Op Amps - AD797AN

[mcp6453[_2_]]

The first console I ever build used NE5532 op amps. The last time I designed a
circuit, I used AD797ANs. In fact, there's a tube of the around here somewhere,
but they are not to be found? The going price seems to be about $10 each from
Digikey. What's the modern day preferred low noise op amp usable as a line
driver, able to drive 600 ohms?

[Peter Larsen[_3_]]

mcp6453 wrote:

> The first console I ever build used NE5532 op amps. The last time I
> designed a circuit, I used AD797ANs. In fact, there's a tube of the
> around here somewhere, but they are not to be found? The going price
> seems to be about $10 each from Digikey. What's the modern day
> preferred low noise op amp usable as a line driver, able to drive 600
> ohms?

Opinion: Analog Devices OP176 for a single, if you want a dual then use Burr
Brown OPA2134.

Note: I think I read something to the effect of some 553x's being more equal
than other 553x's, ie. that they do not all sound similar, some are more
pleasing than others. However in my _opinion_ the OPAx134 has the same
forward sounding charm as the 553x AND a cleaner treble and tighter bass.
The OP176/275's are however much more wellbehaved during turn on and turn
off, in case it matters. An OPA176 is not just half a 275, they updated it
when they made the single, the original design was a dual, the single was on
customer demand and it is in my OPINION the perfect compromise between the
slightly glassy Burr Brown x604 and the gentle and sweet but not overbright
OP 275. I did try replacing OPA176's in a IFM cross-over with 134's and got
a more forward sound, but also turn-on/turn off thumps.

Kind regards

Peter Larsen

[Kevin Aylward[_4_]]

"Peter Larsen" wrote in message
k...

mcp6453 wrote:

>> The first console I ever build used NE5532 op amps. The last time I
>> designed a circuit, I used AD797ANs. In fact, there's a tube of the
>> around here somewhere, but they are not to be found? The going price
>> seems to be about $10 each from Digikey. What's the modern day
>> preferred low noise op amp usable as a line driver, able to drive 600
>> ohms?


>OPAx134 has the same forward sounding charm as the 553x AND a cleaner
>treble and tighter bass. my OPINION the perfect compromise between the
>slightly glassy Burr Brown x604 and the gentle and sweet but not overbright
>OP 275. I did try replacing OPA176's in a IFM cross-over with 134's and got
>a more forward sound

This comes across as a review in an audiophool mag, rather than a somewhat
objective opinion from a somewhat claimed technical NG. However, to your
credit, you do state that it is only your opinion.

Having looked at the (distortion) specs for the OPA602 and the OP275, it is
my technical evaluation that claims that they sound audibly different
(except possible in overload, noise) are simply delusional. This is not an
opinion, but a fact, as in, evolution is a fact.

I also looked at the data sheet for the OPA134, which has, for example,
graphs of IMD of the OPA134 and OP176 overlaid, and again, it is my
technical evaluation that claims that they sound audibly different (except
possible in overload, noise) are simply delusional.

All of the above opamps, have truly insignificant (~0.001%) LF distortion,
and low enough HF distortion that alleged audible differences between them
are simply not credible, when used competently, i.e. closed loop gains not
set to 1000. Normal closed frequency response of all, in normal designs will
be 50khz+, hence not audibly different.

I would like to say I am amazed that this sort of audiophool nonsense is is
still being abounded about, but unfortunately, I am not. Only dismayed.


Kevin Aylward B.Sc.
http://www.kevinaylward.co.uk/ee/index.html

[Mike Rivers]

On 2/19/2011 5:41 AM, Kevin Aylward wrote:

> Having looked at the (distortion) specs for the OPA602 and
> the OP275, it is my technical evaluation that claims that
> they sound audibly different (except possible in overload,
> noise) are simply delusional. This is not an opinion, but a
> fact, as in, evolution is a fact.
>
> I also looked at the data sheet for the OPA134, which has,
> for example, graphs of IMD of the OPA134 and OP176 overlaid,
> and again, it is my technical evaluation that claims that
> they sound audibly different (except possible in overload,
> noise) are simply delusional.

That's based on the evaluation circuit that the manufacturer
used. If Peter did his listening tests by swapping parts in
an existing console or preamp, his "demo" conditions could
have been different than the manufacturer's "test" conditions.

> All of the above opamps, have truly insignificant (~0.001%)
> LF distortion, and low enough HF distortion that alleged
> audible differences between them are simply not credible,
> when used competently

Many things with similar "insignificant" amounts of measured
distortion actually sound different. It's one of the reasons
why there are so many boxes that amplify that compete for
our money. If there were truly no audible differences, why
wouldn't every manufacturer use the same, least expensive
component?

> I would like to say I am amazed that this sort of audiophool
> nonsense is is still being abounded about, but
> unfortunately, I am not. Only dismayed.

And I'm dismayed that a design engineer tosses off the
experience of someone with enough experience to detect
differences in sound under actual working conditions.

Me, I have no opinion. I buy what I need, that I can afford,
and don't worry if there's something that someone has
declared "better" or even "equal."

--
"Today's production equipment is IT based and cannot be
operated without a passing knowledge of computing, although
it seems that it can be operated without a passing knowledge
of audio." - John Watkinson

http://mikeriversaudio.wordpress.com - useful and
interesting audio stuff

[mcp6453[_2_]]

On 2/19/2011 5:41 AM, Kevin Aylward wrote:
>
> This comes across as a review in an audiophool mag, rather than a somewhat
> objective opinion from a somewhat claimed technical NG. However, to your credit,
> you do state that it is only your opinion.
>
> Having looked at the (distortion) specs for the OPA602 and the OP275, it is my
> technical evaluation that claims that they sound audibly different (except
> possible in overload, noise) are simply delusional. This is not an opinion, but
> a fact, as in, evolution is a fact.
>
> I also looked at the data sheet for the OPA134, which has, for example, graphs
> of IMD of the OPA134 and OP176 overlaid, and again, it is my technical
> evaluation that claims that they sound audibly different (except possible in
> overload, noise) are simply delusional.
>
> All of the above opamps, have truly insignificant (~0.001%) LF distortion, and
> low enough HF distortion that alleged audible differences between them are
> simply not credible, when used competently, i.e. closed loop gains not set to
> 1000. Normal closed frequency response of all, in normal designs will be 50khz+,
> hence not audibly different.
>
> I would like to say I am amazed that this sort of audiophool nonsense is is
> still being abounded about, but unfortunately, I am not. Only dismayed.

Kevin, if you were going to buy a lot of singles or duals, like I did with
5534s, 5532s, and AD797s for general purpose audio (summing amps, buffers,
moderate line level gain stages, but not including microphone preamps), what
would you pick? My inventory of go-to op amps is around here somewhere never to
be found. I now have a source for making low run PCBs at a reasonable cost, so
I'm considered building up a few utility circuits I want, such as a hard-wired
mix-minus bus. The fidelity of that bus is totally non-critical, but there's no
reason to use a crappy op amp.

[Mark]

to the op..

I'm not going to get into the techie vs audiophool thing, when you
lie down with dogs, you get up with fleas as we are seeing in some
other threads here..

but I do want to mention a relatively new architecture of op-amp, the
FULLY differential amp (FDA)

http://en.wikipedia.org/wiki/Fully_differential_amplifier

it may have some advantages for you as a differential line driver..

If you decide to design with it, beware some of the things about the
way it works with two feedback paths are counter intuitive. For
example, mis-match resistors in the two feedback paths do NOT upset
the output balance. TI and AD have some good ap notes.


Mark

[Peter Larsen[_3_]]

Kevin Aylward wrote:

> This comes across as a review in an audiophool mag, rather than a
> somewhat objective opinion from a somewhat claimed technical NG.
> However, to your credit, you do state that it is only your opinion.

Good boy Kevin, you actually got the point, I stated my opinion.

However as for your general flak: If you can not hear the
difference between different opamps in a real world application, then you
need to do three things: get some proper test material, get some better
loudspeakers and - most important: learn to really listen.

I have done opamp listening tests by replacing opamps in electronic
cross-overs, as headphone-drivers, in a single opamp stage RIAA and via a CD
with samples of music recorded via a RIAA stage under development by someone
else, the latter test was blind and involved a large handful of testers.

After the blind evaluations came in - panel got a cd in the mail and
returned it with a commented rating of sound examples - the gudanuff opamps
suggested by your local companion in the credo "0.001 percent sounds the
same from all vendors" were replaced with what the testers evaluation
suggested. Testers listened blind and separately and all pointed in the same
general direction and a sound business decision was based on the opinion of
the test panel because of the large agreement in it.

