[Dirk Bruere at NeoPax]

The Auzentech card is shipped with socketed dual opamps on the
outputs, S4580P
The spec seems pretty average and subjectively a bit of top end seems
missing compared to the X-Fi. Whether the X-Fi had too much to start
with is another question...

Anyway, suggested replacements, assuming replacements are necessary at
all?
The 5532 has better slew rate, although I'm not sure about the noise
because the data sheet for the S4580P simply says: 0.8 uV RMS, RIAA
Rs=2k2, f=30kHz LPF
Bandwidth = 15MHz

Dirk

[GregS]

In article . com, "Dirk Bruere at NeoPax" wrote:
The Auzentech card is shipped with socketed dual opamps on the
outputs, S4580P
The spec seems pretty average and subjectively a bit of top end seems
missing compared to the X-Fi. Whether the X-Fi had too much to start
with is another question...

Anyway, suggested replacements, assuming replacements are necessary at
all?
The 5532 has better slew rate, although I'm not sure about the noise
because the data sheet for the S4580P simply says: 0.8 uV RMS, RIAA
Rs=2k2, f=30kHz LPF
Bandwidth = 15MHz

I don't know if the package is the same, but you might take a look at

LM4562
http://cache.national.com/ds/LM/LM4562.pdf
greg

[Dirk Bruere at NeoPax]

On 5 Apr, 14:35, (GregS) wrote:
In article . com, "Dirk Bruere at NeoPax" wrote:

The Auzentech card is shipped with socketed dual opamps on the
outputs, S4580P
The spec seems pretty average and subjectively a bit of top end seems
missing compared to the X-Fi. Whether the X-Fi had too much to start
with is another question...

Anyway, suggested replacements, assuming replacements are necessary at
all?
The 5532 has better slew rate, although I'm not sure about the noise
because the data sheet for the S4580P simply says: 0.8 uV RMS, RIAA
Rs=2k2, f=30kHz LPF
Bandwidth = 15MHz

I don't know if the package is the same, but you might take a look at

LM4562http://cache.national.com/ds/LM/LM4562.pdf
greg

That looks well impressive.
Thanks

Dirk

[Arny Krueger]

"Dirk Bruere at NeoPax" wrote in
message
ups.com
The Auzentech card is shipped with socketed dual opamps
on the outputs, S4580P
The spec seems pretty average and subjectively a bit of
top end seems missing compared to the X-Fi. Whether the
X-Fi had too much to start with is another question...

Anyway, suggested replacements, assuming replacements are
necessary at all?
The 5532 has better slew rate, although I'm not sure
about the noise because the data sheet for the S4580P
simply says: 0.8 uV RMS, RIAA Rs=2k2, f=30kHz LPF
Bandwidth = 15MHz

Have you run the Audio Rightmark program on the card to see what areas of
performance need improving?

[Dirk Bruere at NeoPax]

On 5 Apr, 15:48, "Arny Krueger" wrote:
"Dirk Bruere at NeoPax" wrote in
oglegroups.com

The Auzentech card is shipped with socketed dual opamps
on the outputs, S4580P
The spec seems pretty average and subjectively a bit of
top end seems missing compared to the X-Fi. Whether the
X-Fi had too much to start with is another question...

Anyway, suggested replacements, assuming replacements are
necessary at all?
The 5532 has better slew rate, although I'm not sure
about the noise because the data sheet for the S4580P
simply says: 0.8 uV RMS, RIAA Rs=2k2, f=30kHz LPF
Bandwidth = 15MHz

Have you run the Audio Rightmark program on the card to see what areas of
performance need improving?

No, due to the fact I've only just learned about it!
However, a lot of this audio stuff and what people consider to be a
'good' sound is pretty subjective.
See "valve amps":-)

Dirk

[Eeyore]

Dirk Bruere at NeoPax wrote:

The Auzentech card is shipped with socketed dual opamps on the
outputs, S4580P
The spec seems pretty average

Not at all. 4580s are a lot better than average. Where did you get your data
from ?


and subjectively a bit of top end seems
missing compared to the X-Fi.

It certainly won't be due to the 4580 !


Whether the X-Fi had too much to start
with is another question...

Anyway, suggested replacements, assuming replacements are necessary at
all?
The 5532 has better slew rate, although I'm not sure about the noise
because the data sheet for the S4580P simply says:

Get the original JRC data for the NJM4580. You can see that the nV/sqrt Hz is
well below that of a 5532. It's ~ 3nV/sqrt Hz.


0.8 uV RMS, RIAA
Rs=2k2, f=30kHz LPF
Bandwidth = 15MHz

The 4580 is easily the equal of the 5532. Better in fact. Notably distortion.

