[Dirk Bruere at NeoPax]

I've been looking inside one of these at what components are used.
They seem overall to be pretty good eg AKM converters and 4580 opamps.
What I can't reconcile is why the DCX is so (comparatively) noisy
given the specs of the chips. There are a couple of possibilites that
I can see:
a) It uses a switching PSU for some reason instead of a linear
b) The signals after the DACs but before the opamp o/p are carried on
an unshielded ribbon cable

Any other ideas?
Am I being unreasonable?

Dirk

[Don Pearce]

On Fri, 28 Sep 2007 09:06:49 -0700, Dirk Bruere at NeoPax
wrote:

I've been looking inside one of these at what components are used.
They seem overall to be pretty good eg AKM converters and 4580 opamps.
What I can't reconcile is why the DCX is so (comparatively) noisy
given the specs of the chips. There are a couple of possibilites that
I can see:
a) It uses a switching PSU for some reason instead of a linear
b) The signals after the DACs but before the opamp o/p are carried on
an unshielded ribbon cable

Any other ideas?
Am I being unreasonable?

Dirk

What's the nature of the noise? Can you post a sample?d

d

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

[Dirk Bruere at NeoPax]

On 28 Sep, 17:11, (Don Pearce) wrote:
On Fri, 28 Sep 2007 09:06:49 -0700, Dirk Bruere at NeoPax

wrote:
I've been looking inside one of these at what components are used.
They seem overall to be pretty good eg AKM converters and 4580 opamps.
What I can't reconcile is why the DCX is so (comparatively) noisy
given the specs of the chips. There are a couple of possibilites that
I can see:
a) It uses a switching PSU for some reason instead of a linear
b) The signals after the DACs but before the opamp o/p are carried on
an unshielded ribbon cable

Any other ideas?
Am I being unreasonable?

Dirk

What's the nature of the noise? Can you post a sample?d

d

--
Pearce Consultinghttp://www.pearce.uk.com

Just that when it's connected up to a 1kW amp and speaker system there
is quite an audible hiss with no signal (or small signal) applied.
High quiescent noise. With something like it's (far more expensive)
XTA equivalent this noise does not exist. It's maybe 12dB better at
least. Yet the XTA appears to use no better components IIRC.

Dirk

[Don Pearce]

On Fri, 28 Sep 2007 09:15:54 -0700, Dirk Bruere at NeoPax
wrote:

On 28 Sep, 17:11, (Don Pearce) wrote:
On Fri, 28 Sep 2007 09:06:49 -0700, Dirk Bruere at NeoPax

wrote:
I've been looking inside one of these at what components are used.
They seem overall to be pretty good eg AKM converters and 4580 opamps.
What I can't reconcile is why the DCX is so (comparatively) noisy
given the specs of the chips. There are a couple of possibilites that
I can see:
a) It uses a switching PSU for some reason instead of a linear
b) The signals after the DACs but before the opamp o/p are carried on
an unshielded ribbon cable

Any other ideas?
Am I being unreasonable?

Dirk

What's the nature of the noise? Can you post a sample?d

d

--
Pearce Consultinghttp://www.pearce.uk.com

Just that when it's connected up to a 1kW amp and speaker system there
is quite an audible hiss with no signal (or small signal) applied.
High quiescent noise. With something like it's (far more expensive)
XTA equivalent this noise does not exist. It's maybe 12dB better at
least. Yet the XTA appears to use no better components IIRC.

Dirk

How convinced are you that you are comparing like with like? I mean,
was the gain structure identical between the two setups?

d

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

[Dirk Bruere at NeoPax]

On 28 Sep, 17:18, (Don Pearce) wrote:
On Fri, 28 Sep 2007 09:15:54 -0700, Dirk Bruere at NeoPax



wrote:
On 28 Sep, 17:11, (Don Pearce) wrote:
On Fri, 28 Sep 2007 09:06:49 -0700, Dirk Bruere at NeoPax

wrote:
I've been looking inside one of these at what components are used.
They seem overall to be pretty good eg AKM converters and 4580 opamps.
What I can't reconcile is why the DCX is so (comparatively) noisy
given the specs of the chips. There are a couple of possibilites that
I can see:
a) It uses a switching PSU for some reason instead of a linear
b) The signals after the DACs but before the opamp o/p are carried on
an unshielded ribbon cable

Any other ideas?
Am I being unreasonable?

Dirk

What's the nature of the noise? Can you post a sample?d

d

--
Pearce Consultinghttp://www.pearce.uk.com

Just that when it's connected up to a 1kW amp and speaker system there
is quite an audible hiss with no signal (or small signal) applied.
High quiescent noise. With something like it's (far more expensive)
XTA equivalent this noise does not exist. It's maybe 12dB better at
least. Yet the XTA appears to use no better components IIRC.

Dirk

How convinced are you that you are comparing like with like? I mean,
was the gain structure identical between the two setups?

d

--
Pearce Consultinghttp://www.pearce.uk.com

Yes.
Of course, there's a 10x factor in the cost but I would like to know
*why* there is such a noise difference.
I would also like to know why Behringer uses a switching PSU in its
unit when it would appear to be unnecessary.

