In rec.audio.tech R wrote:
Stewart Pinkerton wrote in
:

On Wed, 15 Dec 2004 22:45:20 GMT, "Rich.Andrews"
wrote:

"Arny Krueger" wrote in news:QN2dnU2d2-ywhV3cRVn-
:

"R" wrote in message
. 1
I am looking for an sound card that features dual DA converters in
parallel. Does anyone know of one?

Please explain more precisely what you mean.


Arny,

What I mean by parallel is that the data stream for one channel feeds 2
dacs
at once and the resultant output of the dacs are tied together. Many of
the
high end CD players and D-A units use that circuit topology as it lowers
the
distortion levels.

You don't mean parallel, you mean a differential pair. Given that
distortion levels with conventional DACs can be 0.001% or less, do you
think this is important?

I don't think i mean a differential pair. Wuld not a differential pair be
2 dacs fed with the same source but one dac has it's invert pin asserted?

Can someone tell me what the difference is here?

Summing the output of two DACs should give "s1 + s2".
A differential pair, as I understand it, would be "s1 - (-s2)"

I don't see a difference. Am I missing something?

"Colin B." wrote in message

In rec.audio.tech R wrote:
Stewart Pinkerton wrote in
:

On Wed, 15 Dec 2004 22:45:20 GMT, "Rich.Andrews"
wrote:

"Arny Krueger" wrote in
news:QN2dnU2d2-ywhV3cRVn- :

"R" wrote in message
. 1
I am looking for an sound card that features dual DA converters
in parallel. Does anyone know of one?

Please explain more precisely what you mean.


Arny,

What I mean by parallel is that the data stream for one channel
feeds 2 dacs at once and the resultant output of the dacs are tied
together. Many of the high end CD players and D-A units use that
circuit topology as it lowers the distortion levels.

You don't mean parallel, you mean a differential pair. Given that
distortion levels with conventional DACs can be 0.001% or less, do
you think this is important?

I don't think i mean a differential pair. Wuld not a differential
pair be 2 dacs fed with the same source but one dac has it's invert
pin asserted?

Can someone tell me what the difference is here?

Summing the output of two DACs should give "s1 + s2".
A differential pair, as I understand it, would be "s1 - (-s2)"

I don't see a difference. Am I missing something?

Your algebra is correct for a simpler case, but it ignores the inversion of
the digital input, and the effect of that inversion on distortion.

"Colin B." wrote in
:

In rec.audio.tech R wrote:
Stewart Pinkerton wrote in
:

On Wed, 15 Dec 2004 22:45:20 GMT, "Rich.Andrews"
wrote:

"Arny Krueger" wrote in news:QN2dnU2d2-ywhV3cRVn-
:

"R" wrote in message
. 1
I am looking for an sound card that features dual DA converters in
parallel. Does anyone know of one?

Please explain more precisely what you mean.


Arny,

What I mean by parallel is that the data stream for one channel feeds
2
dacs
at once and the resultant output of the dacs are tied together. Many
of
the
high end CD players and D-A units use that circuit topology as it
lowers
the
distortion levels.

You don't mean parallel, you mean a differential pair. Given that
distortion levels with conventional DACs can be 0.001% or less, do you
think this is important?

I don't think i mean a differential pair. Wuld not a differential pair
be 2 dacs fed with the same source but one dac has it's invert pin
asserted?

Can someone tell me what the difference is here?

Summing the output of two DACs should give "s1 + s2".
A differential pair, as I understand it, would be "s1 - (-s2)"

I don't see a difference. Am I missing something?


As I see it, in a balanced audio setup the positive dac output would be
the plus side and the inverted dac output would be the negative side. The
common mode noise would be diminished in such a configuration.

In the configuration I am referring to, it would e a simple single ended
unbalanced output but with 2 dacs sharing inputs and outputs.


r

"Rich.Andrews" wrote in message


As I see it, in a balanced audio setup the positive dac output would
be the plus side and the inverted dac output would be the negative
side. The common mode noise would be diminished in such a
configuration.

Random noise generated by either DAC, being uncorrelated, won't cancel and
won't even be diminished. However, it will add geometrically, not
algebraically. The signal adds algebraically for a net 3 dB improvement in
dynamic range, all other sources of noise excluded.

What cancels is even-order nonlinear distortion. Unfortunately, the best
modern DACs already have almost none of this.

"Colin B." wrote:

In rec.audio.tech R wrote:
Stewart Pinkerton wrote in
:

On Wed, 15 Dec 2004 22:45:20 GMT, "Rich.Andrews"
wrote:

"Arny Krueger" wrote in news:QN2dnU2d2-ywhV3cRVn-
:

"R" wrote in message
. 1
I am looking for an sound card that features dual DA converters in
parallel. Does anyone know of one?

Please explain more precisely what you mean.


Arny,

What I mean by parallel is that the data stream for one channel feeds 2
dacs
at once and the resultant output of the dacs are tied together. Many of
the
high end CD players and D-A units use that circuit topology as it lowers
the
distortion levels.

You don't mean parallel, you mean a differential pair. Given that
distortion levels with conventional DACs can be 0.001% or less, do you
think this is important?

I don't think i mean a differential pair. Wuld not a differential pair be
2 dacs fed with the same source but one dac has it's invert pin asserted?

Can someone tell me what the difference is here?

Summing the output of two DACs should give "s1 + s2".
A differential pair, as I understand it, would be "s1 - (-s2)"

I don't see a difference. Am I missing something?

Yup ! ;-)

The principle assumed a conversion error that had a common factor for a given
digital code input. Typically R-2R ladder types.

The output = ( sig1 + error ) - ( -sig2 + error )

= sig 1 + sig2 ( and no error ) = 2 x sig1

Modern converters aren't like this.

Graham