Graham

[Eeyore]

Dirk Bruere at NeoPax wrote:

On 5 Apr, 15:48, "Arny Krueger" wrote:
"Dirk Bruere at NeoPax" wrote in
oglegroups.com

The Auzentech card is shipped with socketed dual opamps
on the outputs, S4580P
The spec seems pretty average and subjectively a bit of
top end seems missing compared to the X-Fi. Whether the
X-Fi had too much to start with is another question...

Anyway, suggested replacements, assuming replacements are
necessary at all?
The 5532 has better slew rate, although I'm not sure
about the noise because the data sheet for the S4580P
simply says: 0.8 uV RMS, RIAA Rs=2k2, f=30kHz LPF
Bandwidth = 15MHz

Have you run the Audio Rightmark program on the card to see what areas of
performance need improving?

No, due to the fact I've only just learned about it!
However, a lot of this audio stuff and what people consider to be a
'good' sound is pretty subjective.
See "valve amps":-)

In the case of valve amps ppl like the colouration they add.

Graham

[Dirk Bruere at NeoPax]

Eeyore wrote:

Dirk Bruere at NeoPax wrote:

On 5 Apr, 15:48, "Arny Krueger" wrote:
"Dirk Bruere at NeoPax" wrote in
oglegroups.com

The Auzentech card is shipped with socketed dual opamps
on the outputs, S4580P
The spec seems pretty average and subjectively a bit of
top end seems missing compared to the X-Fi. Whether the
X-Fi had too much to start with is another question...
Anyway, suggested replacements, assuming replacements are
necessary at all?
The 5532 has better slew rate, although I'm not sure
about the noise because the data sheet for the S4580P
simply says: 0.8 uV RMS, RIAA Rs=2k2, f=30kHz LPF
Bandwidth = 15MHz
Have you run the Audio Rightmark program on the card to see what areas of
performance need improving?
No, due to the fact I've only just learned about it!
However, a lot of this audio stuff and what people consider to be a
'good' sound is pretty subjective.
See "valve amps":-)

In the case of valve amps ppl like the colouration they add.

Graham


I know.
I've played around with DSP progs feeding speakers and technically
measured flat response is not the 'best' sound even though it might be
the most accurate.

--
Dirk

http://www.onetribe.me.uk - the UK's only occult talk show
Hosted by Dirk Bruere and Marc Power

[Dirk Bruere at NeoPax]

Eeyore wrote:

Dirk Bruere at NeoPax wrote:

The Auzentech card is shipped with socketed dual opamps on the
outputs, S4580P
The spec seems pretty average

Not at all. 4580s are a lot better than average. Where did you get your data
from ?


and subjectively a bit of top end seems
missing compared to the X-Fi.

It certainly won't be due to the 4580 !


Whether the X-Fi had too much to start
with is another question...

Anyway, suggested replacements, assuming replacements are necessary at
all?
The 5532 has better slew rate, although I'm not sure about the noise
because the data sheet for the S4580P simply says:

Get the original JRC data for the NJM4580. You can see that the nV/sqrt Hz is
well below that of a 5532. It's ~ 3nV/sqrt Hz.


0.8 uV RMS, RIAA
Rs=2k2, f=30kHz LPF
Bandwidth = 15MHz

The 4580 is easily the equal of the 5532. Better in fact. Notably distortion.

What's the audible effects of slew rates?
From a naive EE POV I would assume that (for a 1V o/p signal) anything
better than 1 volt per 25 uS would be inaudible since that corresponds
with a 40kHz response for a Nyquist 20kHz resolution

--
Dirk

http://www.onetribe.me.uk - the UK's only occult talk show
Hosted by Dirk Bruere and Marc Power

[Eeyore]

Dirk Bruere at NeoPax wrote:

What's the audible effects of slew rates?

With the numbers in question for audio ? NONE, zilch, nix.


From a naive EE POV I would assume that (for a 1V o/p signal) anything
better than 1 volt per 25 uS would be inaudible since that corresponds
with a 40kHz response for a Nyquist 20kHz resolution

As I'm sure you know, the peak dV/dt of a sinwave is 2.pi.f.Vpeak. I hope I
remembered that right.

so for 20kHz @ 25V pk-pk (typically clipping level) its .... 1.6 V/us.

Graham

[Eeyore]

Dirk Bruere at NeoPax wrote:

Eeyore wrote:

In the case of valve amps ppl like the colouration they add.

I know.
I've played around with DSP progs feeding speakers and technically
measured flat response is not the 'best' sound even though it might be
the most accurate.

That's simply an opinion of course. ;~)

Find some truly flat speakers first !

Graham

[Arny Krueger]

"Dirk Bruere at NeoPax" wrote in
message

I've played around with DSP progs feeding speakers and
technically measured flat response is not the 'best'
sound even though it might be the most accurate.