You may also want to read up on the designers intentions with the OP275/176,
applications notes should be readily available. Since you say that it does
not have a sonic signature that is different from the gudanuff brand, then
you contradict those designers stated aim of making it different via
distortion spectra shaping.

> Kevin Aylward B.Sc.

Kind regards

Peter Larsen

[Scott Dorsey]

Peter Larsen > wrote:
>I have done opamp listening tests by replacing opamps in electronic
>cross-overs, as headphone-drivers, in a single opamp stage RIAA and via a CD
>with samples of music recorded via a RIAA stage under development by someone
>else, the latter test was blind and involved a large handful of testers.
>
>After the blind evaluations came in - panel got a cd in the mail and
>returned it with a commented rating of sound examples - the gudanuff opamps
>suggested by your local companion in the credo "0.001 percent sounds the
>same from all vendors" were replaced with what the testers evaluation
>suggested. Testers listened blind and separately and all pointed in the same
>general direction and a sound business decision was based on the opinion of
>the test panel because of the large agreement in it.

And part of the problem here is that test circuits aren't the same as real
world circuits. The input and output impedances have an awful lot to do
with the distortion characteristic of an op-amp in the real world, and
consequently the op-amp that sounds best in one circuit may not operate
reliably at all in another.

Also there are some interesting stability issues on some of the commercial
monolithics... look at the Bode plot on the very-popular OPA2604.

On top of all this, in listening tests you can find that fairly subtle
changes to the noise floor get perceived as tonal changes. In a high
gain circuit, the noise floor on an OPA2604 and an NE5532 will be very
different.

Some of this matters, some of it doesn't. Some of it only matters when you
have a huge number of identical stages in the same signal path. Only listening
will tell you for sure if there's a problem, only careful measurement will
tell you what that problem is.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[mcp6453[_2_]]

On 2/19/2011 9:39 AM, Scott Dorsey wrote:

> And part of the problem here is that test circuits aren't the same as real
> world circuits. The input and output impedances have an awful lot to do
> with the distortion characteristic of an op-amp in the real world, and
> consequently the op-amp that sounds best in one circuit may not operate
> reliably at all in another.
>
> Also there are some interesting stability issues on some of the commercial
> monolithics... look at the Bode plot on the very-popular OPA2604.
>
> On top of all this, in listening tests you can find that fairly subtle
> changes to the noise floor get perceived as tonal changes. In a high
> gain circuit, the noise floor on an OPA2604 and an NE5532 will be very
> different.
>
> Some of this matters, some of it doesn't. Some of it only matters when you
> have a huge number of identical stages in the same signal path. Only listening
> will tell you for sure if there's a problem, only careful measurement will
> tell you what that problem is.

It looks like I may just use 5532s for this project. They're still good op amps
although there are better one. They sure are cheap. I've spent most of the
morning reading op amp opinions. One thing is absolutely true: There is no
consensus.

I'm still surprised that there is not much talk of the AD797. Maybe it didn't
prove as useful for audio as it was thought to be. Maybe it's too expensive.

[Arny Krueger]

"Kevin Aylward" > wrote in
message
> "Peter Larsen" wrote in message
> k...
>
> mcp6453 wrote:
>
>>> The first console I ever build used NE5532 op amps. The
>>> last time I designed a circuit, I used AD797ANs. In
>>> fact, there's a tube of the around here somewhere, but
>>> they are not to be found? The going price seems to be
>>> about $10 each from Digikey. What's the modern day
>>> preferred low noise op amp usable as a line driver,
>>> able to drive 600 ohms?
>
>
>> OPAx134 has the same forward sounding charm as the 553x
>> AND a cleaner treble and tighter bass. my OPINION the
>> perfect compromise between the slightly glassy Burr
>> Brown x604 and the gentle and sweet but not overbright
>> OP 275. I did try replacing OPA176's in a IFM cross-over
>> with 134's and got a more forward sound
>
> This comes across as a review in an audiophool mag,
> rather than a somewhat objective opinion from a somewhat
> claimed technical NG. However, to your credit, you do
> state that it is only your opinion.
> Having looked at the (distortion) specs for the OPA602
> and the OP275, it is my technical evaluation that claims
> that they sound audibly different (except possible in
> overload, noise) are simply delusional. This is not an
> opinion, but a fact, as in, evolution is a fact.
> I also looked at the data sheet for the OPA134, which
> has, for example, graphs of IMD of the OPA134 and OP176
> overlaid, and again, it is my technical evaluation that
> claims that they sound audibly different (except possible
> in overload, noise) are simply delusional.

I would say illusional. There was a time when op amps were far more
marginal. There's a big difference in what happens if you try to squeeze
+22 dB @ 20 KHz into 600 ohms out of a 5532 versus a LM 741. The last
generation of digitral converters required a new generation of high
performance op amps so that they could be tested properly, and the 5532's
use in high performance instrumentation is now pretty well eclipsed. But we
are talking about an insane number of leading zeroes.

> All of the above opamps, have truly insignificant
> (~0.001%) LF distortion, and low enough HF distortion
> that alleged audible differences between them are simply
> not credible, when used competently, i.e. closed loop
> gains not set to 1000. Normal closed frequency response
> of all, in normal designs will be 50khz+, hence not
> audibly different.

Agreed.

> I would like to say I am amazed that this sort of
> audiophool nonsense is is still being abounded about,
> but unfortunately, I am not. Only dismayed.

The practical view of these sorts of things is that the colorations of
properly-implimented modern op amps is often grossly overstated by both
professionals and amateurs.

There is an obvious caveat, which is that while modern op amps can be very
tolerant, it is also easy enough to misapply them.

One common misapplication is using chips with way too much GBW, yielding
ringing and high frequency response artifacts. Vast differences in offset
currents in high impedance circuits can hurt dynamic range.

Modern chips often swing very close to the rails, even with very low
impedance loads. 5532s are good into low impedance loads, but not *that*
good. A TL072 can't touch a 5532 running professional levels into 600 ohm
loads, but can do fine when buffered adequately, if you protect them from
input lockup.

In one demonstration about 30 TL074 op amp sections were cascaded, set for
unity gain using feedback resistors on the order of 10k. There were +/- 15
volt supplies and signal levels around 1 volt. No golden ears have ever
been able to hear the difference in a blind test.

You can say that you hear whatever you want if you do sighted evaluations.
:-(

In another demonstration the much-maligned LM 301 was compared to a straight
wire under similar operational conditions. Same results.

OTOH as you suggested above, there are people who will try to build a mic
preamp with 60 dB gain out of op amps with 1 MHz GBW. As they say, do the
math! :-(

[Arny Krueger]

"mcp6453" > wrote in message


> The first console I ever build used NE5532 op amps. The
> last time I designed a circuit, I used AD797ANs. In fact,
> there's a tube of the around here somewhere, but they are
> not to be found? The going price seems to be about $10
> each from Digikey. What's the modern day preferred low
> noise op amp usable as a line driver, able to drive 600
> ohms?

Here is a heck of a freebie info if you want some *light reading* about op
amps:

http://focus.ti.com/lit/an/slod006b/slod006b.pdf

The current hot chips are the TI

OPA 1612
OPA 1632
OPA 2228
OPA 4228

http://focus.ti.com/lit/sl/slyy013f/slyy013f.pdf

and the National

LM 4562

http://www.national.com/pf/LM/LM4562.html

[Kevin Aylward[_4_]]

"Mike Rivers" wrote in message
...

On 2/19/2011 5:41 AM, Kevin Aylward wrote:

>> Having looked at the (distortion) specs for the OPA602 and
>> the OP275, it is my technical evaluation that claims that
>> they sound audibly different (except possible in overload,
>> noise) are simply delusional. This is not an opinion, but a
>> fact, as in, evolution is a fact.
>>
>> I also looked at the data sheet for the OPA134, which has,
>> for example, graphs of IMD of the OPA134 and OP176 overlaid,
>> and again, it is my technical evaluation that claims that
>> they sound audibly different (except possible in overload,
>> noise) are simply delusional.

>That's based on the evaluation circuit that the manufacturer used. If Peter
>did his listening tests by swapping parts in an existing console or preamp,
>his "demo" conditions could have been different than the manufacturer's
>"test" conditions.

Immaterial. Numerous tests with alleged "golden ears" over many years have
all concluded that specifications such as given by those sort of amp are
simply not audibly detectable. I am not making my claims based on any
specific implementation. My claim is based on knowing what imperfections can
be shown to be not audibly detectable in actual experiments, and what
imperfections the op-amps have.

>> All of the above opamps, have truly insignificant (~0.001%)
>> LF distortion, and low enough HF distortion that alleged
>> audible differences between them are simply not credible,
>> when used competently

>Many things with similar "insignificant" amounts of measured distortion
>actually sound different.

Oh? Care to cite some experiments in support of that position?

