Forums

sound card based sampling oscilloscope

Started by Glen Walpert November 1, 2011
Looks like a nice cheap hobby project, ~40 MSPS effective sampling rate 
on repetitive signals with a 96 kSPS soundcard.
http://www.analog.com/library/analogdialogue/archives/45-11/soundcard.html
On a sunny day (02 Nov 2011 01:43:09 GMT) it happened Glen Walpert
<nospam@null.void> wrote in
<4eb0a02d$0$21176$c3e8da3$e408f015@news.astraweb.com>:

>Looks like a nice cheap hobby project, ~40 MSPS effective sampling rate >on repetitive signals with a 96 kSPS soundcard. >http://www.analog.com/library/analogdialogue/archives/45-11/soundcard.html
Hey that is nice, maybe I can add that AD738 to my PIC scope, that would extend the range a lot: http://panteltje.com/panteltje/pic/scope_pic/
On 02 Nov 2011 01:43:09 GMT, Glen Walpert <nospam@null.void> wrote:

>Looks like a nice cheap hobby project, ~40 MSPS effective sampling rate >on repetitive signals with a 96 kSPS soundcard. >http://www.analog.com/library/analogdialogue/archives/45-11/soundcard.html
That's fun and educational, but one problem is that it's not triggered. It gets its timebase by heterodyning the signal against an independent local oscillator. So it only works with repetitive signals that have very low period jitter. With the timebase suggested in the appnote, the display time/division is difficult to figure out, and it's easy to alias the waveform into something deceptive, like time reversal for instance. In fact, you have to know the input waveform frequency in advance, otherwise it's difficult or impossible to know if the timebase divisor is right. LeCroy did a sampling scope with a free-running, synthesized timebase. I don't recall it being a successful product. John
John Larkin <jjlarkin@highNOTlandTHIStechnologyPART.com> wrote:

>On 02 Nov 2011 01:43:09 GMT, Glen Walpert <nospam@null.void> wrote: > >>Looks like a nice cheap hobby project, ~40 MSPS effective sampling rate >>on repetitive signals with a 96 kSPS soundcard. >>http://www.analog.com/library/analogdialogue/archives/45-11/soundcard.html > >That's fun and educational, but one problem is that it's not >triggered. It gets its timebase by heterodyning the signal against an >independent local oscillator. So it only works with repetitive signals >that have very low period jitter. > >With the timebase suggested in the appnote, the display time/division >is difficult to figure out, and it's easy to alias the waveform into >something deceptive, like time reversal for instance. In fact, you >have to know the input waveform frequency in advance, otherwise it's >difficult or impossible to know if the timebase divisor is right. > >LeCroy did a sampling scope with a free-running, synthesized timebase. >I don't recall it being a successful product.
Maybe they forgot to mention the optional trigger module in the catalog? Otherwise I don't see how such a scope would produce meaningfull readings in any situation. -- Failure does not prove something is impossible, failure simply indicates you are not using the right tools... nico@nctdevpuntnl (punt=.) --------------------------------------------------------------