When you measure the response of a speaker, you are not measuring the
*total* response of the speaker, you are just measuring the sound from the
speaker that happens to fall on the microphone, weighted by the directional
pattern and response curve of the mic.

For example, if you measure a speaker in an anechoic chamber with a good
instrumentation mic, all you are measuing is the on-axis reponse of the
speaker. But, in a regular room, you hear a mixture of the on-axis response
of the speaker weighted by your hearing, plus all of the off-axis responses
of the speaker as modified by reflection around the room and back to your
ears. It is very hard to measure a speaker and weight its response just like
your ears do.


Contrast that with a wire. A wire has just one input and one output. It's
relatively easy to measure the loss in a wire because of that. It if far
easier to figure out what a wire should do if it is nearly perfect.

[Dirk Bruere at NeoPax]

Eeyore wrote:

Dirk Bruere at NeoPax wrote:

What's the audible effects of slew rates?

With the numbers in question for audio ? NONE, zilch, nix.


From a naive EE POV I would assume that (for a 1V o/p signal) anything
better than 1 volt per 25 uS would be inaudible since that corresponds
with a 40kHz response for a Nyquist 20kHz resolution

As I'm sure you know, the peak dV/dt of a sinwave is 2.pi.f.Vpeak. I hope I
remembered that right.

so for 20kHz @ 25V pk-pk (typically clipping level) its .... 1.6 V/us.

Graham

Which is roughly what I stated for 1V p-p (chosen to make the sums
simpler) give or take an omega

--
Dirk

http://www.onetribe.me.uk - The UK's only occult talk show
Presented by Dirk Bruere and Marc Power on ResonanceFM 104.4
http://www.resonancefm.com

[Mr.T]

"Dirk Bruere at NeoPax" wrote in message
...
I've played around with DSP progs feeding speakers and technically
measured flat response is not the 'best' sound even though it might be
the most accurate.

Naturally, for YOU. It depends on YOUR auditory system and your listening
experiences and desires.

Many people do prefer accurate though, and simply adjust their perception to
suit.

MrT.

[Mr.T]

"Arny Krueger" wrote in message
...
When you measure the response of a speaker, you are not measuring the
*total* response of the speaker, you are just measuring the sound from the
speaker that happens to fall on the microphone, weighted by the
directional
pattern and response curve of the mic.

Surely you realise that measurement microphones are usually both
omni-directional and are flat to the required degree, and/or have a known
calibration curve.
(In fact I know you do, so am puzzled why you wrote that?)

For example, if you measure a speaker in an anechoic chamber with a good
instrumentation mic, all you are measuing is the on-axis reponse of the
speaker. But, in a regular room, you hear a mixture of the on-axis
response
of the speaker weighted by your hearing, plus all of the off-axis
responses
of the speaker as modified by reflection around the room and back to your
ears. It is very hard to measure a speaker and weight its response just
like
your ears do.

And very little point anyway since it changes both within the rooom, and
from room to room. And your auditory system partly discriminates directional
and time delayed signals anyway, thus making any simple integration further
pointless IMO.

MrT.

[Eeyore]

Dirk Bruere at NeoPax wrote:

Eeyore wrote:
Dirk Bruere at NeoPax wrote:

What's the audible effects of slew rates?

With the numbers in question for audio ? NONE, zilch, nix.


From a naive EE POV I would assume that (for a 1V o/p signal) anything
better than 1 volt per 25 uS would be inaudible since that corresponds
with a 40kHz response for a Nyquist 20kHz resolution

As I'm sure you know, the peak dV/dt of a sinwave is 2.pi.f.Vpeak. I hope I
remembered that right.

so for 20kHz @ 25V pk-pk (typically clipping level) its .... 1.6 V/us.


Which is roughly what I stated for 1V p-p (chosen to make the sums
simpler) give or take an omega.

Do bear in mind that in order to avoid slew induced distortion, it's wise to stay
well clear of the limiting slew rate too.

Graham

[Dirk Bruere at NeoPax]

Eeyore wrote:

Dirk Bruere at NeoPax wrote:

Eeyore wrote:
Dirk Bruere at NeoPax wrote:

What's the audible effects of slew rates?
With the numbers in question for audio ? NONE, zilch, nix.


From a naive EE POV I would assume that (for a 1V o/p signal) anything
better than 1 volt per 25 uS would be inaudible since that corresponds
with a 40kHz response for a Nyquist 20kHz resolution
As I'm sure you know, the peak dV/dt of a sinwave is 2.pi.f.Vpeak. I hope I
remembered that right.

so for 20kHz @ 25V pk-pk (typically clipping level) its .... 1.6 V/us.

Which is roughly what I stated for 1V p-p (chosen to make the sums
simpler) give or take an omega.