Note, I do not include one off spot measurements such that in actual realty,
there is significant distortion which is being ignored by subterfuge. For
example, a spot full power measurement might show say, 0.01% THD, but a 40db
down level might show a detectable 1% cross-over distortion, but this figure
is simple not stated in the test. Indeed, I am certainly aware of specially
constructed distortions that are abundantly detectable at 0.1% THD when
referenced to a full scale signal, such that when the audio signal is set to
be just below the clip point, it still sounds truly dreadful. However, if
one was actually attempting to present the truth, one would note that the
"real distortion" is much, much higher for most of the time.


>It's one of the reasons why there are so many boxes that amplify that
>compete for our money.

I disagree. There are a lot of boxes out there because everyone wants a
piece of the same pie. Companies will invent any reason they can to get
sales.

>If there were truly no audible differences, why wouldn't every manufacturer
>use the same, least expensive component?

I am *not* claiming that *all* op-amps will sound the same, only those that
meet certain, somewhat indefinite criteria will. A 741 set to a closed loop
gain of 100 will have significantly measurable IMD distortion at high audio
frequencies, and probably be quite audibly detectable.

So, there may well be audible differences between a 741 and OPA602. My claim
is for the range of modern, high performance op-amps that are, essentially,
audibly "perfect".



>> I would like to say I am amazed that this sort of audiophool
>> nonsense is is still being abounded about, but
>> unfortunately, I am not. Only dismayed.

>And I'm dismayed that a design engineer tosses off the experience of
>someone with enough experience to detect differences in sound under actual
>working conditions.

I toss off the experience in the same way as Richard Dawkins tosses off the
experience of those priests that claim that they had a personal experience
with a god. Richard has no hesitation is stating that such people are
deluded.

As far as experience goes, I myself have been a guitar playing, band
performing, electronics bod since the age of 11. I am now 52. So, on that
level, I am at least equally qualified. However I am not backing my claims
based on my own personal experience, which I agree could also be delusional
in principle. My claims are based on being acquainted with sufficient peer
review published literature on the subject. Double-blind controlled tests
have irrefutable determined what sort of levels of frequency response and
distortion errors are not audibly detectable. These have been performed and
published over many years, and all of this is just rehashing what has
already been convincible demonstrated, if only people would take the trouble
and look at this existing evidence. Indeed, Randi James is offering
(http://www.randi.org/site/index.php/jref-news/116-10-tips-for-dealing-with-randi.html)
$1M if someone can demonstrate the value of hi-end speaker cables.

I have looked at the specs for the particular op-amps mentioned, noting
distortions over frequencies and output levels, and different gain settings,
and have summarised this by making the equivalent claim that these
particular op-amps, are essentially, a straight piece of wire with gain for
audio.

Douglas Self (of Soundcraft) summarises much of this here:

http://www.douglas-self.com/ampins/pseudo/subjectv.htm


Kevin Aylward B.Sc.

www.kevinaylward.co.uk

[Kevin Aylward[_4_]]

"mcp6453" wrote in message
...

On 2/19/2011 5:41 AM, Kevin Aylward wrote:
>>
>> This comes across as a review in an audiophool mag, rather than a
>> somewhat
>> objective opinion from a somewhat claimed technical NG. However, to your
>> credit,
>> you do state that it is only your opinion.
>>
>> Having looked at the (distortion) specs for the OPA602 and the OP275, it
>> is my
>> technical evaluation that claims that they sound audibly different
>> (except
>> possible in overload, noise) are simply delusional. This is not an
>> opinion, but
>> a fact, as in, evolution is a fact.
>>
>> I also looked at the data sheet for the OPA134, which has, for example,
>> graphs
>> of IMD of the OPA134 and OP176 overlaid, and again, it is my technical
>> evaluation that claims that they sound audibly different (except possible
>> in
>> overload, noise) are simply delusional.
>>
>> All of the above opamps, have truly insignificant (~0.001%) LF
>> distortion, and
>> low enough HF distortion that alleged audible differences between them
>> are
>> simply not credible, when used competently, i.e. closed loop gains not
>> set to
>> 1000. Normal closed frequency response of all, in normal designs will be
>> 50khz+,
>> hence not audibly different.
>>
>> I would like to say I am amazed that this sort of audiophool nonsense is
>> is
>> still being abounded about, but unfortunately, I am not. Only dismayed.

>Kevin, if you were going to buy a lot of singles or duals, like I did with
>5534s, 5532s, and AD797s for general purpose audio (summing amps, buffers,
>moderate line level gain stages, but not including microphone preamps),
>what
>would you pick? My inventory of go-to op amps is around here somewhere
>never to
>be found. I now have a source for making low run PCBs at a reasonable cost,
>so
>I'm considered building up a few utility circuits I want, such as a
>hard-wired
>mix-minus bus. The fidelity of that bus is totally non-critical, but
>there's no
>reason to use a crappy op amp.


My view is that if you stay away from the likes of 741, 4558, and use a
relatively more modern op-amp which quote these silly low spot distortion
figures, you cant go to much wrong. The fet input ones will have lower noise
when driven by larger source resistances (>10kohms), but generally for low
source resistances (< 1kohms) the bipolar ones are quieter. Its the base
current noise times the source impedance bit. Fets have no base current!.
For guitar inputs, you should certainly use a fet input op-amp. The 5534s,
5532s are quite old now, so I personally would give those amiss nowadays,
but they are cheaper, probably. For a non-discrete mic input i.e. some sort
of instrumentation configuration its probably the likes of the LT1028.

Kevin Aylward B.Sc.
www.kevinaylward.co.uk

[Kevin Aylward[_4_]]

"Arny Krueger" wrote in message
...

"Kevin Aylward" >> wrote in
message
>> "Peter Larsen" wrote in message
>> k...
>>
>> mcp6453 wrote:
>>
>>>>>> The first console I ever build used NE5532 op amps. The
>>>>>> last time I designed a circuit, I used AD797ANs. In
>>>>>> fact, there's a tube of the around here somewhere, but
>>>>>> they are not to be found? The going price seems to be
>>>>>> about $10 each from Digikey. What's the modern day
>>>>>> preferred low noise op amp usable as a line driver,
>>>>>> able to drive 600 ohms?
>>
>>
>>>> OPAx134 has the same forward sounding charm as the 553x
>>>> AND a cleaner treble and tighter bass. my OPINION the
>>>> perfect compromise between the slightly glassy Burr
>>>> Brown x604 and the gentle and sweet but not overbright
>>>> OP 275. I did try replacing OPA176's in a IFM cross-over
>>>> with 134's and got a more forward sound
>>
>> This comes across as a review in an audiophool mag,
>> rather than a somewhat objective opinion from a somewhat
>> claimed technical NG. However, to your credit, you do
>> state that it is only your opinion.
>> Having looked at the (distortion) specs for the OPA602
>> and the OP275, it is my technical evaluation that claims
>> that they sound audibly different (except possible in
>> overload, noise) are simply delusional. This is not an
>> opinion, but a fact, as in, evolution is a fact.
>> I also looked at the data sheet for the OPA134, which
>> has, for example, graphs of IMD of the OPA134 and OP176
>> overlaid, and again, it is my technical evaluation that
>> claims that they sound audibly different (except possible
>> in overload, noise) are simply delusional.

>I would say illusional. There was a time when op amps were far more
>marginal.

Yes.

> There's a big difference in what happens if you try to squeeze +22 dB @
> 20 KHz into 600 ohms out of a 5532 versus a LM 741. The last generation
> of digitral converters required a new generation of high performance op
> amps so that they could be tested properly, and the 5532's use in high
> performance instrumentation is now pretty well eclipsed. But we are
> talking about an insane number of leading zeroes.

Yes. One cant ignore the 1% speaker distortion, so caring about 1.0001% does
sound a bit farfetched.

>> All of the above opamps, have truly insignificant
>> (~0.001%) LF distortion, and low enough HF distortion
>> that alleged audible differences between them are simply
>> not credible, when used competently, i.e. closed loop
>> gains not set to 1000. Normal closed frequency response
>> of all, in normal designs will be 50khz+, hence not
>> audibly different.

>Agreed.

>> I would like to say I am amazed that this sort of
>> audiophool nonsense is is still being abounded about,
>> but unfortunately, I am not. Only dismayed.

>The practical view of these sorts of things is that the colorations of
>properly-implimented modern op amps is often grossly overstated by both
>professionals and amateurs.

Yes.

>There is an obvious caveat, which is that while modern op amps can be very
>tolerant, it is also easy enough to misapply them.

>One common misapplication is using chips with way too much GBW, yielding
>ringing and high frequency response artifacts. Vast differences in offset
>currents in high impedance circuits can hurt dynamic range.

