Hi guys,
I am using DS4420 digitally controlled audio gain controller..But it has 3db steps in its every gain change. So I am getting a zip sound everytime I change the volume. I cannot change the design now. How do I eliminate the zipper noise ? My requirements were I2c controlled addressable(I needed 8 of them) so went for DS4420. (I didnt wanted digital pots so)
I even browsed on internet for articles on it. But found just this Eliminate Zipper noise
More seriously you cannot eliminate the noise, its fundamental to a step-wise volume envelope.
I would call 3db steps "gain control", not "volume control".
Volume control should be done in 1db steps.
Do you also hear the steps during silence (DC blocking coupling caps).
Since you have tried so many of these ICs, it's maybe time to show a diagram of the surrounding parts.
Leo..
I even browsed on internet for articles on it. But found just this Eliminate Zipper noise
Yes that sums up all your options nicely I can thing of nothing more you can do than is contained in that document.
Have you tried putting it into mute mode before you change the gain?
Thank you all for the replies. Yes I tried to put in mute mode before I change the gain. But it has the zipper noise.
Now you know the difference between a gain controller/selector and a real digital volume control...
The proper approach is to multiply by a gain value to full 16 or 24 bit precision and ramp that
control value smoothly (the control value is mixed with the audio signal so needs the same accuracy
and have no components in the audio range). Normally this would be done by the digital sound
source itself as there is usually a DSP involved.
There a lot of chips with 0.5dB steps which is a lot better than 3dB steps, note.
Still would like to see the full diagram.
Course steps does not necessarily add zipper noise.
DC superimposed onto the audio does.
Leo..
Wawa:
Still would like to see the full diagram.
Course steps does not necessarily add zipper noise.
DC superimposed onto the audio does.
Leo..
I agree. When I read this, my first thought was that maybe the gain was being adjusted before the DC offset was removed.