Re: op-amps with wide open-loop bandwidth ?

Dave Moore wrote:
"Kevin Aylward" <see_website@xxxxxxxxxxxxx> wrote in message
news:7TRSf.210450$YJ4.75991@xxxxxxxxxxxxxxxxxxxxxxxxxxxx
Dave Moore wrote:

I'm aware of opamp theory and I understand why some engineers
think there's no reason to "sonically evaluate" opamps.
However,
I have my reasons for doing so and if anyone thinks me a fool for
doing so, so be it. They ain't got'ta be so arrogant and snobbish
about it however.

Dave Moore
( Just a fool that gets results)

Not a fool, just a bit naive.
You are giving little credit to other
qualified people who have actually looked at these issues in depth.
You are making the assumption that *they* are the fools.

You're still absolutely clueless.

Oh?

Let me clue you in. I ask for some information, instead I get
a bunch of mistaken presumptions about why I want the information
and a slew insults to go along with the mistaken presumptions.

I read your other replies in this thread, so I didnt make any
assumptions. It appears that you belive that having an open loop BW at
least equal to the audio BW is nessesary/advantages to get a good sound.
This is easily shown to erroneous. Its pretty basic physics really.

Suppose we have an op amp design. Typically, this might even have a 3db
BW of say, 100Hz! However, this is usually at enormous gain. This design
would have an aspect of distortion characterised by an open loop value,
divided/reduced by the loop gain. Suppose that we now simple added an
internal resister at the appropriate point to broadband the BW to 20Khz.
You appear to be claiming that this new configuration might sound better
ust because it has better BW. This is very unlikely. All the open loop
gain that is now thrown away at the lower frequencies, can now no longer
be used to reduce the distortion at those frequencies. It always pays
you (distortion wise) to have as much loop gain as possible over the
widest frequency range.

Of course, if the amp is slewing prior to the audio BW, then...


And no, I am not making any assumptions about whether
all engineers are the *real* audiophools. I'm quite open minded.
And I ususally believe that people have valid reasons based
on their experiences for believing the things that they do.
One scientist looking at one set of data will draw one
conclusion whereas another studying the same phenomena
with a different but contradictory dataset may draw a
completely different conclusion. In the end perhaps a new
conclusion is reached due to the controversy and everyone is
happy as they were all proved to be right.

Sure, but this open loop BW thing is a no brainer. Its not rocket
science. By itself, the open loop BW is not relevant. Its not debatable
at all. Its all very well understood. Very few physicists argue with the
general theory of relativity, and damn less qualified EE's argue about
the properties of correctly designed feedback amplifiers.



For example, some
actually have extensive professional experience in analogue design
and pro audio, from both a technical point of view and as a
musician. Some of these, after such extensive study, indeed conclude
that it there is no reason to listen to an amplifier in order to
design a straight piece of wire with gain. Its purely a technical
issue based on gains, distortion, bandwidth, noise etc.

The fact that I'd rather prove this to myself doesn't mean that
I don't respect the opinions of others. However there's a bit more to
this story that might be of significance if anyone cared to find
out first before popping off a few rounds and asking questions later.

The issue here, is that I have heard all this before from endless
numbers of people. You have said nothing new. Its already a dead
subject. You jsut are not aware of all the prio art on this matter.

I'm working on more than one front at once. Besides design,
I also do repair and upgrades for a select handful of musicians.

With all due respect, what this tells me is that you are really a
hacker/tech, not a qualified EE. You dont know what you are really
dealing with.

Sure, when I was doing my degree I used to suplement my grant (UK) by
doing repairs for a music shop. However, like, I'm going to do physical
work with that sort of stuff for other people 25 years on?

However... A few weeks ago I did purchase a Marshall AVT150 combo. The
bloody fools have an effects loop that is not inline. This means a
volume pedal, or even a phaser won't work correctly. So, I emailed for
the circuit diagram and modified the effects loop to work correctly. It
even had an irritation of having the clean channel with significantly
more gain then the distortion channel, making it hard to get maximum
volume from the distortion channels when switching. So, sure, I fixed
the gains. But this is all for me personally.

In some cases an upgrade might entail something as simple as
swapping out the opamps in a unit. Under these circumstances
there is indeed a quite noticeable difference in the sound from
one opamp to another.

Maybe for a 709 to an op-37, but in general, little chance of hearing
any difference. Been there mate.

Also, in the years that I have been
doing this I have noticed (as well as the musicians themselves)
very definite characteristics that always follow one particular
opamp around regardless of which piece of gear it is plunked
into.

Anecdotal meanderings. AB blind tests invariable show that this is all
wishfull dreaming.

As I suggested, been playing guitar and messing with electronics audio
both since I was 11. I'm not deaf, yet, so If there was any truth to
this sort of stuff, trust me, I would have discovered it. I actually
used to believe this sort of drivel until I *really* looked into it, in
detail.

Ok, so I suppose someone is going to say at this point, well
of course, because the opamps are being plunked into networks
designed for other opamps. Let me address this. Firstly, sometimes
it's not possible to get a schematic on a particular piece of gear.
Also, often there is a diminishing point of returns to attempt to
trace and draw out the schematic or reverse engineer the circuits
from the PCB's themselves. So, the 'plunk in' option proves to be
the viable alternative

Sure, some opamps might actually oscillate at VHF in the wrong circuit.