Dirk

[Don Pearce]

On Fri, 28 Sep 2007 09:28:09 -0700, Dirk Bruere at NeoPax
wrote:

On 28 Sep, 17:18, (Don Pearce) wrote:
On Fri, 28 Sep 2007 09:15:54 -0700, Dirk Bruere at NeoPax



wrote:
On 28 Sep, 17:11, (Don Pearce) wrote:
On Fri, 28 Sep 2007 09:06:49 -0700, Dirk Bruere at NeoPax

wrote:
I've been looking inside one of these at what components are used.
They seem overall to be pretty good eg AKM converters and 4580 opamps.
What I can't reconcile is why the DCX is so (comparatively) noisy
given the specs of the chips. There are a couple of possibilites that
I can see:
a) It uses a switching PSU for some reason instead of a linear
b) The signals after the DACs but before the opamp o/p are carried on
an unshielded ribbon cable

Any other ideas?
Am I being unreasonable?

Dirk

What's the nature of the noise? Can you post a sample?d

d

--
Pearce Consultinghttp://www.pearce.uk.com

Just that when it's connected up to a 1kW amp and speaker system there
is quite an audible hiss with no signal (or small signal) applied.
High quiescent noise. With something like it's (far more expensive)
XTA equivalent this noise does not exist. It's maybe 12dB better at
least. Yet the XTA appears to use no better components IIRC.

Dirk

How convinced are you that you are comparing like with like? I mean,
was the gain structure identical between the two setups?

d

--
Pearce Consultinghttp://www.pearce.uk.com

Yes.
Of course, there's a 10x factor in the cost but I would like to know
*why* there is such a noise difference.
I would also like to know why Behringer uses a switching PSU in its
unit when it would appear to be unnecessary.

Dirk

I'm just a little puzzled because I have a small Behringer mixer with
a switching supply. It exhibits no excess noise of any kind - just the
specified amount of white noise from the mic inputs.

If the power supply is at fault I would not expect to be hearing white
noise, but some kind of squeal or buzz.Can you make a measurement and
see if it is doing what it is supposed to, and maybe you have simply
underspecified what you need?

d

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

[Dirk Bruere at NeoPax]

On 28 Sep, 17:40, (Don Pearce) wrote:
On Fri, 28 Sep 2007 09:28:09 -0700, Dirk Bruere at NeoPax



wrote:
On 28 Sep, 17:18, (Don Pearce) wrote:
On Fri, 28 Sep 2007 09:15:54 -0700, Dirk Bruere at NeoPax

wrote:
On 28 Sep, 17:11, (Don Pearce) wrote:
On Fri, 28 Sep 2007 09:06:49 -0700, Dirk Bruere at NeoPax

wrote:
I've been looking inside one of these at what components are used.
They seem overall to be pretty good eg AKM converters and 4580 opamps.
What I can't reconcile is why the DCX is so (comparatively) noisy
given the specs of the chips. There are a couple of possibilites that
I can see:
a) It uses a switching PSU for some reason instead of a linear
b) The signals after the DACs but before the opamp o/p are carried on
an unshielded ribbon cable

Any other ideas?
Am I being unreasonable?

Dirk

What's the nature of the noise? Can you post a sample?d

d

--
Pearce Consultinghttp://www.pearce.uk.com

Just that when it's connected up to a 1kW amp and speaker system there
is quite an audible hiss with no signal (or small signal) applied.
High quiescent noise. With something like it's (far more expensive)
XTA equivalent this noise does not exist. It's maybe 12dB better at
least. Yet the XTA appears to use no better components IIRC.

Dirk

How convinced are you that you are comparing like with like? I mean,
was the gain structure identical between the two setups?

d

--
Pearce Consultinghttp://www.pearce.uk.com

Yes.
Of course, there's a 10x factor in the cost but I would like to know
*why* there is such a noise difference.
I would also like to know why Behringer uses a switching PSU in its
unit when it would appear to be unnecessary.

Dirk

I'm just a little puzzled because I have a small Behringer mixer with
a switching supply. It exhibits no excess noise of any kind - just the
specified amount of white noise from the mic inputs.

If the power supply is at fault I would not expect to be hearing white
noise, but some kind of squeal or buzz.Can you make a measurement and
see if it is doing what it is supposed to, and maybe you have simply
underspecified what you need?

d

--
Pearce Consultinghttp://www.pearce.uk.com

I suppose my main question is why do the Behringer and XTA units that
use pretty much the same major components have such different noise
figures. In other words, what justifies that 10x price differential?

Dirk

[Ian Iveson]

Dirk said

Just that when it's connected up to a 1kW amp and speaker
system there
is quite an audible hiss with no signal (or small signal)
applied.
High quiescent noise. With something like it's (far more
expensive)
XTA equivalent this noise does not exist. It's maybe 12dB
better at
least. Yet the XTA appears to use no better components
IIRC.

Maybe it has more of them?

SMPSs produce a range of frequencies, but if you are talking
about random noise, is there any reason to expect it from a
switched mode supply?

Perhaps the more expensive machine has more extensive
noise-reduction? Even if they are identical but with
different power supplies, it could be that the stages are
better isolated or decoupled, including earthing
arrangements, in the quieter machine. Maybe the better one
uses several separately-regulated supplies, for example.

Or possibly noise from the source is filtered in one and not
the other.

Presumably it meets its published specifications.