Do bear in mind that in order to avoid slew induced distortion, it's wise to stay
well clear of the limiting slew rate too.

Graham

Yes.
However, I suppose my original question was whether there is any audible
effects when the slew rate limit is well into the ultrasonics. But maybe
that has more to do with ears than electronics.

--
Dirk

http://www.onetribe.me.uk - The UK's only occult talk show
Presented by Dirk Bruere and Marc Power on ResonanceFM 104.4
http://www.resonancefm.com

[Don Pearce]

On Tue, 10 Apr 2007 12:54:07 +0100, Dirk Bruere at NeoPax
wrote:

Eeyore wrote:

Dirk Bruere at NeoPax wrote:

Eeyore wrote:
Dirk Bruere at NeoPax wrote:

What's the audible effects of slew rates?
With the numbers in question for audio ? NONE, zilch, nix.


From a naive EE POV I would assume that (for a 1V o/p signal) anything
better than 1 volt per 25 uS would be inaudible since that corresponds
with a 40kHz response for a Nyquist 20kHz resolution
As I'm sure you know, the peak dV/dt of a sinwave is 2.pi.f.Vpeak. I hope I
remembered that right.

so for 20kHz @ 25V pk-pk (typically clipping level) its .... 1.6 V/us.

Which is roughly what I stated for 1V p-p (chosen to make the sums
simpler) give or take an omega.

Do bear in mind that in order to avoid slew induced distortion, it's wise to stay
well clear of the limiting slew rate too.

Graham

Yes.
However, I suppose my original question was whether there is any audible
effects when the slew rate limit is well into the ultrasonics. But maybe
that has more to do with ears than electronics.

What matters is not whether slew rate limiting happens in the audible
frequency range, but whether it happens at all. If there are
ultrasonic signals big enough to cause slew rate limiting, there will
be massive intermod effects on audible signals. Any amplifier should
always roll its HF response off at a rate which guarantees that no
signal will cause limiting.

d

--
Pearce Consulting
http://www.pearce.uk.com

[John Phillips]

On 2007-04-10, Don Pearce wrote:
On Tue, 10 Apr 2007 12:54:07 +0100, Dirk Bruere at NeoPax
wrote:

However, I suppose my original question was whether there is any audible
effects when the slew rate limit is well into the ultrasonics. But maybe
that has more to do with ears than electronics.

What matters is not whether slew rate limiting happens in the audible
frequency range, but whether it happens at all. If there are
ultrasonic signals big enough to cause slew rate limiting, there will
be massive intermod effects on audible signals. Any amplifier should
always roll its HF response off at a rate which guarantees that no
signal will cause limiting.

Indeed that's exactly AIUI.

I always thought a properly-specified single RC pole at the input of an
amplifier was standard practice and was sufficient to stop the amplifier
hitting its slew rate limits for any signal that would not cause clipping.

I was puzzled many years ago to see lots of hot air in the audio press
about slew-rate-induced distortion. Of course I was merely a student
of semiconductor device physics at the time so what did I know of
designing amplifiers. However it seemed that this was a non-problem
then and remains a non-problem now. If I am wrong then I am sure
someone will enlighten me.

Perhaps good design practice has to be re-discovered from time to time.

--
John Phillips

[Arny Krueger]

"Mr.T" MrT@home wrote in message

"Arny Krueger" wrote in message
...

"Dirk Bruere at NeoPax" wrote in
message

I've played around with DSP progs feeding speakers and
technically measured flat response is not the 'best'
sound even though it might be the most accurate.


When you measure the response of a speaker, you are not
measuring the *total* response of the speaker, you are
just measuring the sound from the speaker that happens
to fall on the microphone, weighted by the directional
pattern and response curve of the mic.

Surely you realise that measurement microphones are
usually both omni-directional and are flat to the
required degree, and/or have a known calibration curve.
(In fact I know you do, so am puzzled why you wrote that?)

Key words: "sound from the speaker that happens to fall on the microphone".

For example, if you measure a speaker in an anechoic
chamber with a good instrumentation mic, all you are
measuring is the on-axis reponse of the speaker. But, in
a regular room, you hear a mixture of the on-axis
response of the speaker weighted by your hearing, plus
all of the off-axis responses of the speaker as modified
by reflection around the room and back to your ears. It
is very hard to measure a speaker and weight its
response just like your ears do.

And very little point anyway since it changes both within
the rooom, and from room to room.

Please see the post I was responding to. The author seemed to be a bit
mystefied about why setting up a speaker for flat measured response did not
always correspond to best sounding response. The point of my post is that
with speakers, what we measure is not usually the same as what we hear.


And your auditory
system partly discriminates directional and time delayed
signals anyway, thus making any simple integration
further pointless IMO.

MrT.