Let me give you a very recent example (like last night)
of how this might go. I have one professional musician friend
who is patiently waiting to pick up a completely SS circuit
I designed that makes his opamp-plunked and capacitor-upgraded
Pearce G1 SS amp sound like a tube amp.
Also in his rig is a TC electronics
DSP unit that sounds rather dull. I told him that I probably can
do a few things to improve the TC. Well he needs the rig for a gig
monday and told me to go ahead and do whatever I can before then.
About the only thing I'm willing to do whithin those time constraints
and no schematic towork from is an opamp-plunking.
So how'd the plunking go? First off, I took my best guess at
which opamps that I currently have on hand would best
complement the unit.
This entailed yanking the the NE5532 on the front end and
replacing it with a THS2052 and yanking the NE5532 on the tail
end and replacing it with an AD828. The result, (in guitar terms)
much more detail & clarity on the high notes and much improved
tightness and bounce on the low notes. OK, so now the A/B
test.

Quite frankly, I dont belive you.

Put the NE5532's back in. Sure 'nuff. The sound is muddy
again and it's difficult to pull out the low notes whilst finger
picking.

I don't believe you. I have done these sort of tests ad-infinitum.

If there is truly a difference, its unlikely be due to the op-amps.
Something has to be really buggered up with the circuit design itself.


So, next up, trying the unit with a number of different opamps with
'better spec's' than the NE5532's Result, all combinations produced
better sound than the original 5532's as well as imparting quite
predictably whatever flavor or characteristics I have observed
to follow each particular opamp as they have been plunked
into various units over the years.

All anecdotal meanderings. I have done the same, and it is in stark
contrast to your claims. What's more, pro double blind tests on these
*type* of tests *always* show no differences.

In the end I settled for the
two opamps that I predicted would probably best complement
the unit being that of the opamps I currently have on hand, they did
just that, they best complemenbted the unit. So riddle me this. Hoiw
did I know in advance that those two opamps were probably going to
win the contest.

I suppose you also believe in fairies, palm reading, thor, tea cup
reading, astrology?


I've also sat down with various musicians and let them decide
which opamps they best like when plunked into their gear.
9 times out of ten they choose the same ones that would have
chosen.

I dont belive you.

And this is without any pre-suggestion on my part.

So, as you can see, under these circumstances I have very
legitimate reasons for wanting as large of a pool of spec-varied
opamps that I can get my hands on and taking notes on just
what sonic characteristics tend to follow them as they're
plunked into one pice of gear after another.

I think you are simple deluding yourself. A shame really. You sound as
if you could be more objective.



Sure to design a tube distortion
circuit does require listening to it, but thats irrelevant to the
design of the actual amplifier circuits.

I agree. When I design, I start with as clean of an amp as I can
and dirty it up from there. And when it come to designing a "clean
amp" I truly do respect the skills of the more advanced engineers.
But even here, since there is in reality no such thing as an ideal
opamp, the issue becomes just how far which spec's need to go to
dissappear coloration.

As far as an audibly a straight piece of wire with gain, many op-amps
are indeed ideal. You have probably been listening to 741s running at a
closed loop gain of 100, where sure, that may well be a problem. Modern
op amps are audible perfect. Thats just the way it is despite all this
golden ears nonsense.

And this opens up another debate about how
sensitive the ear is to coloration. My conclusion is that the
coloration threshold is probably different for different people. Most
of the more accomplished and talented musicians that I've worked with
seem to have an extraordinary ability to hear subtle nuances.

They certainly claim they do, but in actual tests it all disappears.

Quite
possibly this ability may be one of the reasons that they chose to
become musicians.

Oh dear... now you really are getting out of your field of what
knowledge you have.

People become musicians because, by and large, they were introduced to
playing music at an earlier time. Its the copying or replication bit of
Darwinian evolution (http://www.anasoft.co.uk/replicators/index.html)

Or perhaps some of it is a developed skill.

Well, yes.

Just
as an EE becomes intuitively adept at analyzing a network from doing
so repeatedly, I suppose a musicians hearing ability would improve
with practice also.

Sure, but there is a physics limit.

Regardless, I've found that most
musicians notice exactly the same differences that I do.

Oh...I don't, I'm a musician. Care to provide some credible documentable
support for your "most musicians" claim?

You can claim what you like in this NG, but my experience is way
different, so I dont belive you.

So in answer
to Phil Assholesons question about whether I analyze sound for myself
or for the masses, the answer would be for myself as I don't have
time or the means to set up double-blind studies with mass
participants and 9 times out of ten, my perceptions seem to be inline
with those of the people I deal with.

I don't believe you. My experience is in direct contradiction to your
claim. Do you want me to actually dig up some of the real tests that
have been done on golden ear claims?


So, in short, respect is a two way street.
I'll respect anyone that shows me some.by getting the full story
before hurling stones.


The issue is that you are making claims, that are known to be erroneous,
both technically and experimentally.

Kevin Aylward
431infoEXTRACT@xxxxxxxxxxxxx
http://www.anasoft.co.uk
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.

"There are none more ignorant and useless,than they that seek answers
on their knees, with their eyes closed"


.