Or simply not accounting for cable capacitance caused instability.

>Modern chips often swing very close to the rails, even with very low
>impedance loads. 5532s are good into low impedance loads, but not *that*
>good. A TL072 can't touch a 5532 running professional levels into 600 ohm
>loads, but can do fine when buffered adequately, if you protect them from
>input lockup.

>In one demonstration about 30 TL074 op amp sections were cascaded, set for
>unity gain using feedback resistors on the order of 10k. There were +/- 15
>volt supplies and signal levels around 1 volt. No golden ears have ever
>been able to hear the difference in a blind test.

Yes. Its about 0.005% at 20khz in unity gain. At much higher gains I dare
say, 30 might be detectable.

>You can say that you hear whatever you want if you do sighted evaluations.
>:-(

>In another demonstration the much-maligned LM 301 was compared to a
>straight wire under similar operational conditions. Same results.

>OTOH as you suggested above, there are people who will try to build a mic
>preamp with 60 dB gain out of op amps with 1 MHz GBW. As they say, do the
>math! :-(

Indeed. You need loop gain at the distortion frequency to reduce distortion
at the fundamental, i.e. 40khz and 60khz. Of course, high frequency
modulation is usually only a problem when you have say, synth generated
edges mixing with other frequencies. Most audio just doesn't get full rail
at 20Khz.


Kevin Aylward B.Sc.

www.kevinaylward.co.uk

[Scott Dorsey]

mcp6453 > wrote:
>
>It looks like I may just use 5532s for this project. They're still good op amps
>although there are better one. They sure are cheap. I've spent most of the
>morning reading op amp opinions. One thing is absolutely true: There is no
>consensus.

That's a reasonable suggestion. If you don't have super low or super high
impedances, and you don't need a super low noise floor, the 5532 is pretty
damn good.

It will do bad things at really high gains, so don't do that. It's not
happy as a phono front end, so don't do that either.

>I'm still surprised that there is not much talk of the AD797. Maybe it didn't
>prove as useful for audio as it was thought to be. Maybe it's too expensive.

The distortion signature is different than most typical op-amps, and that
might be a good thing. I liked the MC34082 for the same reason, I think,
just that it was different and so if you mixed it in the cascaded distortion
products wouldn't all be the same since the op-amps wouldn't all be the same
in the signal path.

With the AD797 you pay a little bit extra for careful trimming and really
good CMRR. That is probably not a big deal for you, but it can be.

But you know, I have heard some decent sounding equipment using LM301s,
built by people who understood the limitations and kept them within fairly
narrow gain ranges (and never put anything reactive in the feedback loop,
no no no). So go figure.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Scott Dorsey]

Kevin Aylward > wrote:
>
>Immaterial. Numerous tests with alleged "golden ears" over many years have
>all concluded that specifications such as given by those sort of amp are
>simply not audibly detectable. I am not making my claims based on any
>specific implementation. My claim is based on knowing what imperfections can
>be shown to be not audibly detectable in actual experiments, and what
>imperfections the op-amps have.

Bob Pease does a great demo where he has a board with an op-amp on it and
an attenuator. He runs a 1 KHz square wave through it, and shows it on
a scope. It's a beautiful, crisp square wave.

Then he pulls out a larger board with 10 op amps and attenuators, and it
also produces a beautiful, crisp square wave.

Then he reaches under the lectern and pulls out a board with 100 op amps
in it, and hooks it up, and you know... the square wave doesn't look so
square any more.

Then he goes back behind the curtain and comes out carrying a huge board
with a thousand op-amps on it. He hooks up the input, output, and huge
power leads... and the thing that comes out hardly looks like a square
wave at all.

I doubt I could hear any difference in an audio signal with one stage, but
I bet even my mostly-deaf father could hear a difference in a thousand.
--scott


--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Kevin Aylward[_4_]]

"Peter Larsen" wrote in message
k...

Kevin Aylward wrote:

>> This comes across as a review in an audiophool mag, rather than a
>> somewhat objective opinion from a somewhat claimed technical NG.
>> However, to your credit, you do state that it is only your opinion.

>Good boy Kevin, you actually got the point, I stated my opinion.

>However as for your general flak: If you can not hear the
>difference between different opamps in a real world application,

I didn't say all op-amps, I said the ones you listed.

> then you
>need to do three things:

Here we go.. Yeah, its the ghosts doooo exist, but you are just not
sensitive enough to see them like I do, sort of thing. Only the chosen
priests can hear and see the emperors new clothes.

>get some proper test material, get some better
>loudspeakers

Oh... you mean those speakers generating 0.001% THD made from unobtainium?

>and - most important: learn to really listen.

I do. Oh... but I do...

>I have done opamp listening tests by replacing opamps in electronic
>cross-overs, as headphone-drivers, in a single opamp stage RIAA and via a
>CD
>with samples of music recorded via a RIAA stage under development by
>someone
>else, the latter test was blind and involved a large handful of testers.

As have I, 30 years ago when my hearing was a tad better than today.
However, it means squat. What matters is properly conducted tests. When
these are done, its a no contest. Amplifiers that meet certain performance
levels are indistinguishable. Those op-amps meet such levels of performance.

>After the blind evaluations came in - panel got a cd in the mail and
>returned it with a commented rating of sound examples - the gudanuff opamps
>suggested by your local companion in the credo "0.001 percent sounds the
>same from all vendors" were replaced with what the testers evaluation
>suggested. Testers listened blind and separately and all pointed in the
>same
>general direction and a sound business decision was based on the opinion of
>the test panel because of the large agreement in it.

Yeah sure mate. Publish the full documented test, then repeat it under
controls set by James Randi and then claim the $1M.

http://www.randi.org/site/index.php/jref-news/116-10-tips-for-dealing-with-randi.html

http://www.douglas-self.com/ampins/pseudo/subjectv.htm is well worth
reading.

I used to believe in all sorts of daft things when I was 12. Indeed, even at
27, after much (erroneous) study I thought Einstein was wrong. It takes a
bit of courage to finally admit to oneself that they were totally out to
lunch on some ideas that they had. We often believe what we want to believe,
not what the evidence truly tells us. Fundamentally, I don't believe in
magic (http://www.kevinaylward.co.uk/replicators/magic.html). Many audio
claims are equivalent to magic.

>You may also want to read up on the designers intentions with the
>OP275/176,
>applications notes should be readily available. Since you say that it does
>not have a sonic signature that is different from the gudanuff brand, then
>you contradict those designers stated aim of making it different via
>distortion spectra shaping.

I design analogue integrated circuits every day, 37 1/2 hours a week, and
what I will say here, is that you have been led way up the garden path. I
dare say I do contradict the marketing departments, the designers are
another matter entirely.


Kevin Aylward B.Sc.
www.kevinaylward.co.uk

[Mike Rivers]

On 2/19/2011 2:57 PM, Kevin Aylward wrote:

> Immaterial.

Horse****.

> Numerous tests with alleged "golden ears" over
> many years have all concluded that specifications such as
> given by those sort of amp are simply not audibly
> detectable.

I'll agree with that, and add that most listeners do not
have golden ears.

> I am not making my claims based on any specific
> implementation. My claim is based on knowing what
> imperfections can be shown to be not audibly detectable in
> actual experiments, and what imperfections the op-amps have.

And I'm saying that there are some audible imperfections
that are not described by the specifications in an op-amp
spec sheet. That's all. Do you not agree that not everything
using the same op-amp sounds the same?

>> Many things with similar "insignificant" amounts of
>> measured distortion actually sound different.
>
> Oh? Care to cite some experiments in support of that position?

Yes. Right in my own studio. But I didn't publish a paper on
it because I figured that either everybody knew that or
someone would think it was nonsense. I see that we have some
of each right here. But then I'm talking about listening to
real audio program material, not single frequency or dual
frequency lab tests that yield numbers.

> I disagree. There are a lot of boxes out there because
> everyone wants a piece of the same pie. Companies will
> invent any reason they can to get sales.

And they sell to different people because some like one and
some like the other. And some are influenced by the
preferences of others.

> I am *not* claiming that *all* op-amps will sound the same,
> only those that meet certain, somewhat indefinite criteria
> will. A 741 set to a closed loop gain of 100 will have
> significantly measurable IMD distortion at high audio
> frequencies, and probably be quite audibly detectable.

OK, but who designs with 741s any more? And for that matter,
who cares about distortion if, to a presumably educated
listener, it sounds good? Isn't that why tube gear,
particularly tube guitar amplifiers, sells well?

> So, there may well be audible differences between a 741 and
> OPA602. My claim is for the range of modern, high
> performance op-amps that are, essentially, audibly "perfect".

OK, so why don't all manufacturers use the same op amp that
meets their purchase specifications?