Ian

[Arny Krueger]

"Dirk Bruere at NeoPax" wrote in
message
ups.com

I've been looking inside one of these at what components
are used. They seem overall to be pretty good eg AKM
converters and 4580 opamps. What I can't reconcile is why
the DCX is so (comparatively) noisy given the specs of
the chips. There are a couple of possibilites that I can
see:

a) It uses a switching PSU for some reason instead of a
linear

Switchmode power supplies are often chosen to obtain economical operation
with a variety of power line voltages and frequencies.

b) The signals after the DACs but before the opamp o/p
are carried on an unshielded ribbon cable

Very typical.

Any other ideas?

Gain structure?

Am I being unreasonable?

The methodology you are using to evaluate noise

"connected up to a 1kW amp and speaker system there
is quite an audible hiss with no signal (or small signal) applied."

is meaningless.

Noise measurements are relevant only he context of standard gain and
bandwidth.

It's my understanding that the DCX 2496 is partcularly susceptible to gain
structure abuse because it has the possibility of multiple gain-setting
elements in each signal path.

[Eeyore]

Dirk Bruere at NeoPax wrote:

I've been looking inside one of these at what components are used.
They seem overall to be pretty good eg AKM converters and 4580 opamps.
What I can't reconcile is why the DCX is so (comparatively) noisy
given the specs of the chips. There are a couple of possibilites that
I can see:
a) It uses a switching PSU for some reason instead of a linear
b) The signals after the DACs but before the opamp o/p are carried on
an unshielded ribbon cable

Neither of those would really give you *noise*.

There's far more to designing quiet circuitry than the chips alone.

Graham

[Eeyore]

Dirk Bruere at NeoPax wrote:

(Don Pearce) wrote:

What's the nature of the noise? Can you post a sample?d


Just that when it's connected up to a 1kW amp and speaker system there
is quite an audible hiss with no signal (or small signal) applied.

Not that surprising is it ?

With a 1kW amp you have 20dB more gain than with say a 10W amp.


High quiescent noise. With something like it's (far more expensive)
XTA equivalent this noise does not exist. It's maybe 12dB better at
least. Yet the XTA appears to use no better components IIRC.

How about measuring it ?

Graham

[Eeyore]

Dirk Bruere at NeoPax wrote:

I would also like to know why Behringer uses a switching PSU in its
unit when it would appear to be unnecessary.

Unnecessary in what way ?

I'll bet it's universal input and that saves cost and/or stockholding issues
(which also incur cost).

Graham

[Eeyore]

Dirk Bruere at NeoPax wrote:

I suppose my main question is why do the Behringer and XTA units that
use pretty much the same major components have such different noise
figures. In other words, what justifies that 10x price differential?

Simple. Its design skill.

Graham

[Eeyore]

Ian Iveson wrote:

Perhaps the more expensive machine has more extensive
noise-reduction? Even if they are identical but with
different power supplies, it could be that the stages are
better isolated or decoupled, including earthing
arrangements, in the quieter machine. Maybe the better one
uses several separately-regulated supplies, for example.

Or possibly noise from the source is filtered in one and not
the other.

LMAO !

Barking up all the wrong trees as usual !

Graham

[Eeyore]

"François Yves Le Gal" wrote:

Dirk Bruere at NeoPax wrote:

There are a couple of possibilites that
I can see:

c. Poorly set levels. the Behringer is very sensitive to level matching and
offers a narrow optimum window.

Since they both operate at exactly the same levels how can you come up with that
conclusion ?

The OP should ensure he hasn't selected any '-10' option on the Behringer unit
(some have that - not sure about this one).

Graham

[dezza]

These guys offer a power supply mod that claims decreased noise.
Perhaps they may know the answer to your questions.


http://www.selectronic.fr/dcx2496_us.asp#11


Dezza.

[Eeyore]

"François Yves Le Gal" wrote:

Eeyore wrote:

Since they both operate at exactly the same levels how can you come up with that
conclusion ?

I'm quite familair with the DCX quirks...

Can you elaborate in that case ?

Graham

[Eeyore]

dezza wrote:

These guys offer a power supply mod that claims decreased noise.
Perhaps they may know the answer to your questions.

http://www.selectronic.fr/dcx2496_us.asp#11

Mostly snake oil I'm sure.

Graham

[Ian Iveson]

Eeyore

LMAO !

Glad you are able to enjoy yourself in spite of your
inadequate personality.

Barking up all the wrong trees as usual !

Just trying to illustrate that there is more to a design
than a few component choices.

Any suggestions to contribute, Eeyore? The OP asked for
ideas, and you have done nothing but posture and mock.

What's a DCX2496 anyway?

Ian

[Don Pearce]

On Fri, 28 Sep 2007 17:35:43 GMT, "Ian Iveson"
wrote:

Dirk said

Just that when it's connected up to a 1kW amp and speaker
system there
is quite an audible hiss with no signal (or small signal)
applied.
High quiescent noise. With something like it's (far more
expensive)
XTA equivalent this noise does not exist. It's maybe 12dB
better at
least. Yet the XTA appears to use no better components
IIRC.

Maybe it has more of them?

SMPSs produce a range of frequencies, but if you are talking
about random noise, is there any reason to expect it from a
switched mode supply?

No. If a power supply causes noise, it won't be random. In any case a
good audio design is not sensitive to power supply noise.