[Arny Krueger]

"John Phillips" wrote
in message

I always thought a properly-specified single RC pole at
the input of an amplifier was standard practice and was
sufficient to stop the amplifier hitting its slew rate
limits for any signal that would not cause clipping.

Hold that thought! ;-)

I was puzzled many years ago to see lots of hot air in
the audio press about slew-rate-induced distortion.

Doing their part to increase global warming, methinks!

Of course I was merely a student
of semiconductor device physics at the time so what did I
know of designing amplifiers. However it seemed that
this was a non-problem
then and remains a non-problem now. If I am wrong then I
am sure someone will enlighten me.

The signals coming into the line input of an audio interface should be
pretty well tamed-down by the preceeding stages of amplification and
processing.

For example, a mic preamp seems to be far more likely to be encountering
hard-to-follow signals.

Interestingly enough, one of the places where highly-regarded mic preamps
distinguish themselves from more ordinary products is their resistance to IM
from broadband signals at their inputs.

Perhaps good design practice has to be re-discovered from
time to time.

[GregS]

In article , wrote:
On 2007-04-10, Don Pearce wrote:
On Tue, 10 Apr 2007 12:54:07 +0100, Dirk Bruere at NeoPax
wrote:

However, I suppose my original question was whether there is any audible
effects when the slew rate limit is well into the ultrasonics. But maybe
that has more to do with ears than electronics.

What matters is not whether slew rate limiting happens in the audible
frequency range, but whether it happens at all. If there are
ultrasonic signals big enough to cause slew rate limiting, there will
be massive intermod effects on audible signals. Any amplifier should
always roll its HF response off at a rate which guarantees that no
signal will cause limiting.

Indeed that's exactly AIUI.

I always thought a properly-specified single RC pole at the input of an
amplifier was standard practice and was sufficient to stop the amplifier
hitting its slew rate limits for any signal that would not cause clipping.

I don't think it was standard practice, but probably used more after the TIM thingy.
Many good amplifiers do incorporate RF filtering.

greg

I was puzzled many years ago to see lots of hot air in the audio press
about slew-rate-induced distortion. Of course I was merely a student
of semiconductor device physics at the time so what did I know of
designing amplifiers. However it seemed that this was a non-problem
then and remains a non-problem now. If I am wrong then I am sure
someone will enlighten me.

Perhaps good design practice has to be re-discovered from time to time.

[Don Pearce]

On Tue, 10 Apr 2007 10:39:11 -0400, "Arny Krueger"
wrote:

Interestingly enough, one of the places where highly-regarded mic preamps
distinguish themselves from more ordinary products is their resistance to IM
from broadband signals at their inputs.

I have a GSM phone sitting right next to my Behringer mixer right now.
Not a hint of induced noise.

d

--
Pearce Consulting
http://www.pearce.uk.com

[John Phillips]

On 2007-04-10, Arny Krueger wrote:
"John Phillips" wrote
in message

I always thought a properly-specified single RC pole at
the input of an amplifier was standard practice and was
sufficient to stop the amplifier hitting its slew rate
limits for any signal that would not cause clipping.

Hold that thought! ;-)

I was puzzled many years ago to see lots of hot air in
the audio press about slew-rate-induced distortion.

Doing their part to increase global warming, methinks!

Of course I was merely a student
of semiconductor device physics at the time so what did I
know of designing amplifiers. However it seemed that
this was a non-problem
then and remains a non-problem now. If I am wrong then I
am sure someone will enlighten me.

The signals coming into the line input of an audio interface should be
pretty well tamed-down by the preceeding stages of amplification and
processing.

I am sure that's generally true, but as a designer I am equally sure
I could not assume its truth in all circumstances. The only safe
approach seems to be that of assuming a very wide bandwidth input and
making sure the amplifier still behaves.

For example, a mic preamp seems to be far more likely to be encountering
hard-to-follow signals.

Another place to make no unwarranted assumptions.

Interestingly enough, one of the places where highly-regarded mic preamps
distinguish themselves from more ordinary products is their resistance to IM
from broadband signals at their inputs.

A result of good design for the purpose. Assumptions seem to be the
root of many a cock-up.

--
John Phillips

[Arny Krueger]

"Don Pearce" wrote in message

On Tue, 10 Apr 2007 10:39:11 -0400, "Arny Krueger"
wrote:

Interestingly enough, one of the places where
highly-regarded mic preamps distinguish themselves from
more ordinary products is their resistance to IM from
broadband signals at their inputs.

I have a GSM phone sitting right next to my Behringer
mixer right now. Not a hint of induced noise.

Sometimes higher frequencies create less fuss.

I notice that keying a FRS walkie-talkie in the sanctuary causes a click in
the SR system at church, but no audio.