> As far as experience goes, I myself have been a guitar
> playing, band performing, electronics bod since the age of
> 11. I am now 52. So, on that level, I am at least equally
> qualified.

There's nothing there that tells me that you're a critical
listener. You're probably half deaf and have no business
making aural judgments. Leave that to the golden ears and
those who just have good taste.

> My claims are based on being
> acquainted with sufficient peer review published literature
> on the subject. Double-blind controlled tests have
> irrefutable determined what sort of levels of frequency
> response and distortion errors are not audibly detectable.

I'm not making any claims, but I don't have a problem with
anything that doesn't sound bad to me. Double blind testing
doesn't impress me. I know that I hear differently every
day, and it wouldn't surprise me if, in a double blind test
between two devices, one day I'd think one sounded like the
standard, and another day I'd think the other one did. I
just don't see the point in arguing tiny percentages of
distortion in building blocks when there are bigger things
that affect what gets to our brains. Some are electronic,
some are acoustic, some are psychological. But I acknowledge
that there are some things that sound bad all the time.
They're not really part of this discussion.


--
"Today's production equipment is IT based and cannot be
operated without a passing knowledge of computing, although
it seems that it can be operated without a passing knowledge
of audio." - John Watkinson

http://mikeriversaudio.wordpress.com - useful and
interesting audio stuff

[PStamler]

While the religious war rages on, may I suggest downloading Sam
Groner's tests of the performance of a heap of popular (and not so
popular) opamps:

www.tinyurl.com/opamptests

It's about 35 megs of the most valuable data a designer who works with
opamps can find. They show that opamps spec'd at .001% distortion can
produce truly ghastly distortion levels under the kind of real-world
conditions they face in pro audio. They show the major problems most
opamps have with operating in non-inverting mode. And driving pro
loads. And being driven by higher impedances. And, sometimes, dealing
with thermal problems at low frequencies. And distortion at higher
frequencies (how about 1-2% THD at 20kHz, presumably with IMD to
match)?

Just possibly, some of his data may point to *why* pieces of equipment
may sound different from one another. I've posted comparisons of a
good-sounding equalizer (set flat) and a bad-sounding one (also set
flat), with CCIR-IM measured at a modest +10dBu. The good one has
reasonably clean performance; the bad one has a forest of IM products
surrounding the two test frequencies. Not coincidentally, this product
has the reputation of sounding "crispy-crunchy"; with that much high-
frequency IM, I don't wonder.

To the original poster: If Groner's data are meaningful, the 5532 is
still not a bad choice; you can get a modest but real improvement in
distortion performance (and a MAJOR improvement in DC performance) by
using an LT1469 (single amp LT1468). Not too expensive, either.

I think one reason the 797 never caught on, in addition to the high
cost, was that it acquired a reputation for being hard to stabilize in
real-world applications. Its input noise current is also comparatively
high, which makes it persnickety about input impedances.

If you really need to drive 600 ohms often, though, a John Hardy JH990
is your best friend. Not cheap, but very, very, very clean into a 600
ohm load.

Peace,
Paul

[Ralph Barone]

In article
>,
PStamler > wrote:

> While the religious war rages on, may I suggest downloading Sam
> Groner's tests of the performance of a heap of popular (and not so
> popular) opamps:
>
> www.tinyurl.com/opamptests
>
> It's about 35 megs of the most valuable data a designer who works with
> opamps can find. They show that opamps spec'd at .001% distortion can
> produce truly ghastly distortion levels under the kind of real-world
> conditions they face in pro audio. They show the major problems most
> opamps have with operating in non-inverting mode. And driving pro
> loads. And being driven by higher impedances. And, sometimes, dealing
> with thermal problems at low frequencies. And distortion at higher
> frequencies (how about 1-2% THD at 20kHz, presumably with IMD to
> match)?
>
> Just possibly, some of his data may point to *why* pieces of equipment
> may sound different from one another. I've posted comparisons of a
> good-sounding equalizer (set flat) and a bad-sounding one (also set
> flat), with CCIR-IM measured at a modest +10dBu. The good one has
> reasonably clean performance; the bad one has a forest of IM products
> surrounding the two test frequencies. Not coincidentally, this product
> has the reputation of sounding "crispy-crunchy"; with that much high-
> frequency IM, I don't wonder.
>
> To the original poster: If Groner's data are meaningful, the 5532 is
> still not a bad choice; you can get a modest but real improvement in
> distortion performance (and a MAJOR improvement in DC performance) by
> using an LT1469 (single amp LT1468). Not too expensive, either.
>
> I think one reason the 797 never caught on, in addition to the high
> cost, was that it acquired a reputation for being hard to stabilize in
> real-world applications. Its input noise current is also comparatively
> high, which makes it persnickety about input impedances.
>
> If you really need to drive 600 ohms often, though, a John Hardy JH990
> is your best friend. Not cheap, but very, very, very clean into a 600
> ohm load.
>
> Peace,
> Paul

Thanks. Good to see a meaningful contribution to the discussion that
attempts to bring both sides together, rather than widening the gulf.

[Kevin Aylward[_4_]]

"Mike Rivers" wrote in message
...

On 2/19/2011 2:57 PM, Kevin Aylward wrote:


>> Numerous tests with alleged "golden ears" over
>> many years have all concluded that specifications such as
>> given by those sort of amp are simply not audibly
>> detectable.

>I'll agree with that, and add that most listeners do not have golden ears.

>> I am not making my claims based on any specific
>> implementation. My claim is based on knowing what
>> imperfections can be shown to be not audibly detectable in
>> actual experiments, and what imperfections the op-amps have.

>And I'm saying that there are some audible imperfections that are not
>described by the specifications in an op-amp spec sheet.

In principle yes. Often data sheets deliberately miss data which is known to
show the device in a bad light. I know this as fact as when I was an
analogue i.c. design engineer at Texas Instruments (1997-2001), I wrote the
data sheets, or gave the data to marketing to make them. It is also time
consuming to detail everything.

However...experience tells me certain things. An op-amp that is
*specifically* designed at targeting the silly low distortion audio market
is very unlikely to have fundamental flaws that impede their use in the
market they are designed for. One does not need a statement from the pope,
that he is a catholic.


>That's all. Do you not agree that not everything using the same op-amp
>sounds the same?

Yes. A 741 set to a gain of unity and a gain of 1000, at the same output
level, will probably sound different. A OPA604 set to unity gain and a gain
of ten, at the same output level will sound the same.

However, by construction, signal levels need to be set such that various
obvious artefacts are not generated due to the systems not being operated in
their guaranteed linear operating range. For example, in unity gain, most
op-amps will clip on the input before the output clips. My claim about
specific op-amps is when they are not frequency, voltage or current limiting
conditions. A competent design will ensure that this is usually the case.

>>>> Many things with similar "insignificant" amounts of
>>>> measured distortion actually sound different.
>>
>> Oh? Care to cite some experiments in support of that position?

>Yes. Right in my own studio.

and I am sure you realise that this proves nothing.

>But I didn't publish a paper on it because I figured that either everybody
>knew that or someone would think it was nonsense. I see that we have some
>of each right here. But then I'm talking about listening to real audio
>program material, not single frequency or dual frequency lab tests that
>yield numbers.

Here we go again..Yes, I agree, this is all quite nonsense. I will quote
from Douglas Self again:

http://www.douglas-self.com/ampins/pseudo/subjectv.htm

Quote:

"5: ARTICLES OF FAITH: THE TENETS OF SUBJECTIVISM.
All of the alleged effects listed below have received considerable
affirmation in the audio press, to the point where some are treated as
facts. The reality is that none of them has in the last fifteen years proved
susceptible to objective confirmation. This sad record is perhaps equalled
only by students of parapsychology. I hope that the brief statements below
are considered fair by their proponents. If not I have no doubt I shall soon
hear about it:

"Sinewaves are steady-state signals that represent too easy a test for
amplifiers, compared with the complexities of music."
This is presumably meant to imply that sinewaves are in some way
particularly easy for an amplifier to deal with, the implication being that
anyone using a THD analyser must be hopelessly naive. Since sines and
cosines have an unending series of non-zero differentials, "steady" hardly
comes into it. I know of no evidence that sinewaves of randomly varying
amplitude (for example) would provide a more searching test of amplifier
competence.
I believe this outlook is the result of anthropomorphic thinking about
amplifiers; treating them as though they think about what they amplify.
Twenty sinewaves of different frequencies may be conceptually complex to us,
and the output of a symphony orchestra much more so, but to an amplifier
both composite signals resolve to a single instantaneous voltage that must
be increased in amplitude and presented at low impedance. The rate of change
of this voltage has a maximum set by the frequency response and amplitude
capability of the channel and is not generally greater for more complex
signals; you do not get hgher slew rate with bigger orchestras. You must
remember that an amplifier has no perspective on the signal arriving at its
input, but literally takes it as it comes. "

Mike, you should understand that there is nothing new in your arguments, and
that they have all been dismissed soundly by competent analysis. The physic
of electronics processing of *any* signal is well understood. There is
absolutely nothing magic about audio.