Perhaps the more expensive machine has more extensive
noise-reduction? Even if they are identical but with
different power supplies, it could be that the stages are
better isolated or decoupled, including earthing
arrangements, in the quieter machine. Maybe the better one
uses several separately-regulated supplies, for example.

Noise reduction? That is a tool of last resort for a lousy recording -
it has nothing to do with this kind of product. Isolation and
decoupling have nothing to do with noise. Sure you can reduce noise by
decoupling, but you will reduce the wanted signal by the same amount.
As for earthing arrangements, that will affect hum, not noise. Again
the regulation of power supplies has nothing to do with this.

Have you actually given this a moment's thought, or are you just
writing technical-sounding phrases at random?

Or possibly noise from the source is filtered in one and not
the other.

Please stop this! Noise is by definition in the audio band. If you
filter the noise, you filter the signal too.

You are just writing garbage.

Presumably it meets its published specifications.

Ian

At last something sensible! The thing to do is measure. If it meets
its spec, tough - it was bought without sufficient thought (or it is
not being used correctly). If it doesn't, it can go back for a refund.

d


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

[Arny Krueger]

"Eeyore" wrote in
message
"François Yves Le Gal" wrote:

Eeyore wrote:

Since they both operate at exactly the same levels how
can you come up with that conclusion ?

I'm quite familair with the DCX quirks...

Can you elaborate in that case ?

The DCX is an interesting box because it is basically a collection of
software-based processing blocks that can be patched together by software to
create a wide variety of signal processing systems, many of which can have
gain either explicitly or implicitly. Due to the tiny display, it is easy
for people in the field to lose track of what they've actually programmed.

The inside of the DCX 2496 box is very spare, because there are as I
recally only about 4 circuit cards - one for the front panel, one for the
rear, one with the DSP and converters on it, and one for the switchmode
power supply. The switchmode power supply is in its own metal enclosure.
Other than routing analog and digital signals in and out of the DSP, there
is no dedicated signal-handling circuity.

By changing the programming of the DSP, the DXC 2496 could be just about any
kind of signal processor that can be created with a DSP.

[Arny Krueger]

"dezza" wrote in message
oups.com

These guys offer a power supply mod that claims decreased
noise.

Claims are one thing, actual performance is a big secret because they make
no specific claims.

Perhaps they may know the answer to your questions.

http://www.selectronic.fr/dcx2496_us.asp#11


Oh my!

I'm gonna be sick! :-(

[Eeyore]

Arny Krueger wrote:

"dezza" wrote in message

These guys offer a power supply mod that claims decreased
noise.

Claims are one thing, actual performance is a big secret because they make
no specific claims.

I was intruiged by their claim that a crystal oscillator isn't stable enough for
processing audio at high quality ! They must know something that the rest of the
industry doesn't. ;~)

And the Black Gate caps - LOL ! Just for the output too, nowhere else. What
charlatans.

Graham

[Eeyore]

"François Yves Le Gal" wrote:

Eeyore wrote:

Can you elaborate in that case ?

The DCX becomes quite noisy if you don't work as close as it's max input
level as possible. This results in a +22 dB output level which has to be
attenuated in most cases.

Using the digital input is another story.

That suggests the analogue inputs don't operate at a suitable level. There's no
way an input buffer is going to add any appreciable noise. An input buffer using
4580s wouldn't have more than -110dB of noise typically.

However, you would be able to tell that if the digital input resulted in a
larger output and I doubt it does.

Graham

[Dirk Bruere at NeoPax]

Ian Iveson wrote:
Dirk said

Just that when it's connected up to a 1kW amp and speaker
system there
is quite an audible hiss with no signal (or small signal)
applied.
High quiescent noise. With something like it's (far more
expensive)
XTA equivalent this noise does not exist. It's maybe 12dB
better at
least. Yet the XTA appears to use no better components
IIRC.

Maybe it has more of them?

SMPSs produce a range of frequencies, but if you are talking
about random noise, is there any reason to expect it from a
switched mode supply?

Perhaps the more expensive machine has more extensive
noise-reduction? Even if they are identical but with
different power supplies, it could be that the stages are
better isolated or decoupled, including earthing
arrangements, in the quieter machine. Maybe the better one
uses several separately-regulated supplies, for example.

Or possibly noise from the source is filtered in one and not
the other.

Presumably it meets its published specifications.

It could even be something like how the PCB is layed out and the number
of layers. The XTA board, for example is *much* bigger.

--
Dirk

http://www.transcendence.me.uk/ - Transcendence UK
Remote Viewing classes in London

[Dirk Bruere at NeoPax]

Arny Krueger wrote:
"Dirk Bruere at NeoPax" wrote in
message
ups.com

I've been looking inside one of these at what components
are used. They seem overall to be pretty good eg AKM
converters and 4580 opamps. What I can't reconcile is why
the DCX is so (comparatively) noisy given the specs of
the chips. There are a couple of possibilites that I can
see:

a) It uses a switching PSU for some reason instead of a
linear

Switchmode power supplies are often chosen to obtain economical operation
with a variety of power line voltages and frequencies.

b) The signals after the DACs but before the opamp o/p
are carried on an unshielded ribbon cable

Very typical.

Any other ideas?

Gain structure?

Am I being unreasonable?

The methodology you are using to evaluate noise

"connected up to a 1kW amp and speaker system there
is quite an audible hiss with no signal (or small signal) applied."

is meaningless.

