Mixing 4 audio channels to 3?

D

DaveC

Guest
I want to connect the audio out (headphone jack) of 2 computer sound cards to
a desktop & woofer speaker-amp combination. (The original input was via USB
only and I'm modifying this for analog audio input.)

This is my guess at the necessary resistors to mix down these 4 outputs to
the 3 inputs in the amplifier (L, R, sub):

<http://i41.tinypic.com/97mpud.jpg>

The data sheet for the Philips TDA8510J amp IC shows an application for 2
channel input:

<http://i40.tinypic.com/5ets9w.jpg>

Suggestions welcome.

Thanks.
 
If I'm on the right track, what are the suggested values for resistors?

Thahks.
 
On Wed, 26 Oct 2011 08:50:55 -0700 (PDT), NT <meow2222@care2.com>
wrote:

On Oct 26, 7:58 am, DaveC <inva...@invalid.net> wrote:
I want to connect the audio out (headphone jack) of 2 computer sound cards to
a desktop & woofer speaker-amp combination. (The original input was via USB
only and I'm modifying this for analog audio input.)

This is my guess at the necessary resistors to mix down these 4 outputs to
the 3 inputs in the amplifier (L, R, sub):

http://i41.tinypic.com/97mpud.jpg

Lose Rs 1-4, keep 5-8, lose R11, and lose C1,2,3, then itll be good.


NT
Click to expand...
Is there something wrong with my browser? I can't see any of these
component references. Actually I see what you mean. One problem here
is that R9 and R10 are going to cause crosstalk. Without some active
electronics there is no way around that apart from taking the sub from
one channel only. This is actually quite common.

d
 
Is there something wrong with my browser? I can't see any of these
component references.
Click to expand...
Sorry. They didn't make onto the jpg file. Added here:

<http://i40.tinypic.com/5ets9w.jpg>

Actually I see what you mean. One problem here
is that R9 and R10 are going to cause crosstalk. Without some active
electronics there is no way around that apart from taking the sub from
one channel only. This is actually quite common.
d
Click to expand...
OK, can do that.

Suggested values for resistors?

Thanks.
 
Lose Rs 1-4, keep 5-8, lose R11, and lose C1,2,3, then itll be good.
NT
Click to expand...
Callouts added:

<http://i40.tinypic.com/5ets9w.jpg>

R11 provides some adjustment for sub. Otherwise no way to have any difference
between L & R channel volume and sub volume.

The caps are suggested on the data sheet example:

<http://i40.tinypic.com/5ets9w.jpg>

Why would you do away with them?

What values for the resistors do you suggest?

Thanks.
 
On Wed, 26 Oct 2011 10:10:12 -0700, DaveC <invalid@invalid.net> wrote:

Lose Rs 1-4, keep 5-8, lose R11, and lose C1,2,3, then itll be good.
NT

Callouts added:

http://i40.tinypic.com/5ets9w.jpg

R11 provides some adjustment for sub. Otherwise no way to have any difference
between L & R channel volume and sub volume.

The caps are suggested on the data sheet example:

http://i40.tinypic.com/5ets9w.jpg

Why would you do away with them?

What values for the resistors do you suggest?

Thanks.
Click to expand...
For that particular circuit with that chip, the caps are necessary
because those inputs are not at ground potential. If you are feeding
normal hi fi unit inputs the caps aren't needed. Resistor values
around 5 to 10 k would be what you need.

d
 
On Oct 26, 7:58 am, DaveC <inva...@invalid.net> wrote:
I want to connect the audio out (headphone jack) of 2 computer sound cards to
a desktop & woofer speaker-amp combination. (The original input was via USB
only and I'm modifying this for analog audio input.)

This is my guess at the necessary resistors to mix down these 4 outputs to
the 3 inputs in the amplifier (L, R, sub):

http://i41.tinypic.com/97mpud.jpg
Click to expand...
Lose Rs 1-4, keep 5-8, lose R11, and lose C1,2,3, then itll be good.


