Does Op-amp Rolling Work?

[NTTY]

That's OK. As far as I know subjective impressions aren't banned on this site. I know some people would prefer that they were; if that's your position, take it up with the moderators. Meanwhile I will continue to occasionally post a personal impression and you may continue to totally ignore them.

It’s more than ok, this forum is very open.
I recently joined a French forum and got banned after a week.
Reason is that I posted an article from @DonH56 here about amplifiers’ damping factor.
And someone replied that ASR was non-sense, that @amirm was an incompetent "Emir", and that the appraisal of low value items emanated from scientistic people. So all of that was completely off topic. I thought all of it was extremely rude, to say the least. Couple of other members and I reacted and I understand we got all banned.

The link for those interested (in French), second post.

 

[KL....]

It’s more than ok, this forum is very open.
I recently joined a French forum and got banned after a week.
Reason is that I posted an article from @DonH56 here about amplifiers’ damping factor.
And someone replied that ASR was non-sense, that @amirm was an incompetent "Emir", and that the appraisal of low value items emanated from scientistic people. So all of that was completely off topic. I thought all of it was extremely rude, to say the least. Couple of other members and I reacted and I understand we got all banned.

The link for those interested (in French), second post.

Thank you, very harsh of that forum, especially indicateing that ASR was non-sense (it is not), @amirm was incompetent (clearly is not).... bit sad but forward we journey (always more to learn and enjoy), mustn't we :=)

 

[pma]

I am interested. Looking forward!

Thank you, I think I'll do it. Quite interestingly, in a non-inverting circuit with gain 6dB, NE5532 is better than LM4562 in THD vs. frequency test at 3V above 2kHz!!

 

[Old_School_Brad]

Thank you, I think I'll do it. Quite interestingly, in a non-inverting circuit with gain 6dB, NE5532 is better than LM4562 in THD vs. frequency test at 3V above 2kHz!!

And Pavel, if you're up for it, maybe give a brief introduction to what you're doing and why, along with a bit of an explanation. I know some things, but I'm always eager to learn more, and a little context can be really helpful. That said, I totally understand if you'd rather not.

 

[audio_tony]

I have a test rig to test dual operational amplifiers, at low gain and high gain.

View attachment 438438

If there are readers interested here, I can make quite detailed comparison of LME49720 (LM4562), NE5532, OPA2134, TL072 and LM1458 in a new separate thread. THDlevel, THDfrequency, CCIF IMD, DIM24 and DIM48, square responses. Just let me know.

View attachment 438439

I proposed doing this in the other thread, but my suggestion was met with disdain.

 

[solderdude]

Pavel, before you take on a workload basically verifying data sheets of a few well known op-amps and doing all kinds of tests including 1458 (as well as TL072) at various voltage supply rails and loads I think that some measurement buffs would like to see the as good as negligible differences in certain applications/loads this will only give some more plots and not debunk power supply decoupling issues, PSRR nor RF immunity etc.

I suspect people prefer to 'see' how a few of those expensive discrete thingies and modern op-amps perform opposite 5532.

This also will not satisfy the 'but this one sounds better' crowd.
They might prefer 'audio recordings' reproduced on an excellent DAC, well made recordings and the outputs recorded with a TOTL ADC in hires, exactly level matched (simply swapping op-amps should do that automatically), the source file (upsampled to the same format as the recorded material) so these can be played back at home on 'their perfect sounding gear' so they can check how the op-amps sound.

That's taking on a lot of work which won't win over any of the 'hearing' crowd and would only confirm what most people here already know.
Namely, all op-amps will measure differently, some will be SOTA and others will be 'more than good enough' in 'buffer to few times gain circuits and no one will be able to tell recordings apart with statistical relevance when tested blind.

You're welcome to try but fear it will only take up a lot of time and bring nothing more than what you already posted.

 

[pma]

Pavel, before you take on a workload basically verifying data sheets of a few well known op-amps and doing all kinds of tests including 1458 (as well as TL072) at various voltage supply rails and loads I think that some measurement buffs would like to see the as good as negligible differences in certain applications/loads this will only give some more plots and not debunk power supply decoupling issues, PSRR nor RF immunity etc.

Yes of course, because it is not generally known how much e.g. the common mode distortion (i.e. noninverting gain and unequal impedances in +IN and -IN opamp inputs) plays a decisive role and it is not evident from standard datasheets at first sight. It is covered in Samuel Groner's (is he a member here?) tests but magnified by his special test setup. You might be interested how much the quite usual gain +2 circuit with unequal input impedances sorts the opamps and that the 5532 is better in this respect than 49720/4562. And how very poor is 2134 - but this is quite well known about the JFET input opamps an this phenomenon was reduced in the expensive (but IMO still best JFET opamps) 627/637.
So, the intention is definitely not to replicate the datasheet plots.

If I were to make a conclusion, I would never say that opamps do not make a difference in audio circuits. Because it strongly depends on the circuit schematics. We may say that in case of the unity gain buffer the difference would be very small. But that's all.

 

[DonH56]

It’s more than ok, this forum is very open.
I recently joined a French forum and got banned after a week.
Reason is that I posted an article from @DonH56 here about amplifiers’ damping factor.
And someone replied that ASR was non-sense, that @amirm was an incompetent "Emir", and that the appraisal of low value items emanated from scientistic people. So all of that was completely off topic. I thought all of it was extremely rude, to say the least. Couple of other members and I reacted and I understand we got all banned.

The link for those interested (in French), second post.

It's a shame how many die-hard subjectivists dismiss without thought the chance to learn reasons why amplifiers sound different (and other things). Seems to be the way of the world these days; if I do not like your position, you must be personally destroyed for taking a position contrary to my own, irrespective of truth or fiction.