Noise measurements are relevant only he context of standard gain and
bandwidth.

It's my understanding that the DCX 2496 is partcularly susceptible to gain
structure abuse because it has the possibility of multiple gain-setting
elements in each signal path.

True of the XTA as well.
The context you mention is identical settings (as far as possible) on
both units.

--
Dirk

http://www.transcendence.me.uk/ - Transcendence UK
Remote Viewing classes in London

[Dirk Bruere at NeoPax]

Eeyore wrote:

Dirk Bruere at NeoPax wrote:

(Don Pearce) wrote:
What's the nature of the noise? Can you post a sample?d

Just that when it's connected up to a 1kW amp and speaker system there
is quite an audible hiss with no signal (or small signal) applied.

Not that surprising is it ?

With a 1kW amp you have 20dB more gain than with say a 10W amp.


High quiescent noise. With something like it's (far more expensive)
XTA equivalent this noise does not exist. It's maybe 12dB better at
least. Yet the XTA appears to use no better components IIRC.

How about measuring it ?

Graham


Don't have the XTA around anymore.

--
Dirk

http://www.transcendence.me.uk/ - Transcendence UK
Remote Viewing classes in London

[Dirk Bruere at NeoPax]

Eeyore wrote:

dezza wrote:

These guys offer a power supply mod that claims decreased noise.
Perhaps they may know the answer to your questions.

http://www.selectronic.fr/dcx2496_us.asp#11

Mostly snake oil I'm sure.

Graham


I don't see their I/O unit being an improvement over the Behringer
The PSU OTOH might be of interest.

--
Dirk

http://www.transcendence.me.uk/ - Transcendence UK
Remote Viewing classes in London

[Dirk Bruere at NeoPax]

Eeyore wrote:

"François Yves Le Gal" wrote:

Eeyore wrote:

Can you elaborate in that case ?
The DCX becomes quite noisy if you don't work as close as it's max input
level as possible. This results in a +22 dB output level which has to be
attenuated in most cases.

Using the digital input is another story.

That suggests the analogue inputs don't operate at a suitable level. There's no
way an input buffer is going to add any appreciable noise. An input buffer using
4580s wouldn't have more than -110dB of noise typically.

However, you would be able to tell that if the digital input resulted in a
larger output and I doubt it does.

Graham


Using it with digital I/P reduces quiescent noise. Subjectively halving
it, which I suppose is what I might expect.

--
Dirk

http://www.transcendence.me.uk/ - Transcendence UK
Remote Viewing classes in London

[Ian Iveson]

Don Pearce wrote

Just that when it's connected up to a 1kW amp and
speaker
system there
is quite an audible hiss with no signal (or small
signal)
applied.
High quiescent noise. With something like it's (far more
expensive)
XTA equivalent this noise does not exist. It's maybe
12dB
better at
least. Yet the XTA appears to use no better components
IIRC.

Maybe it has more of them?

SMPSs produce a range of frequencies, but if you are
talking
about random noise, is there any reason to expect it from
a
switched mode supply?

No. If a power supply causes noise, it won't be random.

Well, my question was rhetorical, which should be obvious. I
can see no inherent reason why a SMPS should produce more
noise than a linear supply.

You are wrong to say that power supplies never produce
random noise. Of course they do. You may say it is not
significant in this case, and that may be true.

In any case a
good audio design is not sensitive to power supply noise.

The quality of this particular audio design is at issue,
obviously. And why should a good audio design be insensitive
to noise that, by your account, never exists in the first
place?

Perhaps the more expensive machine has more extensive
noise-reduction? Even if they are identical but with
different power supplies, it could be that the stages are
better isolated or decoupled, including earthing
arrangements, in the quieter machine. Maybe the better one
uses several separately-regulated supplies, for example.

Noise reduction? That is a tool of last resort for a lousy
recording -
it has nothing to do with this kind of product.

Only if your thoughts are limited to some narrow definition
of noise reduction.

Isolation and decoupling have nothing to do with noise.

Don't be silly.

Sure you can reduce noise by decoupling,

But you said...

but you will reduce the wanted signal by the same amount.

Depends what and how you decouple, obviously.

As for earthing arrangements, that will affect hum, not
noise. Again
the regulation of power supplies has nothing to do with
this.

Have you actually given this a moment's thought, or are
you just
writing technical-sounding phrases at random?

Technical sounding? Not my intention. The OP asked for
ideas, so I wrote a few down. I suppose I should expect
territorial cackling from resident self-styled clever-clogs.

Or possibly noise from the source is filtered in one and
not
the other.

Please stop this!

No.

Noise is by definition in the audio band.

Noise is not by definition in the audio band, stupid. The
noise the OP refers to is in the audio band.

If you filter the noise, you filter the signal too.

Could be a product with digital stages for all I know...what
is a DCX2496 anyway?

You are just writing garbage.

Ideas, not answers, around where I judge the OP's
misconceptions lie. A discussion of what is not true is more
likely to be helpful than your loutish posturing, IMO.

Presumably it meets its published specifications.

At last something sensible! The thing to do is measure. If
it meets
its spec, tough - it was bought without sufficient thought
(or it is
not being used correctly). If it doesn't, it can go back
for a refund.

