DC Blocking

[fdeck]

I would like to know if it is possible to take a (hopefully) inexpensive USB audio interface, and remove the DC blocking on the audio inputs. Presumably, I could trace the audio input signals to the input of the ADC, and insert my own buffered signal at that point.

But I am concerned that the audio interface might contain some sort of digital filtering that effectively blocks the DC signal. Does anybody know if this is the case?

FWIW it's for a non-audio application. I already have the necessary software.

If this can't be done with a commercial interface, are there any inexpensive kits that I could modify for this purpose?

[Iain McNeill]

So you're trying to use a sound card to do data logging of some sort?
I've wondered the same thing.

It probably depends on the ADC/codec. I would think that any delta/sigma oversampling ADC would inherently have a high pass built in but I've no idea what frequency it is.

You'll also have to worry about offsets. The inputs will have to be biased to Vcc/2 regardless of gain.

If you've got a spare soundcard to trash and want to try I'm sure a lot of people here would be interested in the results.

[fdeck]

I might just give it a try. So far as high pass filters, I have one cause for hope. The audio files that I have recorded always have a DC offset that is dependent on the sound card and channel. You'd think that a digital DC blocker would be more perfect than that.

I will handle the gain and offsets, if necessary, through some sort of calibration. Temperature drift will be whatever it turns out to be. I will certainly share my results!

[jcx]

High Pass filtering is common in audio ADC, look up the chip used in your card

the high quality AK5394 appears to have a pin to enable/disable the high pass so hacking the board could work

http://www.asahi-kasei.co.jp/akm/en/...5394a_f03e.pdf

[fdeck]

Bummer. That's a fly in the ointment, since all of the computers in the house have motherboard audio that I don't want to hack. I was hoping to do this with a cheap USB audio interface, which I would have to buy without knowing the codec.

Those dirty dogs, spoiling a perfectly good hack.

[kevinkr]

Consider using a voltage controlled oscillator instead to drive the a/d, depending on the precision you need this might be enough. - No modifications required to the onboard sound hardware, and the accuracy is largely determined by the accuracy of your VCO.. (Calibration would be required.)

You will obviously need software that can convert the frequency measured back to a representation of the original dc value that generated it.

And it would be fairly slow..

Alternatively perhaps you can acquire a decent data acquisition card somewhere, not sure what your options are for USB though.