>> I am *not* claiming that *all* op-amps will sound the same,
>> only those that meet certain, somewhat indefinite criteria
>> will. A 741 set to a closed loop gain of 100 will have
>> significantly measurable IMD distortion at high audio
>> frequencies, and probably be quite audibly detectable.

>OK, but who designs with 741s any more? And for that matter, who cares
>about distortion if, to a presumably educated listener, it sounds good?

Designing equipment that is not meant to distort, not distort is
straightforward today. Claims that there are audible differences between the
various "high end" op-amps, when operated correctly and designed into
products competently, are not supportable.

>Isn't that why tube gear, particularly tube guitar amplifiers, sells well?

Yes, many people like distortion.

>> So, there may well be audible differences between a 741 and
>> OPA602. My claim is for the range of modern, high
>> performance op-amps that are, essentially, audibly "perfect".

>OK, so why don't all manufacturers use the same op amp that meets their
>purchase specifications?

Because the purchase specifications are not not sufficient for all
manufactures. Specification do differentiate products.
Purchase negotiation price differences.
Because many engineers at audio companies, are not.
Personal preferences of the boss.
etc...


>> As far as experience goes, I myself have been a guitar
>> playing, band performing, electronics bod since the age of
>> 11. I am now 52. So, on that level, I am at least equally
>> qualified.

>There's nothing there that tells me that you're a critical listener. You're
>probably half deaf and have no business making aural judgments. Leave that
>to the golden ears and those who just have good taste.

And this is a good debating argument?

>> My claims are based on being
>> acquainted with sufficient peer review published literature
>> on the subject. Double-blind controlled tests have
>> irrefutable determined what sort of levels of frequency
>> response and distortion errors are not audibly detectable.

>I'm not making any claims, but I don't have a problem with anything that
>doesn't sound bad to me. Double blind testing doesn't impress me.

Then you believe in magic.

>I know that I hear differently every day, and it wouldn't surprise me if,
>in a double blind test between two devices, one day I'd think one sounded
>like the standard, and another day I'd think the other one did.

So, you admit that it is psychological, not real.

> I just don't see the point in arguing tiny percentages of distortion in
> building blocks when there are bigger things that affect what gets to our
> brains.

And that's exactly my point. The blemishes in the op-amps under discussion
are a non-event.

Kevin Aylward B.Sc.
www.kevinaylward.co.uk

[Kevin Aylward[_4_]]

"Scott Dorsey" wrote in message ...

Kevin Aylward > wrote:
>
>>Immaterial. Numerous tests with alleged "golden ears" over many years have
>>all concluded that specifications such as given by those sort of amp are
>>simply not audibly detectable. I am not making my claims based on any
>>specific implementation. My claim is based on knowing what imperfections
>>can
>>be shown to be not audibly detectable in actual experiments, and what
>>imperfections the op-amps have.

>Bob Pease does a great demo where he has a board with an op-amp on it and
>an attenuator. He runs a 1 KHz square wave through it, and shows it on
>a scope. It's a beautiful, crisp square wave.

>Then he pulls out a larger board with 10 op amps and attenuators, and it
>also produces a beautiful, crisp square wave.

>Then he reaches under the lectern and pulls out a board with 100 op amps
>in it, and hooks it up, and you know... the square wave doesn't look so
>square any more.

>Then he goes back behind the curtain and comes out carrying a huge board
>with a thousand op-amps on it. He hooks up the input, output, and huge
>power leads... and the thing that comes out hardly looks like a square
>wave at all.

>I doubt I could hear any difference in an audio signal with one stage, but
>I bet even my mostly-deaf father could hear a difference in a thousand.


I do not know what the point is here. Sure, if we have 1000 741s set to a
gain of 100, signals wont look too hot at the end of the chain. 100 of the
specific op-amps under discussion, set to a gain of 10 (with following
attenuators), will look a lot better. Of hand I cant say just how much
because I would have to set up a simulation.


Kevin Aylward B.Sc.

www.kevinaylward.co.uk

[Kevin Aylward[_4_]]

"PStamler" wrote in message
...

>While the religious war rages on, may I suggest downloading Sam
>Groner's tests of the performance of a heap of popular (and not so
>popular) opamps:

>www.tinyurl.com/opamptests

>It's about 35 megs of the most valuable data a designer who works with
>opamps can find.

Oh...

>They show that opamps spec'd at .001% distortion can
>produce truly ghastly distortion levels under the kind of real-world
>conditions they face in pro audio.

Where?

This is a big document.

I looked at the summary:

"With 600 ohm loading, this figure deteriorates typically by one or two
orders of magnitude"

This would make 0.001% 0.1%, hardly "ghastly"

I am quite aware of real world conditions. I have been a real world analogue
designer for 30 years. You will need to clarify exactly what conditions and
what specific op-amp you are referring to, that this document shows that
there are alleged issues with. The summary did not highlight any fundamental
flaws with any op-amp as far as I read.

>They show the major problems most
>opamps have with operating in non-inverting mode.

Oh? "major" for "most". Which ones?

> And driving pro
>loads. And being driven by higher impedances. And, sometimes, dealing
>with thermal problems at low frequencies. And distortion at higher
>frequencies (how about 1-2% THD at 20kHz, presumably with IMD to
>match)?

Some amps may well have highish 20Khz distortion, especially when operated
non optimally. So?

>Just possibly, some of his data may point to *why* pieces of equipment
>may sound different from one another. I've posted comparisons of a
>good-sounding equalizer (set flat) and a bad-sounding one (also set
>flat), with CCIR-IM measured at a modest +10dBu. The good one has
>reasonably clean performance; the bad one has a forest of IM products
>surrounding the two test frequencies. Not coincidentally, this product
>has the reputation of sounding "crispy-crunchy"; with that much high-
>frequency IM, I don't wonder.

This is quite a different matter entirely to issue at hand. Whether or not
what the claim above is true or not is immaterial. My comments were
specifically addressed on some allegations that certain, specific op-amps
sounded different. No one is making the daft claim that "all equipment
sounds the same".


>If you really need to drive 600 ohms often, though, a John Hardy JH990
>is your best friend. Not cheap, but very, very, very clean into a 600
>ohm load.

Other than high impedance headphones, there is pretty much no reason to
design an audio system to drive 600 ohms at all. This is just an unfortunate
carry over from telecommunication systems. Telecom systems relied on
characteristic impedances and matching to avoid effects such as "singing"
(oscillation) due to the 2 to 4 wire nature of the system. i.e. amplifiers
for both directions at once on a single pair of wires. There is no rational
reason to continue with 600 ohm loads for audio systems.

Power amp inputs and other audio bits of kit should be designed to be > 10K
ohms, preferably >100k ohms, and driven by < 100 ohm drive impedances.
Conceivable, one could have a mixer driving 100 X 1KW amps in parallel for a
concert, but these form a tiny fraction of systems. Transmission line
matching is not required, as the wavelengths for audio at the cable lengths
usually used, are so large that transmission line effects are quite
insignificant.

Kevin Aylward B.Sc.
www.kevinaylward.co.uk

[Scott Dorsey]

Kevin Aylward > wrote:
>
>In principle yes. Often data sheets deliberately miss data which is known to
>show the device in a bad light. I know this as fact as when I was an
>analogue i.c. design engineer at Texas Instruments (1997-2001), I wrote the
>data sheets, or gave the data to marketing to make them. It is also time
>consuming to detail everything.

Were you by any chance responsible for those awful Excalibur op-amps?
--scott


--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Sean Conolly]

"Arny Krueger" > wrote in message
...
> Here is a heck of a freebie info if you want some *light reading* about op
> amps:
>
> http://focus.ti.com/lit/an/slod006b/slod006b.pdf

Thanks!

Sean

[William Sommerwerck]

> http://focus.ti.com/lit/an/slod006b/slod006b.pdf

I browsed the first few chapters, and it's really good.

[Peter Larsen[_3_]]

William Sommerwerck wrote:

>> http://focus.ti.com/lit/an/slod006b/slod006b.pdf

> I browsed the first few chapters, and it's really good.

A very helpful contribution to the debate indeed, thanks Arny!