Here I have some sympathy with your analysis, although not
with your tone. If the OP bought a cheap product hoping that
easy mods would bring it up to the level of a higher spec
machine from the same producer, then it's likely to be a
forlorn hope.

OTOH, the OP wonders why there is such a difference. That
question has not been fully answered.

Ian

[Don Pearce]

On Sat, 29 Sep 2007 15:28:10 GMT, "Ian Iveson"
wrote:

Don Pearce wrote

Just that when it's connected up to a 1kW amp and
speaker
system there
is quite an audible hiss with no signal (or small
signal)
applied.
High quiescent noise. With something like it's (far more
expensive)
XTA equivalent this noise does not exist. It's maybe
12dB
better at
least. Yet the XTA appears to use no better components
IIRC.

Maybe it has more of them?

SMPSs produce a range of frequencies, but if you are
talking
about random noise, is there any reason to expect it from
a
switched mode supply?

No. If a power supply causes noise, it won't be random.

Well, my question was rhetorical, which should be obvious. I
can see no inherent reason why a SMPS should produce more
noise than a linear supply.


Really? The you know absolutely zero about switching power supplies.
But that noise should never find its way into the audio path, if
things are correctly designed.

You are wrong to say that power supplies never produce
random noise. Of course they do. You may say it is not
significant in this case, and that may be true.

What I was saying is that switch mode power supplies produce noise
spurs at discrete frequencies. Of course there is also thermal noise
in there, as there must be with anything electronic, but that will be
very far from the dominant noise. Effectively it is perfectly ok to
say it isn't there.

In any case a
good audio design is not sensitive to power supply noise.

The quality of this particular audio design is at issue,
obviously. And why should a good audio design be insensitive
to noise that, by your account, never exists in the first
place?

Is it? I wasn't aware that the design integrity was in question at
this stage. As for that second - not even worthy of an answer.

Perhaps the more expensive machine has more extensive
noise-reduction? Even if they are identical but with
different power supplies, it could be that the stages are
better isolated or decoupled, including earthing
arrangements, in the quieter machine. Maybe the better one
uses several separately-regulated supplies, for example.

Noise reduction? That is a tool of last resort for a lousy
recording -
it has nothing to do with this kind of product.

Only if your thoughts are limited to some narrow definition
of noise reduction.


You have some noise - you reduce it. That noise, in this case, is
audible noise, which means it is in the audio band. Do you have some
magic notion of how you might reduce its level without also reducing
the wanted signal, which occupies the same frequencies?

Isolation and decoupling have nothing to do with noise.

Don't be silly.

Not silly. Noise is to do with signal levels and noise sources.
Decoupling (do you mean power supplies?)is about hum.

Sure you can reduce noise by decoupling,

But you said...

Try and follow. I am doing a reductio ad absurdem proof.

but you will reduce the wanted signal by the same amount.

Depends what and how you decouple, obviously.

If you decouple anything but the signal, you won't reduce the noise,
which is part of that signal. What did you think you might decouple?

As for earthing arrangements, that will affect hum, not
noise. Again
the regulation of power supplies has nothing to do with
this.

Have you actually given this a moment's thought, or are
you just
writing technical-sounding phrases at random?

Technical sounding? Not my intention. The OP asked for
ideas, so I wrote a few down. I suppose I should expect
territorial cackling from resident self-styled clever-clogs.

You didn't write down ideas, you just wrote some random words.

Or possibly noise from the source is filtered in one and
not
the other.

Please stop this!

No.

Noise is by definition in the audio band.

Noise is not by definition in the audio band, stupid. The
noise the OP refers to is in the audio band.

That is what makes it noise. If it were outside the audio band, it
would be inaudible, and hence not noise. Do try and keep up.

If you filter the noise, you filter the signal too.

Could be a product with digital stages for all I know...what
is a DCX2496 anyway?

You mean you wrote all of that without even looking it up to find out
what it was? Now we at least know why you wrote such garbage.

You are just writing garbage.

Ideas, not answers, around where I judge the OP's
misconceptions lie. A discussion of what is not true is more
likely to be helpful than your loutish posturing, IMO.

Since you didn't even know what it was you were talking about, who are
you to judge what the OP's misconceptions might be? You are the one
doing the posturing here.

I only get ****ed off with people who come in here and spout tripe.
Look back at my discussion with the OP if you want to know how I deal
with people who are being reasonable. Your contribution had zero
content, and was clearly designed to do nothing but make you sound
clever. Sadly everything you wrote was ********, so you just looked an
idiot.

Presumably it meets its published specifications.

At last something sensible! The thing to do is measure. If
it meets
its spec, tough - it was bought without sufficient thought
(or it is
not being used correctly). If it doesn't, it can go back
for a refund.

Here I have some sympathy with your analysis, although not
with your tone. If the OP bought a cheap product hoping that
easy mods would bring it up to the level of a higher spec
machine from the same producer, then it's likely to be a
forlorn hope.

OTOH, the OP wonders why there is such a difference. That
question has not been fully answered.