[Don Pearce]

On Tue, 10 Apr 2007 12:11:27 -0400, "Arny Krueger"
wrote:

"Don Pearce" wrote in message

On Tue, 10 Apr 2007 10:39:11 -0400, "Arny Krueger"
wrote:

Interestingly enough, one of the places where
highly-regarded mic preamps distinguish themselves from
more ordinary products is their resistance to IM from
broadband signals at their inputs.

I have a GSM phone sitting right next to my Behringer
mixer right now. Not a hint of induced noise.

Sometimes higher frequencies create less fuss.

I notice that keying a FRS walkie-talkie in the sanctuary causes a click in
the SR system at church, but no audio.


Yebbut GSM phones seem to have supernatural powers of penetration.
There are plenty of professional installations that are as susceptible
as hell to them.

d

--
Pearce Consulting
http://www.pearce.uk.com

[GregS]

In article , (Don Pearce) wrote:
On Tue, 10 Apr 2007 12:11:27 -0400, "Arny Krueger"
wrote:

"Don Pearce" wrote in message

On Tue, 10 Apr 2007 10:39:11 -0400, "Arny Krueger"
wrote:

Interestingly enough, one of the places where
highly-regarded mic preamps distinguish themselves from
more ordinary products is their resistance to IM from
broadband signals at their inputs.

I have a GSM phone sitting right next to my Behringer
mixer right now. Not a hint of induced noise.

Sometimes higher frequencies create less fuss.

I notice that keying a FRS walkie-talkie in the sanctuary causes a click in
the SR system at church, but no audio.


Yebbut GSM phones seem to have supernatural powers of penetration.
There are plenty of professional installations that are as susceptible
as hell to them.

FM carriers usually just creates an on/off thump. Digital modulation will
usually create a lot of havoc. I would imaging the carrier to be constantly on,
not sure. GSM goes up to almost 2gHz, penetrating cabinets and holes much
easier. While FRS is about 490mHz. I tried holding an FM transmitter close
to some old NASA Collins digital equipment one time. The carrier alone did nothing, but
when I talked loudly into the microphone, the equipment glitched and went off sync.
The test worked, as I guessed.

greg

[Peter Larsen]

Dirk Bruere at NeoPax wrote:

However, I suppose my original question was whether there is any
audible effects when the slew rate limit is well into the
ultrasonics.

Yes, I encountered one opamp sounding plain boring when under RF stress.

Dirk


Kind regards

Peter Larsen

[Peter Larsen]

Eeyore wrote:

Dirk Bruere at NeoPax wrote:

The Auzentech card is shipped with socketed dual opamps on the
outputs, S4580P
The spec seems pretty average

Not at all. 4580s are a lot better than average.

They don't do really bad things. I did find a circuit using them boring,
and vastly improved after an opamp replacement.

It certainly won't be due to the 4580 !

On the probabilities here we disagree.

The 4580 is easily the equal of the 5532.

It could go either way, I have heard thingies with 553x being harsh or
being forward. I do not know why, but I recall that 553x comes in many
brands, makes and flavors.

Better in fact. Notably
distortion.

You may recall my modded DAT, Preben Friis's measurements suggested that
distortion was lower before the opamp replacement, much to my surprise
since it got so much cleaner by being modded, but it must be included
that coupling cap issues also were addressed.

Graham


Kind regards

Peter Larsen

[Peter Larsen]

Dirk Bruere at NeoPax wrote:

The Auzentech card is shipped with socketed dual opamps on the
outputs, S4580P

socketed, that is nice, complete url to the manufacturer?

The spec seems pretty average and subjectively a bit of top
end seems missing compared to the X-Fi.

Hmmm ....

Whether the X-Fi had too much to start
with is another question...

I have not heard that specific product.

Anyway, suggested replacements, assuming replacements
are necessary at all?

What is the context of the sound card?

The 5532 has better slew rate, although I'm not sure
about the noise because the data sheet for the S4580P
simply says: 0.8 uV RMS, RIAA
Rs=2k2, f=30kHz LPF
Bandwidth = 15MHz

Analog Devices OPA(?) 275 comes to mind as very well sounding and very
well behaved. Burr Brown 2134 has a charming midrange and is also a nice
upgrading tool to work with, but slmewhat more prone to turn-on(/turn
off thumps. Burr Brown X604 seems to have superior extremes of the audio
range, but can come across as bright.

NOTE however that effects of unbiased couplling caps can cloud the issue
and that very much audio has passed fantazillions of "industry standard"
components prior to reaching your playback system. Which is to say that
modding may be irrelevant, I would actively discourage it as an
unsuitable risk vs. possible benefit if not for the socketing of the
original IC's

Industry standard components DO tend to be those that pass approval in
tech deparment AND in finance department, the latter departments tend to
do their dollar math in 0.000000000001 cent increments.