On subjective views, my goal is usually to figure out why the difference, though often enough it is just bias or poor test controls.

 

[gamerpaddy]

i wanted to try it myself.

my test: a simple gain=1 buffer like in 99% of dac and amps supplied by a normal 12v wall wart psu with some caps.

i recorded 4 tracks on my audio interface to have the same clock on adc and dac.
test music was a 27mb FLAC played trough foobar2000.

ive put a NE5532, OPA1656 and LM4562 in there and a "short-bridge"

84.68 MB file on MEGA

mega.nz

the zip contains 4 audio files labeled A B D C
results are here:

Spoiler: results

Spoiler: no cheating

Spoiler: ok then, theres no way back

A LM4562
B NE5532
C OPA1656
D NONE

 

[pma]

New thread with op-amp measurements:

Big test of DIP dual op-amps for audio: LM4562, NE5532, OPA2134, TL072 and LM1458

There are questions if it matters which type of op-amp is used in the audio amplifier and my answer is that it depends on circuit schematics and op-amp parameters. Some of those parameters are obvious from datasheets, but others may be quite hidden and not clear from the first view. For the test...

www.audiosciencereview.com

 

[pma]

my test: a simple gain=1 buffer like in 99% of dac and amps supplied by a normal 12v wall wart psu with some caps.

Do you have any supply bypass capacitors installed on the breadboard? Can you show the bottom view of the breadboard?

 

[gamerpaddy]

just two 10mF caps between V+ V- to gnd
and two resistors for balancing so i can use a virtual ground and not needing a dual rail supply

 

[HaveMeterWillTravel]

Can you define what is 1. "meh" chip, 2. good chip, 3. better chip, 4. best chip, and 5—you can't beat this cheap I meant chip?

If not how do the manufacturers know how many buyers would buy what?

This is unlike Dell PC where a Celeron performs different from a Pentium and costs different from a Pentium. Meanwhile the other parts are often swapped without informing the customer since they do not affect the performance. There are lots of comparators, DACs, sensors on motherboards and chips, btw. We just don't hear about them because the only thing snake oil buyers know are that one op-amp in that particular location. Let's totally forget that the amplifier module has its own differential-amplifier input stage than you cannot replace.

Furthermore to add, we are talking about a component which does not bottleneck the performance. This is like giving a 400W computer a choice of 600W and 800W power supplies, you don't see Dell doing that either.

There is a reason why everyone just didn't keep using the GAP/R K2-W op-amp, then the TL072 or μA709.

There are many many modern op-amps which not suitable for this function. SOTA op-amps intended for general audio use crossed over the "audibly indistinguishable" quite some time ago. Only in certain specific applications are there differences in suitability, such as driving very low or higher impedance headphones.

 

[HaveMeterWillTravel]

I would spend hours going through it. They seemed like gift from heaven! Free schematics for all sorts of things.

There is so many great reference material from the various chip manufactures as well as measurement equipment companies ( HP, Keithley, etc).
I would postulate any reasonably smart EE and only some years of experience could build a nicely designed circuit for the majority of applications out there.

i wanted to try it myself.


View attachment 438482

SMD to PDIP adapters were a godsend.

However I have been using the new TI AMP-PDK-EVM and it's daughter cards. One can use the cards even for non-amp products.

 

[solderdude]

just two 10mF caps between V+ V- to gnd
and two resistors for balancing so i can use a virtual ground and not needing a dual rail supply

View attachment 438497

Best practice (certainly for fast op-amps) would be to also use 10nF-100nF ceramic caps directly from powerpins to common/ground as well as the used electrolytics.
Just like the board PMA uses.

 

[Earwax]

This might prove to be helpful as there is a sound comparison for the OP-amps He picked.

 

[Bruce Morgen]

Ron is Danny Richie's former video producer and the audiophool sidekick who got him into the utter foolishness of op-amp rolling -- IMO that alone should make one take anything he presents with a boulder size chunk of good old NaCl. The fact that he's also buddied up with Randy, the "cheapaudioman" and a literal comedian, simply confirms his subterranean level of credibility.

 

[DaneelOlivaw]

i wanted to try it myself.

my test: a simple gain=1 buffer like in 99% of dac and amps supplied by a normal 12v wall wart psu with some caps.

i recorded 4 tracks on my audio interface to have the same clock on adc and dac.
test music was a 27mb FLAC played trough foobar2000.

ive put a NE5532, OPA1656 and LM4562 in there and a "short-bridge"

84.68 MB file on MEGA

mega.nz

the zip contains 4 audio files labeled A B D C
results are here:

Spoiler: results

Spoiler: no cheating

Spoiler: ok then, theres no way back

A LM4562
B NE5532
C OPA1656
D NONE

View attachment 438482

I tried to hear a difference between A B C D and failed.
I used my Sabaj A20D with which i was able to hear differences of the Samples on https://www.oppodigital.com/hra/dsd-by-davidelias.aspx (vision of her).
From there here i tested:

HighRes
dsd64 2.822Mhz, PCM 88.2khz, PCM flac 88.2khz
no difference between them but superior to:

Standard:
MQA ('352.8khz'), PCM 44.1khz
no difference between them but superior to:

MP3 (320k-44.1khz)

So my ears cant be too bad.

 

[amirm]

Ron is Danny Richie's former video producer

He is producing all of his videos. He was hosting him on his channel for a while but then spun off. But Ron remained as the video producer.

 

[Bruce Morgen]

He is producing all of his videos. He was hosting him on his channel for a while but then spun off. But Ron remained as the video producer.

Correction duly noted.