Ian

My feeling about all this is that it is quite hard to make this sort
of kit much noisier than theory allows, and it is very probable that
the OP has simply not got used to it yet, and needs to read up some
more on the gain structure and position his signal more optimally
between noise floor and clipping. My experience of Behringer equipment
leads me to the conclusion that they do not do this stuff wrong, they
understand noise floors and won't release a product with a needless
10dB of excess noise.

d

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

[Dirk Bruere at NeoPax]

Don Pearce wrote:


My feeling about all this is that it is quite hard to make this sort
of kit much noisier than theory allows, and it is very probable that
the OP has simply not got used to it yet, and needs to read up some
more on the gain structure and position his signal more optimally
between noise floor and clipping. My experience of Behringer equipment
leads me to the conclusion that they do not do this stuff wrong, they
understand noise floors and won't release a product with a needless
10dB of excess noise.

I've been using DCXs almost every day for a year in many different
configurations, as well as XTA and BSS Soundweb kit. From the POV of
quiescent noise ie noise when there is no signal present, the Behringer
outperforms a Soundweb 9088ii by at least 10dB - I measured this on
Harman premises with their own equipment. Previously I had spent months
****ing about thinking that I must have made a mistake because everyone
kept telling me what a great bit of kit it was (and is, but not from the
noise POV).

XTA, OTOH, is excellent from this POV.
I want to know why, given that the designs are not too dissimilar.
[Soundweb is an old design and doesn't count in this question]

--
Dirk

http://www.transcendence.me.uk/ - Transcendence UK
Remote Viewing classes in London

[Ian Iveson]

I have no inclination to nurse you through your mendacious
backtracking bluster Don. Sorry.

Ian

"Don Pearce" wrote in message
...
On Sat, 29 Sep 2007 15:28:10 GMT, "Ian Iveson"
wrote:

Don Pearce wrote

Just that when it's connected up to a 1kW amp and
speaker
system there
is quite an audible hiss with no signal (or small
signal)
applied.
High quiescent noise. With something like it's (far
more
expensive)
XTA equivalent this noise does not exist. It's maybe
12dB
better at
least. Yet the XTA appears to use no better components
IIRC.

Maybe it has more of them?

SMPSs produce a range of frequencies, but if you are
talking
about random noise, is there any reason to expect it
from
a
switched mode supply?

No. If a power supply causes noise, it won't be random.

Well, my question was rhetorical, which should be obvious.
I
can see no inherent reason why a SMPS should produce more
noise than a linear supply.

Really? The you know absolutely zero about switching power
supplies.
But that noise should never find its way into the audio
path, if
things are correctly designed.


You are wrong to say that power supplies never produce
random noise. Of course they do. You may say it is not
significant in this case, and that may be true.

What I was saying is that switch mode power supplies
produce noise
spurs at discrete frequencies. Of course there is also
thermal noise
in there, as there must be with anything electronic, but
that will be
very far from the dominant noise. Effectively it is
perfectly ok to
say it isn't there.

In any case a
good audio design is not sensitive to power supply
noise.

The quality of this particular audio design is at issue,
obviously. And why should a good audio design be
insensitive
to noise that, by your account, never exists in the first
place?

Is it? I wasn't aware that the design integrity was in
question at
this stage. As for that second - not even worthy of an
answer.

Perhaps the more expensive machine has more extensive
noise-reduction? Even if they are identical but with
different power supplies, it could be that the stages
are
better isolated or decoupled, including earthing
arrangements, in the quieter machine. Maybe the better
one
uses several separately-regulated supplies, for example.

Noise reduction? That is a tool of last resort for a
lousy
recording -
it has nothing to do with this kind of product.

Only if your thoughts are limited to some narrow
definition
of noise reduction.


You have some noise - you reduce it. That noise, in this
case, is
audible noise, which means it is in the audio band. Do you
have some
magic notion of how you might reduce its level without
also reducing
the wanted signal, which occupies the same frequencies?

Isolation and decoupling have nothing to do with noise.

Don't be silly.

Not silly. Noise is to do with signal levels and noise
sources.
Decoupling (do you mean power supplies?)is about hum.

Sure you can reduce noise by decoupling,

But you said...

Try and follow. I am doing a reductio ad absurdem proof.

but you will reduce the wanted signal by the same
amount.

Depends what and how you decouple, obviously.

If you decouple anything but the signal, you won't reduce
the noise,
which is part of that signal. What did you think you might
decouple?

As for earthing arrangements, that will affect hum, not
noise. Again
the regulation of power supplies has nothing to do with
this.

Have you actually given this a moment's thought, or are
you just
writing technical-sounding phrases at random?

Technical sounding? Not my intention. The OP asked for
ideas, so I wrote a few down. I suppose I should expect
territorial cackling from resident self-styled
clever-clogs.

You didn't write down ideas, you just wrote some random
words.

Or possibly noise from the source is filtered in one and
not
the other.

Please stop this!

No.

Noise is by definition in the audio band.

Noise is not by definition in the audio band, stupid. The
noise the OP refers to is in the audio band.

That is what makes it noise. If it were outside the audio
band, it
would be inaudible, and hence not noise. Do try and keep
up.

If you filter the noise, you filter the signal too.

Could be a product with digital stages for all I
know...what
is a DCX2496 anyway?

You mean you wrote all of that without even looking it up
to find out
what it was? Now we at least know why you wrote such
garbage.

You are just writing garbage.

Ideas, not answers, around where I judge the OP's
misconceptions lie. A discussion of what is not true is
more
likely to be helpful than your loutish posturing, IMO.

Since you didn't even know what it was you were talking
about, who are
you to judge what the OP's misconceptions might be? You
are the one
doing the posturing here.