All evaluations above are subjective and in one context where distortion
measurement was made the meausured distortion of the modded apparatus
increased by modding rather than decreasing, it also sounded very much
better, but specs were against the mod and showed that the original
component choice was very sane.

Dirk


Kind regards

Peter Larsen

[Eeyore]

Peter Larsen wrote:

Eeyore wrote:
Dirk Bruere at NeoPax wrote:

The Auzentech card is shipped with socketed dual opamps on the
outputs, S4580P
The spec seems pretty average

Not at all. 4580s are a lot better than average.

They don't do really bad things. I did find a circuit using them boring,
and vastly improved after an opamp replacement.

Using what kind of replacement ?


It certainly won't be due to the 4580 !

On the probabilities here we disagree.

Tell me more.


The 4580 is easily the equal of the 5532.

It could go either way, I have heard thingies with 553x being harsh or
being forward. I do not know why, but I recall that 553x comes in many
brands, makes and flavors.

That is true. I recall such comments being made in the late 70s early 80s. The
chip design may have altered ( only Philips/Signetics made it then and Philips
no longer make it btw ) as may the circuits it's used in.


Better in fact. Notably distortion.

You may recall my modded DAT,

I'm not familiar with that actually.


Preben Friis's measurements suggested that
distortion was lower before the opamp replacement, much to my surprise
since it got so much cleaner by being modded, but it must be included
that coupling cap issues also were addressed.

The coupling cap. Was that a small value zero bias electrolytic ? That makes
some sense.

Graham

[Peter Larsen]

Arny Krueger wrote:

[Grahamps question]

The coupling cap. Was that a small value zero bias
electrolytic ? That makes some sense.

The SV3800 uses _very_ large coupling caps that by design are
unpolarized. BB 2604's were used as replacement opmaps leading to
subjectively improved audio detail but also to a measureable increase in
distortion as detected by Preben Friis. We made a couple of comparative
recordings, I'll probably put them on my site if I care to re-do it some
day.

If its the mod that Peter posted info about here a few
years back,

It is about the SV3800 mod, where Preben Friis kindly provided his own
SV3800 for comparison. The SV3800 PCB is not of a quality that is
intended for repeat soldering, and traces run under the glue that keeps
the opamps in place during solder. I'd rather advice against the mod due
to the risk rather than reocmmend it for the obtained vastly improved
audio quality.

one salient feature was a rather dramatic change in FR below 100 Hz.

That was _intentional_ since the machine was used for live location
recording, the caps that could not be simply omitted where chosen as
small as possible to avoid overly large lumps of polypropylene dangling
inside a machine that would be frequently transported.

Input and output coupling caps were bypassed rather than replaced. I
don't see the point of having the preamp end with a coupling cap and the
DAT begin with a new, one must be enough. Whether that is a suitable
general practice for everybody is entirely a different issue, in this
context of a minimalistic recording setup every omitted component helps.

It is still in its modded state a very good machine - the equal of the
Fostex FR2, albeit differently sounding, but DAT tapes tend to be less
very good than they were, or the transport is flawed or both, and I got
tired of losing recordings due to tape debris (from new tapes!) on the
heads and now I use a less good sounding harddisk recorder that brings
the recording home every time.


Kind regards

Peter Larsen

[Arny Krueger]

"Eeyore" wrote in
message
Peter Larsen wrote:

Eeyore wrote:
Dirk Bruere at NeoPax wrote:

The Auzentech card is shipped with socketed dual
opamps on the outputs, S4580P
The spec seems pretty average

Not at all. 4580s are a lot better than average.

They don't do really bad things. I did find a circuit
using them boring, and vastly improved after an opamp
replacement.

Using what kind of replacement ?


It certainly won't be due to the 4580 !

On the probabilities here we disagree.

Tell me more.


The 4580 is easily the equal of the 5532.

It could go either way, I have heard thingies with 553x
being harsh or being forward. I do not know why, but I
recall that 553x comes in many brands, makes and flavors.

That is true. I recall such comments being made in the
late 70s early 80s. The chip design may have altered (
only Philips/Signetics made it then and Philips no longer
make it btw ) as may the circuits it's used in.


Better in fact. Notably distortion.

You may recall my modded DAT,

I'm not familiar with that actually.


Preben Friis's measurements suggested that
distortion was lower before the opamp replacement, much
to my surprise since it got so much cleaner by being
modded, but it must be included that coupling cap issues
also were addressed.

The coupling cap. Was that a small value zero bias
electrolytic ? That makes some sense.

If its the mod that Peter posted info about here a few years back, one
salient feature was a rather dramatic change in FR below 100 Hz.