Kind regards

Peter Larsen

[Mike Rivers]

On 2/20/2011 7:17 AM, Kevin Aylward wrote:

> In principle yes. Often data sheets deliberately miss data
> which is known to show the device in a bad light. I know
> this as fact as when I was an analogue i.c. design engineer
> at Texas Instruments (1997-2001), I wrote the data sheets,
> or gave the data to marketing to make them. It is also time
> consuming to detail everything.

So why do we have so many op amps anyway?

> Yes. A 741 set to a gain of unity and a gain of 1000, at the
> same output level, will probably sound different. A OPA604
> set to unity gain and a gain of ten, at the same output
> level will sound the same.

But will a 741 and an OPA604 in the same circuit at the same
gain sound the same?

> However, by construction, signal levels need to be set such
> that various obvious artefacts are not generated due to the
> systems not being operated in their guaranteed linear
> operating range.

OK, so they "need to be set." Perhaps some do and some
don't. This may make an op amp that sounds good in one box
sound bad in another. Those aren't things that can be
predicted by reading the op amp spec sheet.

> Mike, you should understand that there is nothing new in
> your arguments, and that they have all been dismissed
> soundly by competent analysis.

I'm not trying to break any new ground here. I'm just saying
that when it comes to products that I use (that is a box
full of op amps) some sound better than others, and that's
not just because one uses "better" op amps.

> Designing equipment that is not meant to distort, not
> distort is straightforward today. Claims that there are
> audible differences between the various "high end" op-amps,
> when operated correctly and designed into products
> competently, are not supportable.

That may be true, but differences, whether supportable by op
amp specs or not, are audible between various end products.

>> OK, so why don't all manufacturers use the same op amp
>> that meets their purchase specifications?
>
> Because the purchase specifications are not not sufficient
> for all manufactures. Specification do differentiate products.
> Purchase negotiation price differences.
> Because many engineers at audio companies, are not.
> Personal preferences of the boss.

I guess that at least in part answers my question about why
we have so many op amps.

>> I'm not making any claims, but I don't have a problem with
>> anything that doesn't sound bad to me. Double blind
>> testing doesn't impress me.
>
> Then you believe in magic.

No, I don't need to believe in anything of the sort. I use
what's available. I don't use what I think sounds bad. I
don't care what you use, or why you use it.

>> I know that I hear differently every day, and it wouldn't
>> surprise me if, in a double blind test between two
>> devices, one day I'd think one sounded like the standard,
>> and another day I'd think the other one did.
>
> So, you admit that it is psychological, not real.

No, I'm suggesting that human hearing is not consistent. So
fine differences are not of the greatest importance to me.

>> I just don't see the point in arguing tiny percentages of
>> distortion in building blocks when there are bigger things
>> that affect what gets to our brains.
>
> And that's exactly my point. The blemishes in the op-amps
> under discussion are a non-event.

But the differences in what we as end users of products that
use op amps are apparent, more so than the blemishes of the
op amps.



--
"Today's production equipment is IT based and cannot be
operated without a passing knowledge of computing, although
it seems that it can be operated without a passing knowledge
of audio." - John Watkinson

http://mikeriversaudio.wordpress.com - useful and
interesting audio stuff

[PStamler]

Kevin, instead of talking at such length, I suggest you read the data
from Groner's tests. Everyone else probably should too:

www.tinyurl.com/opamptests

People keep saying that, *properly applied*, opamps can provide clean
performance. Groner's data give useful information about what
applications may be proper for various chips. For example, some give
reasonably decent performance in inverting mode, into high-Z loads,
but hideous performance in non-inverting mode or driving 600 ohms; a
good (or rather bad) example is the TL071, which in Groner's tests
exhibits something like 0.017% THD at 1kHz, +20dBu, inverting, but >1%
THD, same conditions but inverting, and >2%, inverting but driving 600
ohms.

Read the data. There's a good deal of myth-busting in there too; among
the myths the data begin to refute is the one that says ICs are awful,
but discrete opamps are much better. (With a couple of exceptions --
like the Hardy 990 -- the discrete opamps Groner tested perform a lot
worse than, say, a 5532.)

A long evening with Groner's book will persuade you that (a) it's
possible for IC opamps to offer very high performance; (b) you need to
match the IC to the job carefully (and Groner gives valuable data to
help do so); and (c) it's still hard to beat a 5532/4. Not Impossible
in the right application, but hard. Except for DC characteristics, of
course.

Peace,
Paul

[hank alrich]

PStamler > wrote:

> Kevin, instead of talking at such length, I suggest you read the data
> from Groner's tests. Everyone else probably should too:
>
> www.tinyurl.com/opamptests

Thanks, Paul.

--
shut up and play your guitar * http://hankalrich.com/
http://armadillomusicproductions.com/who'slistening.html
http://www.sonicbids.com/HankandShaidriAlrichwithDougHarman

[PStamler]

On Feb 20, 5:11*pm, PStamler > wrote:
> a
> good (or rather bad) example is the TL071, which in Groner's tests
> exhibits something like 0.017% THD at 1kHz, +20dBu, inverting, but >1%
> THD, same conditions but inverting,

Make that "same conditions but non-inverting". Slip of the finger.

Peace,
Paul

[Arny Krueger]

"Kevin Aylward" > wrote in
message
> "PStamler" wrote in message
> ...
>
>> While the religious war rages on, may I suggest
>> downloading Sam Groner's tests of the performance of a
>> heap of popular (and not so popular) opamps:
>
>> www.tinyurl.com/opamptests
>
>> It's about 35 megs of the most valuable data a designer
>> who works with opamps can find.
>
> Oh...
>
>> They show that opamps spec'd at .001% distortion can
>> produce truly ghastly distortion levels under the kind
>> of real-world conditions they face in pro audio.
>
> Where?

> This is a big document.

> I looked at the summary:

> "With 600 ohm loading, this figure deteriorates typically
> by one or two orders of magnitude"

> This would make 0.001% 0.1%, hardly "ghastly"

Not only that, but it presumes full-scale signal at 20 KHz which rarely if
ever exist in the natural world.


> Other than high impedance headphones, there is pretty
> much no reason to design an audio system to drive 600
> ohms at all.

....for exactly the reason stated below:

> Power amp inputs and other audio bits of kit should be
> designed to be > 10K ohms, preferably >100k ohms, and
> driven by < 100 ohm drive impedances. Conceivable, one
> could have a mixer driving 100 X 1KW amps in parallel for
> a concert, but these form a tiny fraction of systems.


And, power amps generally require less than 2 volts for full output.

Besides, if you want to drive low impedances with 10 volts RMS, there are
reasonbly-priced op amps that will do the job. Just don't damn the more
humble stuff because its not the same as the best.

[Arny Krueger]

"PStamler" > wrote in message

> On Feb 20, 5:11 pm, PStamler > wrote:
>> a
>> good (or rather bad) example is the TL071, which in
>> Groner's tests exhibits something like 0.017% THD at
>> 1kHz, +20dBu, inverting, but >1% THD, same conditions
>> but inverting,
>
> Make that "same conditions but non-inverting". Slip of
> the finger.

Had to be, because with few exceptions inverting operation is easier to pull
off.

[Arny Krueger]

"Peter Larsen" > wrote in message
k
> William Sommerwerck wrote:
>
>>> http://focus.ti.com/lit/an/slod006b/slod006b.pdf
>
>> I browsed the first few chapters, and it's really good.
>
> A very helpful contribution to the debate indeed, thanks
> Arny!

I stumbled into it by accident. No talent required! ;-)

[g]

On Feb 19, 5:41*am, "Kevin Aylward" >
wrote:
> "Peter Larsen" *wrote in message
>
> k...
>
> mcp6453 wrote:
> >> The first console I ever build used NE5532 op amps. The last time I
> >> designed a circuit, I used AD797ANs. In fact, there's a tube of the
> >> around here somewhere, but they are not to be found? The going price
> >> seems to be about $10 each from Digikey. What's the modern day
> >> preferred low noise op amp usable as a line driver, able to drive 600
> >> ohms?
> >OPAx134 has the same forward sounding charm as the 553x AND a cleaner
> >treble and tighter bass. my OPINION the perfect compromise between the
> >slightly glassy Burr Brown x604 and the gentle and sweet but not overbright
> >OP 275. I did try replacing OPA176's in a IFM cross-over with 134's and got
> >a more forward sound
>
> This comes across as a review in an audiophool mag, rather than a somewhat
> objective opinion from a somewhat claimed technical NG. However, to your
> credit, you do state that it is only your opinion.
>
> Having looked at the (distortion) specs for the OPA602 and the OP275, it is
> my technical evaluation that claims that they sound audibly different
> (except possible in overload, noise) are simply delusional. This is not an
> opinion, but a fact, as in, evolution is a fact.
>
> I also looked at the data sheet for the OPA134, which has, for example,
> graphs of IMD of the OPA134 and OP176 overlaid, and again, it is my
> technical evaluation that claims that they sound audibly different (except
> possible in overload, noise) are simply delusional.
>
> All of the above opamps, have truly insignificant (~0.001%) LF distortion,
> and low enough HF distortion that alleged audible differences between them
> are simply not credible, when used competently, i.e. closed loop gains not
> set to 1000. Normal closed frequency response of all, in normal designs will
> be 50khz+, hence not audibly different.
>
> I would like to say I am amazed that this sort of audiophool *nonsense is is
> still being abounded about, but unfortunately, I am not. Only dismayed.
>
> Kevin Aylward B.Sc.http://www.kevinaylward.co.uk/ee/index.html