I only get ****ed off with people who come in here and
spout tripe.
Look back at my discussion with the OP if you want to know
how I deal
with people who are being reasonable. Your contribution
had zero
content, and was clearly designed to do nothing but make
you sound
clever. Sadly everything you wrote was ********, so you
just looked an
idiot.

Presumably it meets its published specifications.

At last something sensible! The thing to do is measure.
If
it meets
its spec, tough - it was bought without sufficient
thought
(or it is
not being used correctly). If it doesn't, it can go back
for a refund.

Here I have some sympathy with your analysis, although not
with your tone. If the OP bought a cheap product hoping
that
easy mods would bring it up to the level of a higher spec
machine from the same producer, then it's likely to be a
forlorn hope.

OTOH, the OP wonders why there is such a difference. That
question has not been fully answered.

Ian

My feeling about all this is that it is quite hard to make
this sort
of kit much noisier than theory allows, and it is very
probable that
the OP has simply not got used to it yet, and needs to
read up some
more on the gain structure and position his signal more
optimally
between noise floor and clipping. My experience of
Behringer equipment
leads me to the conclusion that they do not do this stuff
wrong, they
understand noise floors and won't release a product with a
needless
10dB of excess noise.

d

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

[Don Pearce]

On Sat, 29 Sep 2007 16:52:16 GMT, "Ian Iveson"
wrote:

Sorry.

Accepted.

d

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

[Dirk Bruere at NeoPax]

François Yves Le Gal wrote:
On Sat, 29 Sep 2007 13:50:31 +0100, Eeyore
wrote:

That suggests the analogue inputs don't operate at a suitable level.

Yup, hit them as hard as you can - below saturation of course - and the DCX
will be OK.

I use digital inputs to get around that problem.
So the quiescent noise I'm getting is either analogue from the DACs and
opamps or something peculiar with the arithmetic.

--
Dirk

http://www.transcendence.me.uk/ - Transcendence UK
Remote Viewing classes in London

[Dirk Bruere at NeoPax]

François Yves Le Gal wrote:
On Sat, 29 Sep 2007 22:54:28 +0100, Dirk Bruere at NeoPax
wrote:

I use digital inputs to get around that problem.

OK. Then the outputs are at +22 dB, noise will be perceptible if the power
amps are wide open/the drivers are high efficiency/etc. Try to pad the
levels.

I should have said earlier, but this is with everything set at 0dB

--
Dirk

http://www.transcendence.me.uk/ - Transcendence UK
Remote Viewing classes in London

[Eeyore]

Dirk Bruere at NeoPax wrote:

Eeyore wrote:
dezza wrote:

These guys offer a power supply mod that claims decreased noise.
Perhaps they may know the answer to your questions.

http://www.selectronic.fr/dcx2496_us.asp#11

Mostly snake oil I'm sure.


I don't see their I/O unit being an improvement over the Behringer
The PSU OTOH might be of interest.

Possibly worth evaluating. It's the only mod they offer that seems to have any
scientific basis for any alleged improvements.

Graham

[Dirk Bruere at NeoPax]

François Yves Le Gal wrote:
On Sun, 30 Sep 2007 14:35:24 +0100, Dirk Bruere at NeoPax
wrote:

I should have said earlier, but this is with everything set at 0dB

Everything?

No EQ filters set, in and out sliders to 0dB

--
Dirk

http://www.transcendence.me.uk/ - Transcendence UK
Remote Viewing classes in London

[Dirk Bruere at NeoPax]

François Yves Le Gal wrote:
On Mon, 01 Oct 2007 00:39:10 +0100, Dirk Bruere at NeoPax
wrote:

No EQ filters set, in and out sliders to 0dB

Tthe 0 dB are local levels and translate to +22 dB in the real world.



So what does that mean? That if I put a 1V signal into the Behringer
with 0dB set throughout I'll get 22V out????

--
Dirk

http://www.transcendence.me.uk/ - Transcendence UK
Remote Viewing classes in London

[Arny Krueger]

"Dirk Bruere at NeoPax" wrote in
message
François Yves Le Gal wrote:
On Mon, 01 Oct 2007 00:39:10 +0100, Dirk Bruere at NeoPax
wrote:

No EQ filters set, in and out sliders to 0dB

Tthe 0 dB are local levels and translate to +22 dB in
the real world.


So what does that mean? That if I put a 1V signal into
the Behringer with 0dB set throughout I'll get 22V out????

Where this seems to be headed is that:

(1) If the DCX 2496 is set up to be overall, a unity gain device
(2) If the input is 1 volt

then:

The output will be 1 volt.

However, the ca. -96 dB dynamic range of the device will be referenced to
+22 dB which is about 10 volts. The noise at the analog output terminals of
the DCX 2496 will be about 74 dB down. If you put that into a really large
power amp whose gain control is set to max, and the speakers are fairly
efficient, then there will be considerable audible noise.

If you set up the DCX 2496 so that its overall gain is 5 or 14 dB, and you
set the gain of the power amp to be 14 dB less than max, then the noise
floor of the DCX2496 will be effectively 14 dB lower, or about 88 dB down.
This will reduce the noise at the speakers to a very significant degree.
There will still be about 6 dB worth of headroom in the DCX 2496, which
should be enough for a controlled circumstance like this.