[time-nuts] Using cheap sound cards for measurements

[Lux, Jim (337C)]

On 8/22/09 5:37 AM, "Hal Murray" <hmurray at megapathdsl.net> wrote:

> 
> 
>>> Add internal or external (5/10 MHz) clock.
> 
>> None will have that...<grin> Maybe a multiple of 48 or 44.1 kHz
> 
> The key idea is to have an external clock input.  I don't think the frequency
> is critical for our usage.
> 
> All the audio chips I've looked at have a DSP filter that scales with the
> clock frequency.  I'd expect them to work as long as the clock isn't too
> fast.  You will just get a strange sampling rate.
> 
> The analog front end has a rough anti-aliasing filter.  You might get
> troubles if the clock is too slow and the input signal has energy in the band
> that gets aliased down on top of the signal.
>

That's an interesting idea.  I would imagine that the clock going into the
chip is probably some multiple of the sample rate (e.g. 48kHz*16*2 = 1.536
MHz), so you could pick the closest 1/N from 10 MHz and pump that in.

However, what about the USB interface?  These are inexpensive devices, and
I'll bet all the rates are carefully chosen so that everything shares one
clock.