators, sync, chroma, sweep, deflection, high voltage, and other circuitry for the amazing (early 1950s!) all-analog NTSC monochrome-compatible color over-the-air broadcast television system. Early color TVs had on the order of two dozen tubes in therm (some of which contained multiple active elements). The circuits (many of them tuned) in them had to be tolerant of significant variation and aging of the active componenets.

Hence Never Twice Same Color.

 

[mhardy6647]

Hence Never Twice Same Color.

Dang.
Rolling on the floor again.
Twice in one morning.
I am an old-ish person; I am gonna hurt myself.
Pull a ligament or something.

 

[DVDdoug]

fwiw, I listen to mostly jazz / big band / vocal pop.

Sadly they don't make an op-amp for that use.....

Amir is being sarcastic, or snarky, or something. He's been showing his sense of humor lately!

But probably 99% of "audiophiles" would take it seriously. Most audiophiles are nuts! ...Usually "not believing in" blind listening tests or measurements. This is one of the very-few rational-scientific audio-related resources.

You've probably seen these links before, but:
What is a blind ABX test?
Audiophoolery

Despite the goofy name, Audiophoolery is a serious article that describes the few REAL characteristics of sound quality. That should help you to avoid the thousands of undefined audiophile terms like "detail", or 'clarity" that seem to mean something, but really don't.

and can clearly hear that after all the ones I tried... My ears are not lying to me.

"Clearly" you are NOT doing proper level-matched blind listening tests with statistically valid results.

And there is no logical reason for an op-amp change to cause "boomy" bass or "bright" sound, etc. Those are simply not the kind of defects or changes you'd get.

 

[DVDdoug]

Let the record show (and I am going to drift, briefly, on topic) -- I don't even roll tubes... and heaven knows I could.

I haven't even SEEN a tube in decades! I owned a couple of tube amps into the 80s but I hadn't used them for a long time and I gave the away. One of them was a MacIntosh... "perfect sound" and probably valuable now, but it was mono and I only had one.

MacIntosh amps were rock-solid designs so I'm sure "rolling" the tubes would make no difference (unless you use the wrong tube) and the amp would work perfectly as the tubes age until the tubes die or go out-of-spec.

 

[wadude]

Let the record show (and I am going to drift, briefly, on topic) -- I don't even roll tubes... and heaven knows I could.
As has been pointed out here, most explicitly (in my memory) by @SIY, it is perfectly possible to design and construct a vacuum tube audio amplifier circuit that is very sensitive to minor variations of the electrical characteristics of an individual vacuum tube. Such a circuit may well perform demonstrably differently with individual vacuum tubes.

The industry that designed and manufactured vacuum tube devices for a living worked diligently and with outstanding skill to minimize or eliminate the sensitivity of such circuits to minor tube-to-tube variation -- not to mention age-related drift in vacuum tube performance.

ASIDE: Consider the requirement of very broad-band RF tuners, oscillators, detectors, clippers, demodulators, sync, chroma, sweep, deflection, high voltage, and other circuitry for the amazing (early 1950s!) all-analog NTSC monochrome-compatible color over-the-air broadcast television system. Early color TVs had on the order of two dozen tubes in therm (some of which contained multiple active elements). The circuits (many of them tuned) in them had to be tolerant of significant variation and aging of the active componenets.

I reckon that it is perfectly possible to design a touchy enough audio circuit around an IC op amp (or a discrete one!) such that different op amps would perform markedly differently. I believe the goal of good engineering is to minimize or eliminate such idiocyncratic performance in an electronic appliance.

PS ... and never forget that vacuum tube operational amplifiers are always an option! OK, now I am being snarky again -- and I shall stop.

View attachment 464975

What about substitutions? I used to sub 6CA7s for EL34s to tighten up the bass. I never AB’d them, just replaced whole sets at a time when the old ones were worn.

 

[JeffS7444]

@olds1959special, no one is questioning whether you experienced something. The problem is that you give no indication that you made these comparisons with any sort of scientific rigor, nor do you "show your work", so it's not possible to review what you did. Slight differences in gain can influence what you perceive.

 

[Groove01]

Are we all just experiencing some sort of illusion?

Very likely. It’s a similar argument to cables and power cords.

A literal crap ton of people will swear cables make a difference. Likewise, millions of people think they understand economics. In both cases, they really don’t. The fact they find other people who agree with them doesn’t make them any less wrong, it just creates an insurmountable barrier to growth and understanding.

Even “if” the op amp makes a difference, your speakers, their placement in the room, and the room itself will impact the sound a 1000x more.

Unless your speakers, placement, and room are fully optimized- why waste time and effort for very, very low return.

For fun? OK, sure I can buy that - but just refer back to paragraph 2.

 

[Joachim Herbert]

Consider the requirement of very broad-band RF tuners, oscillators, detectors, clippers, demodulators, sync, chroma, sweep, deflection, high voltage, and other circuitry for the amazing (early 1950s!) all-analog NTSC monochrome-compatible color over-the-air broadcast television system. Early color TVs had on the order of two dozen tubes in therm (some of which contained multiple active elements). The circuits (many of them tuned) in them had to be tolerant of significant variation and aging of the active componenets.

Troubleshooting back in the day involved gently knocking on the tubes with a screwdrivers handle. Flickering images meant faulty tubes. About 80 percent of all cases were closed that way.

 

[emmodad]

What would you recommend for Tuvan Throat Singing?

poorly-laid-out, poorly-decoupled breadboarded amp design with dodgy supply rail stability
everybody remember asking their freshman op-amp circuit design lab grad student instructor why the speaker output sounded like a putt-putt motorboat?

 

[olds1959special]

A while ago I built a headphone amp and tested op-amps in it. I used these instructions:

How to Build the CMoy Pocket Amplifier

tangentsoft.com

I made a 'power amp' without a volume control, and powered it with a 24V 5A power supply.