NT
 
For that particular circuit with that chip, the caps are necessary
because those inputs are not at ground potential. If you are feeding
normal hi fi unit inputs the caps aren't needed. Resistor values
around 5 to 10 k would be what you need.
Click to expand...
And the value for the potentiometer?

Thanks.
 
On Wed, 26 Oct 2011 11:18:39 -0700, DaveC <invalid@invalid.net> wrote:

For that particular circuit with that chip, the caps are necessary
because those inputs are not at ground potential. If you are feeding
normal hi fi unit inputs the caps aren't needed. Resistor values
around 5 to 10 k would be what you need.

And the value for the potentiometer?

That depends entirely on the input impedance of the subwoofer. You
Click to expand...
could try a 100k pot and connect it as a normal volume control - one
end to the incoming signal, the other end to ground and the slider to
the subwoofer.

d
 
That depends entirely on the input impedance of the subwoofer. You
could try a 100k pot and connect it as a normal volume control - one
end to the incoming signal, the other end to ground and the slider to
the subwoofer.

d
Click to expand...
The data sheet says that the 2 single-ended inputs (R & L) are 50K impedance.
The bridge-tied load ("BTL") input -- used for the sub -- is 25K impedance.

Does this suggest any change to your recommendation of a 100K pot tied to
ground?

Thanks.
 
On Wed, 26 Oct 2011 11:52:49 -0700, DaveC <invalid@invalid.net> wrote:

That depends entirely on the input impedance of the subwoofer. You
could try a 100k pot and connect it as a normal volume control - one
end to the incoming signal, the other end to ground and the slider to
the subwoofer.

d

The data sheet says that the 2 single-ended inputs (R & L) are 50K impedance.
The bridge-tied load ("BTL") input -- used for the sub -- is 25K impedance.

Does this suggest any change to your recommendation of a 100K pot tied to
ground?

Thanks.
Click to expand...
No, that would be about right.

d
 
spam@spam.com (Don Pearce) writes:

On Wed, 26 Oct 2011 08:50:55 -0700 (PDT), NT <meow2222@care2.com
wrote:

On Oct 26, 7:58 am, DaveC <inva...@invalid.net> wrote:
I want to connect the audio out (headphone jack) of 2 computer sound cards to
a desktop & woofer speaker-amp combination. (The original input was via USB
only and I'm modifying this for analog audio input.)

This is my guess at the necessary resistors to mix down these 4 outputs to
the 3 inputs in the amplifier (L, R, sub):

http://i41.tinypic.com/97mpud.jpg

Lose Rs 1-4, keep 5-8, lose R11, and lose C1,2,3, then itll be good.


NT

Is there something wrong with my browser? I can't see any of these
component references. Actually I see what you mean. One problem here
is that R9 and R10 are going to cause crosstalk. Without some active
electronics there is no way around that apart from taking the sub from
one channel only. This is actually quite common.
Click to expand...
I don't see component references, either. However, assuming R1-R4 are
the resistors from the inputs to ground, they may be necessary to provide
a proper impedance to the source driving them.

To reduce crosstalk, I'd replace the two resistors from the 220nF
cap junctions to the pot with four resistors from each of the four inputs
to the pot. Presumably, the output impedance of the amps is low compared
to the resistors reducing crosstalk.

For mixing signals there is a method using amps and feedback where the
connection between multiple sources is at a virtual ground, eliminating
crosstalk.
 
On Oct 26, 5:09 pm, s...@spam.com (Don Pearce) wrote:
On Wed, 26 Oct 2011 08:50:55 -0700 (PDT), NT <meow2...@care2.com
wrote:

On Oct 26, 7:58 am, DaveC <inva...@invalid.net> wrote:
I want to connect the audio out (headphone jack) of 2 computer sound cards to
a desktop & woofer speaker-amp combination. (The original input was via USB
only and I'm modifying this for analog audio input.)

This is my guess at the necessary resistors to mix down these 4 outputs to
the 3 inputs in the amplifier (L, R, sub):

http://i41.tinypic.com/97mpud.jpg

Lose Rs 1-4, keep 5-8, lose R11, and lose C1,2,3, then itll be good.