Believe what you hear.

greg

[g]

On Feb 19, 9:49*am, mcp6453 > wrote:
> On 2/19/2011 9:39 AM, Scott Dorsey wrote:
>
>
>
> > And part of the problem here is that test circuits aren't the same as real
> > world circuits. *The input and output impedances have an awful lot to do
> > with the distortion characteristic of an op-amp in the real world, and
> > consequently the op-amp that sounds best in one circuit may not operate
> > reliably at all in another.
>
> > Also there are some interesting stability issues on some of the commercial
> > monolithics... look at the Bode plot on the very-popular OPA2604.
>
> > On top of all this, in listening tests you can find that fairly subtle
> > changes to the noise floor get perceived as tonal changes. *In a high
> > gain circuit, the noise floor on an OPA2604 and an NE5532 will be very
> > different.
>
> > Some of this matters, some of it doesn't. *Some of it only matters when you
> > have a huge number of identical stages in the same signal path. *Only listening
> > will tell you for sure if there's a problem, only careful measurement will
> > tell you what that problem is.
>
> It looks like I may just use 5532s for this project. They're still good op amps
> although there are better one. They sure are cheap. I've spent most of the
> morning reading op amp opinions. One thing is absolutely true: There is no
> consensus.
>
> I'm still surprised that there is not much talk of the AD797. Maybe it didn't
> prove as useful for audio as it was thought to be. Maybe it's too expensive.

Its usually good to stay with the same inputs. fets may change things
for the worse.

greg

[Mike Rivers]

On 2/20/2011 6:11 PM, PStamler wrote:

> A long evening with Groner's book will persuade you that (a) it's
> possible for IC opamps to offer very high performance; (b) you need to
> match the IC to the job carefully (and Groner gives valuable data to
> help do so); and (c) it's still hard to beat a 5532/4.

Yup. And there are even commercial manufacturers who who
tend to use the same op amp throughout a design, which, as
your reference suggests, might be fine in some places and
not so good in other places.

There isn't a lot of call for an op amp to drive a 600 ohm
load in the universe of people who are looking for equipment
that doesn't come from (and cost like) that which is
designed by designers who are aware of the differences in
application. So some are able to get away with less
thoughtful designs at lower cost.

Also, product specifications tend to be misleading to those
who don't fully understand them. For example, a specified
output impedance of 50 ohms won't necessarily properly drive
a 600 ohm load, but will probably be fine into a 5000 ohm
load which is more typical of the input of devices that
would be used in conjunction with gear which has a lower
cost output design. People who buy $250 mixers don't use
them to drive $20,000 re-purposed disk mastering compressors.

Perhaps a better question to ask (or maybe it was asked - I
forgot what started this discussion) is:

What IC op amp can I buy a handful of to incorporate
anywhere in my multi-channel mic preamp without regard for
perhaps mis-applying it and getting poor performance at one
or more points in the chain? I don't have an output drive
requirement for loads under 2.5K, nor do I need greater
output level than +24 dBu. I can manage internal levels so
as to keep the op amps within the range where they have
adequately low distortion.

Any recommendations? Or is this the sort of question that
promotes an answer that suggests one op amp for the input, a
different one for the output, and perhaps a third one or two
in between?

In other words, is there an op amp that will let me use it
anywhere in this "product" without worry that I'm
compromising performance?

I used to work for someone who used 2N706 transistors for
everything. I still have some in my cigar box of old
transistors. But it wasn't audio equipment. ;)





--
"Today's production equipment is IT based and cannot be
operated without a passing knowledge of computing, although
it seems that it can be operated without a passing knowledge
of audio." - John Watkinson

http://mikeriversaudio.wordpress.com - useful and
interesting audio stuff

[Mark]

On Feb 21, 7:47*am, "Arny Krueger" > wrote:
> "Peter Larsen" > wrote in message
>
> k
>
> > William Sommerwerck wrote:
>
> >>>http://focus.ti.com/lit/an/slod006b/slod006b.pdf
>
> >> I browsed the first few chapters, and it's really good.
>
> > A very helpful contribution to the debate indeed, thanks
> > Arny!
>
> I stumbled into it by accident. No talent required! ;-)



more good reading:

Section 5 is particularly interesting
http://www.linearaudio.nl/Documents/Baxandall%20power%20amp%20design.pdf


http://www.analog.com/library/analogdialogue/archives/39-05/op_amp_applications_handbook.html

http://focus.ti.com/lit/an/slod006b/slod006b.pdf

also look up Fully Differential Amplifiers

Mark

[hank alrich]

G > wrote:

> On Feb 19, 5:41 am, "Kevin Aylward" >
> wrote:
> > "Peter Larsen" wrote in message
> >
> > k...
> >
> > mcp6453 wrote:
> > >> The first console I ever build used NE5532 op amps. The last time I
> > >> designed a circuit, I used AD797ANs. In fact, there's a tube of the
> > >> around here somewhere, but they are not to be found? The going price
> > >> seems to be about $10 each from Digikey. What's the modern day
> > >> preferred low noise op amp usable as a line driver, able to drive 600
> > >> ohms?
> > >OPAx134 has the same forward sounding charm as the 553x AND a cleaner
> > >treble and tighter bass. my OPINION the perfect compromise between the
> > >slightly glassy Burr Brown x604 and the gentle and sweet but not overbright
> > >OP 275. I did try replacing OPA176's in a IFM cross-over with 134's and got
> > >a more forward sound
> >
> > This comes across as a review in an audiophool mag, rather than a somewhat
> > objective opinion from a somewhat claimed technical NG. However, to your
> > credit, you do state that it is only your opinion.
> >
> > Having looked at the (distortion) specs for the OPA602 and the OP275, it is
> > my technical evaluation that claims that they sound audibly different
> > (except possible in overload, noise) are simply delusional. This is not an
> > opinion, but a fact, as in, evolution is a fact.
> >
> > I also looked at the data sheet for the OPA134, which has, for example,
> > graphs of IMD of the OPA134 and OP176 overlaid, and again, it is my
> > technical evaluation that claims that they sound audibly different (except
> > possible in overload, noise) are simply delusional.
> >
> > All of the above opamps, have truly insignificant (~0.001%) LF distortion,
> > and low enough HF distortion that alleged audible differences between them
> > are simply not credible, when used competently, i.e. closed loop gains not
> > set to 1000. Normal closed frequency response of all, in normal designs will
> > be 50khz+, hence not audibly different.
> >
> > I would like to say I am amazed that this sort of audiophool nonsense is is
> > still being abounded about, but unfortunately, I am not. Only dismayed.
> >
> > Kevin Aylward B.Sc.http://www.kevinaylward.co.uk/ee/index.html
>
> Believe what you hear.
>
> greg

And how's that spec sheet sound, anyway?

--
shut up and play your guitar * http://hankalrich.com/
http://armadillomusicproductions.com/who'slistening.html
http://www.sonicbids.com/HankandShaidriAlrichwithDougHarman

[Arny Krueger]

"G" > wrote in message

>
> Believe what you hear.

Many people hear what they believe. High end audio is largely based on this
concept.

[PStamler]

On Feb 21, 8:26*am, Mike Rivers > wrote:

> Perhaps a better question to ask (or maybe it was asked - I
> forgot what started this discussion) is:
>
> What IC op amp can I buy a handful of to incorporate
> anywhere in my multi-channel mic preamp without regard for
> perhaps mis-applying it and getting poor performance at one
> or more points in the chain? I don't have an output drive
> requirement for loads under 2.5K, nor do I need greater
> output level than +24 dBu. I can manage internal levels so
> as to keep the op amps within the range where they have
> adequately low distortion.
>
> Any recommendations? *Or is this the sort of question that
> promotes an answer that suggests one op amp for the input, a
> different one for the output, and perhaps a third one or two
> in between?

If you really want to maximize performance, it's best to optimize each
stage individually. But if you really want to use one chip everyplace,
I'd say the LT1469 is your pal. (Well, almost everyplace -- I'd still
prefer a FET input for EQ stages, though, for DC reasons.)

Peace,
Paul