Good digitally controlled volume IC?

Professional audio recording and studio engineering, post #45,894
Author:
Date:
Subject:
 David Gravereaux
 2008-07-08 20:51:35
 Good digitally controlled volume IC?
Hi guys,

I was wondering what you circuit hackers use?

I'm pondering the TI (Burr-Brown) PGA4311
<http://focus.ti.com/lit/ds/symlink/pga4311.pdf>, 4 channels, $5.95/1k,
looks good.

The Cirrus CS3310
<http://www.cirrus.com/en/pubs/proDatasheet/CS3310_F1.pdf> is just 2
channel and I need three for my application. It's rather expensive.

The Maxim DS1841 <http://datasheets.maxim-ic.com/en/ds/DS1841.pdf> is a
bit minimalist as I'll have to implement a zero-cross delay for
clickless changes, is missing an output buffer and has only 128 steps.
It sure is cheap though.

This is post a D/A converter and needs to just drive a 10K load for an
amplifier input (short line, same box and PSU reference).

We found division in the digital domain not to have a pleasing behavior,
so I need to go analog to avoid quantizing artifacts. FYI, the input
for the A/D will have something similar, too, for adjusting gain
structure, but I'm not at that half of it yet.
Author:
Date:
Subject:
 LAB
 2008-07-09 09:04:31
 Re: Good digitally controlled volume IC?
>> I'm pondering the TI (Burr-Brown) PGA4311
<http://focus.ti.com/lit/ds/symlink/pga4311.pdf>, 4 channels, $5.95/1k,
looks good.

It seems to be very good. It's 6 years old, then package is big enough
to let you test it without soldering problems.

Gianluca
Author:
Date:
Subject:
 David Gravereaux
 2008-07-09 01:08:15
 Re: Good digitally controlled volume IC?
LAB wrote:
> >> I'm pondering the TI (Burr-Brown) PGA4311
> <http://focus.ti.com/lit/ds/symlink/pga4311.pdf>, 4 channels, $5.95/1k,
> looks good.
>
> It seems to be very good. It's 6 years old, then package is big enough
> to let you test it without soldering problems.

Yes, I thought the package size was a bit bulky ;) Thanks for double check.
Author:
Date:
Subject:
 Scott Dorsey
 2008-07-09 09:01:08
 Re: Good digitally controlled volume IC?
David Gravereaux <[email protected]> wrote:
>Hi guys,
>
>I was wondering what you circuit hackers use?

If you want actual good sound, the solution is stepped attenuators with
relay control or a motor-controlled pot. None of the digital volume
control devices sound very good to my ears. They sure are popular, though.

>I'm pondering the TI (Burr-Brown) PGA4311
><http://focus.ti.com/lit/ds/symlink/pga4311.pdf>, 4 channels, $5.95/1k,
>looks good.
>
>The Cirrus CS3310
><http://www.cirrus.com/en/pubs/proDatasheet/CS3310_F1.pdf> is just 2
>channel and I need three for my application. It's rather expensive.

I have not used either of these.

>The Maxim DS1841 <http://datasheets.maxim-ic.com/en/ds/DS1841.pdf> is a
>bit minimalist as I'll have to implement a zero-cross delay for
>clickless changes, is missing an output buffer and has only 128 steps.
>It sure is cheap though.

I have used this one for communications applications; it's not something
I would pick for hi-fi work unless my power budget was really tight. But
if you are short on power, it could be a good pick.

>This is post a D/A converter and needs to just drive a 10K load for an
>amplifier input (short line, same box and PSU reference).

Note that many of the ladder-type volume control devices are very, very
picky about load impedance, and that includes the classic parts.

>We found division in the digital domain not to have a pleasing behavior,
>so I need to go analog to avoid quantizing artifacts. FYI, the input
>for the A/D will have something similar, too, for adjusting gain
>structure, but I'm not at that half of it yet.

This sounds to me like either you have a dithering issue, or else your
analogue section isn't as linear as you want it to be. Digital volume
control should be a lot more neutral than doing it with any of these
gadgets (although it may be less neutral than a mechanical pot).

You could also look into the TDA7448, which seems popular. But the whole
idea kind of gives me the willies.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Author:
Date:
Subject:
 David Gravereaux
 2008-07-09 13:46:06
 Re: Good digitally controlled volume IC?
Thanks Scott,

The general idea is to shift the headroom knowing what the playing level
is to be. IOW, when playing the system at low levels that can exercise
low resolution bit depths, up the gain going into the A/D and drop it on
the D/A side.

This is all just *speculation* on part of the manager (not me), as we
have yet to measure this thing end-to-end connected.