NT

Is there something wrong with my browser? I can't see any of these
component references. Actually I see what you mean. One problem here
is that R9 and R10 are going to cause crosstalk. Without some active
electronics there is no way around that apart from taking the sub from
one channel only. This is actually quite common.

d
Click to expand...
I was hinting that the OP might decide to add numbers to the
components, otherwise its sufficiently tedious to talk about them that
people often just wont bother.

Yes the caps are necessary, but for the amp IC, not for the mixer.
House them together and its the same difference, house them apart and
the distiction matters.


NT
 
Yes the caps are necessary, but for the amp IC, not for the mixer.
House them together and its the same difference, house them apart and
the distiction matters.
NT
Click to expand...
This is unclear to me.

Do you mean that if the mixer resistor matrix is housed close to the amp IC
that the capacitors won't be needed? And if they are separated and connected
by shielded audio cables that the caps will be needed?

Thanks.
 
Latest rev:

<http://i40.tinypic.com/300feaq.jpg>

All resistors 5K-10K. Pot 100K.

Comments welcome.

Thanks.
 
DaveC <invalid@invalid.net> writes:

If I'm on the right track, what are the suggested values for resistors?
Click to expand...
10 kohms for all fixed resistor should work...


--
Tomi Engdahl (http://www.iki.fi/then/)
Take a look at my electronics web links and documents at
http://www.epanorama.net/
 
On Oct 27, 5:17 am, DaveC <inva...@invalid.net> wrote:
Yes the caps are necessary, but for the amp IC, not for the mixer.
House them together and its the same difference, house them apart and
the distiction matters.
NT

This is unclear to me.

Do you mean that if the mixer resistor matrix is housed close to the amp IC
that the capacitors won't be needed? And if they are separated and connected
by shielded audio cables that the caps will be needed?

Thanks.
Click to expand...
Nothing like that, no. The amp IC requires the caps, so theyre needed,
no question about it. But theyre needed for the amp, not for mixing
purposes, so if the 2 circuits can be disconnected by the end user,
then the caps need to be with the IC amp part, not the mixer part.


NT
 
On Wed, 26 Oct 2011 21:17:16 -0700, DaveC <invalid@invalid.net> wrote:

Yes the caps are necessary, but for the amp IC, not for the mixer.
House them together and its the same difference, house them apart and
the distiction matters.
NT

This is unclear to me.

Do you mean that if the mixer resistor matrix is housed close to the amp IC
that the capacitors won't be needed? And if they are separated and connected
by shielded audio cables that the caps will be needed?

Thanks.
Click to expand...
---
Since the chip is being powered by single positive DC supply, The
only way to get signals which swing below ground into and out of the
chip is to use capacitive coupling.

--
JF
 
"John Fields" <jfields@austininstruments.com> wrote in message
news:irhia7h795aq53t9enlqls8i7ia6eppofq@4ax.com...
On Wed, 26 Oct 2011 21:17:16 -0700, DaveC <invalid@invalid.net> wrote:

Yes the caps are necessary, but for the amp IC, not for the mixer.
House them together and its the same difference, house them apart and
the distiction matters.
NT

This is unclear to me.

Do you mean that if the mixer resistor matrix is housed close to the amp
IC
that the capacitors won't be needed? And if they are separated and
connected
by shielded audio cables that the caps will be needed?

Thanks.

---
Since the chip is being powered by single positive DC supply, The
only way to get signals which swing below ground into and out of the
chip is to use capacitive coupling.
Click to expand...
Or, as is not uncommonly done these days in portable digital players,
establish a floating ground for the purpose of providing and signal for the
(floating) load, so that the difference between the output and the floating
ground can become negative.
 
Or, as is not uncommonly done these days in portable digital players,
establish a floating ground for the purpose of providing and signal for the
(floating) load, so that the difference between the output and the floating
ground can become negative.
Click to expand...
Which raises the question I was pondering recently: is it OK to connect 2
computers' sound cards' grounds together?

The PS for the powered speakers is via "wall wart" (not grounded).

Thanks.
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top