I used solid silver wiring and silver bearing solder.

 

[kemmler3D]

@olds1959special, no one is questioning whether you experienced something. The problem is that you give no indication that you made these comparisons with any sort of scientific rigor, nor do you "show your work", so it's not possible to review what you did. Slight differences in gain can influence what you perceive.

Yep, there is no doubt something was heard. This happens roughly every time someone listens to two comparable devices in the same day.

The interesting question is why it was heard. If we have no solid info there, there's not much of a discussion to be had.

 

[solderdude]

A while ago I built a headphone amp and tested op-amps in it. I used these instructions:

How to Build the CMoy Pocket Amplifier

tangentsoft.com

I made a 'power amp' without a volume control, and powered it with a 24V 5A power supply.

I used solid silver wiring and silver bearing solder.

The silver wire and solder was a waste of money.

For C'Moy headphone amps the op-amps make a lot of difference as most op-amps are easily driven into clipping.
Most op-amps are not intended to provide the required currents in low impedance loads.
Here op-amps matter.

These circuits usually work pretty well with 300ohm and 600ohm headphones but the vast majority of op-amps don't want to see low impedance loads from headphones.

In the P4 and the power amps all of these op-amps do not have to supply any current and performance is the same.

 

[WillBrink]

I have been experimenting with op-amps in my Fosi P4 and can clearly hear that after all the ones I tried, the Sparkos SS3602 sounds excellent. I previously thought it sounded harsh and bright, with boomy bass, but even if this is true to some degree overall it is best choice of op-amp I've tried in this application. Why is it then, that folks here deny the benefits of op-amp rolling? My ears are not lying to me. The only challenge is to choose a single op-amp to stick with, but having tried five so far, I am tired of experimenting. So I plan to leave things alone now. That's my goal, to achieve the best result and leave it alone.

Using the stock op-amp did not provide the most musical enjoyment, and that's my point, why can't people here recognize this? Put your measurement devices away for a second and try listening to music.

We have heard that about speaker wire and power cables, and yes, their ears were lying to them, so...

 

[pma]

A while ago I built a headphone amp and tested op-amps in it. I used these instructions:

How to Build the CMoy Pocket Amplifier

tangentsoft.com

I made a 'power amp' without a volume control, and powered it with a 24V 5A power supply.

I used solid silver wiring and silver bearing solder.

Horrible build quality on a breadboard. Perfect way to get oscillations and induced interference.

 

[tengiz]

Well, we listen to different kind of music. You have to match the op-amp to music you are listening to. The op-amp that is good for say, classical music, is not going to sound good with rock as those audiences never mix. This is why everyone has their favorite op-amp. Different music, different audio "processor."

I second that.

But it goes deeper. Much deeper.

Let’s talk about the C code used in DSP filters. You think the loop structure doesn’t affect the sound? Think again.

• goto: crude, direct, unfiltered — perfect for punk rock.
• while: introspective, elegant, academic — ideal for classical and chamber music.
• for: balanced, controlled, versatile — the everyman of loops, suitable for anything from jazz to indie pop.

Don’t even get me started on do-while — you don't hear it, you feel it, like free jazz.

And now for the real rabbit hole:

Ask me about the use of the register modifier for function-local variables.

Yes, yes — I know the compilers say it’s obsolete.

But I beg to differ. I hear the difference. Especially in high-resolution FLACs and with well-matched op-amps. Try it with a Baroque fugue or some gritty doom metal. It’s there.

They’ve been hiding this from us for too long.
The real audiophile secrets are buried in compiler flags and microcode.

So tell me this:

• Do static inline functions sound warmer, or just smeared?
• Are recursive functions better for ambient, or are they strictly prog rock territory?
• And what’s the best compiler flag for jazz?

Let’s compare notes.
Hardware matters. Code structure matters. Music genre matters. The rest is lies and marketing.

There is more that the industry's been hiding from us.

Resist!

 

[EERecordist]

I second that.

But it goes deeper. Much deeper.

Let’s talk about the C code used in DSP filters. You think the loop structure doesn’t affect the sound? Think again.

• goto: crude, direct, unfiltered — perfect for punk rock.
• while: introspective, elegant, academic — ideal for classical and chamber music.
• for: balanced, controlled, versatile — the everyman of loops, suitable for anything from jazz to indie pop.

Don’t even get me started on do-while — you don't hear it, you feel it, like free jazz.

And now for the real rabbit hole:

Ask me about the use of the register modifier for function-local variables.

Yes, yes — I know the compilers say it’s obsolete.

But I beg to differ. I hear the difference. Especially in high-resolution FLACs and with well-matched op-amps. Try it with a Baroque fugue or some gritty doom metal. It’s there.

They’ve been hiding this from us for too long.
The real audiophile secrets are buried in compiler flags and microcode.

So tell me this:

• Do static inline functions sound warmer, or just smeared?
• Are recursive functions better for ambient, or are they strictly prog rock territory?
• And what’s the best compiler flag for jazz?

Let’s compare notes.
Hardware matters. Code structure matters. Music genre matters. The rest is lies and marketing.

There is more that the industry's been hiding from us.

Resist!

Not only the processor, but the compiler will affect the music! Real designers are all about bare metal code. Libraries are for the lazy.

 

[Salt]

A while ago I built a headphone amp and tested op-amps in it. I used these instructions:

How to Build the CMoy Pocket Amplifier

tangentsoft.com

I made a 'power amp' without a volume control, and powered it with a 24V 5A power supply.

I used solid silver wiring and silver bearing solder.

Where is the OPA? Maybe my presbyopia is tricking me ...

 

[Toni Mas]

View attachment 464859


JSmith

Well stop!