[Richard Crowley]

"Peter Larsen" wrote...
Dirk Bruere at NeoPax wrote:

The Auzentech card is shipped with socketed dual opamps on the
outputs, S4580P

socketed, that is nice, complete url to the manufacturer?

www.auzentech.com

[Eeyore]

Arny Krueger wrote:

"Eeyore" wrote

The coupling cap. Was that a small value zero bias
electrolytic ? That makes some sense.

If its the mod that Peter posted info about here a few years back, one
salient feature was a rather dramatic change in FR below 100 Hz.

In *frequency response* !

Do you have any details of this ?

Graham-

[Peter Larsen]

Eeyore wrote:

Arny Krueger wrote:

"Eeyore" wrote

The coupling cap. Was that a small value zero bias
electrolytic ? That makes some sense.

If its the mod that Peter posted info about here a few years back, one
salient feature was a rather dramatic change in FR below 100 Hz.

In *frequency response* !

Yes.

Do you have any details of this ?

I can re-measure it some day, it was the *AIM* of the coupling cap mod
part of it to raise the -3 dB point to the vicinity of 20 Hz, and it was
not a flaw of the total mod that an approximation to that spec was
obtained, nor did that outcome have anything to do with the opamp swap.
The shape of the roll-off did get kinda strange because of the peculiar
circuit around the record level pots ...

I just don't care much for recording ventilation noise, train subsonics
and the strange rumble caused by people walking in the parts of a
building that are adjacent to the concert hall in the New Carlsberg
Glyptotek.

Graham-


Kind regards

Peter Larsen

[Arny Krueger]

"Peter Larsen" wrote in
message
Eeyore wrote:

Arny Krueger wrote:

"Eeyore" wrote

The coupling cap. Was that a small value zero bias
electrolytic ? That makes some sense.

If its the mod that Peter posted info about here a few
years back, one salient feature was a rather dramatic
change in FR below 100 Hz.

In *frequency response* !

Yes.

Do you have any details of this ?

I can re-measure it some day, it was the *AIM* of the
coupling cap mod part of it to raise the -3 dB point to
the vicinity of 20 Hz, and it was not a flaw of the total
mod that an approximation to that spec was obtained, nor
did that outcome have anything to do with the opamp swap.
The shape of the roll-off did get kinda strange because
of the peculiar circuit around the record level pots ...

I just don't care much for recording ventilation noise,
train subsonics and the strange rumble caused by people
walking in the parts of a building that are adjacent to
the concert hall in the New Carlsberg Glyptotek.

I don't fault the strategy of limiting LF response to improve SQ of live
recordings. I do it quite often.

The usual consquences of limiting LF response within reason and good taste
are a cleaner-sounding recording, particularly when reproduced on systems
with limited dynamic range at very low frequencies, which includes the vast
majority of consumer music listening systems.

[Mr.T]

"Peter Larsen" wrote in message
...
Not at all. 4580s are a lot better than average.

They don't do really bad things. I did find a circuit using them boring,
and vastly improved after an opamp replacement.

"Boring" a real technical term there :-)

Maybe if you told us exactly what TECHNICAL parameter was "vastly improved"
by the replacement?
You DID do proper measurements to determine that you weren't just imagining
things, right?

MrT.

[Peter Larsen]

"Mr.T" wrote:

"Peter Larsen" wrote in message
...

Not at all. 4580s are a lot better than average.

They don't do really bad things. I did find a circuit using
them boring, and vastly improved after an opamp replacement.

"Boring" a real technical term there :-)

Yes.

Maybe if you told us exactly what TECHNICAL parameter was
"vastly improved" by the replacement?

Detail, spatial definition and rendition, perspective - in short: stereo
imaging - and intertransient silence.

You DID do proper measurements to determine that you weren't
just imagining things, right?

I did not, and I specifically mentioned that measured distortion
performance of the thingamajic in question was worse after the mod. If
you want to think that it did not in fact sound better after the mod,
then think so and be happy. I am happy with the many good recordings I
got from it after the mod, and we can then both be happy and believe
that each other is a fool as per your implicit suggestion.

MrT.


Kind regards

Peter Larsen

[Peter Larsen]

Arny Krueger wrote:

The usual consquences of limiting LF response within reason
and good taste are a cleaner-sounding recording,

Actually not, things like voice and instrument box sounds suffer, it is
bad for guitars and the entire violoncel family as well as for concert
grands and for perceived 3-dimensinality

particularly when reproduced on systems
with limited dynamic range at very low frequencies, which
includes the vast majority of consumer music listening
systems.

But the outcome is that playback systems that do not have an extended lf
response sound cleaner and also rooms with a system with extended LF
response in them but with LF resonance issues sound a lot cleaner, and
for these reasons I generally will use high pass filtering in post,
often it appears to me to be the optimum trade-off in terms of midrange
clarity.


Kind regards

Peter Larsen