[time-nuts] Using cheap sound cards for measurements

Demian Martin demianm_1 at yahoo.com
Fri Aug 21 22:34:52 UTC 2009


A decent "home studio" soundcard would have easily 60+ dB of separation
across it range and probably much higher. For the outrageous sum of around
$150 you can get the ESI Juli@ from eBay (and real dealers) with better than
60 dB of separation, below 5 Hz to 100 KHz response, differential inputs and
good Windows, Mac and Linux support. It seems any other approach would
probably value time at below minimum wage or less for the effort. Here is a
review showing 100+ dB separation http://ixbtlabs.com/articles2/esi-julia/
The card uses local crystals for its oscillators so it has lower jitter than
the cheap solutions. If it helped the measurement you could drive the
crystal input with an external oscillator etc. For more extended low
frequency response you could bypass the input caps but the dc offsets would
need attention. 

Demian Martin
Product Design Services

Date: Fri, 21 Aug 2009 22:51:40 +0200
From: Christian Vogel <vogelchr at vogel.cx>
Subject: Re: [time-nuts] Using cheap sound cards for measurements
To: Discussion of precise time and frequency measurement
	<time-nuts at febo.com>
Message-ID: <4A8F08DC.6010708 at vogel.cx>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed

Hi Lux,
> Syncing inexpensive cards is a real chore (and the only reason to be 
> thinking about using this in the first place is to keep the cost to a 
> minimum, otherwise, you might as well build a special purpose little 
> box with counters & A/Ds, and an interface)
I've had too many problems with cheap (onboard) soundcards in the past, 
even when using them for their intended purpose, so I would not advice 
to use them for anything quantitative.

But if you *really* want to syncronize inexpensive soundcards, it's 
rather trivial, see for example
   http://quicktoots.linuxaudio.org/toots/el-cheapo/ .

Just buy a few dozens for a EUR/$ each and hunt down the ones with 
identical oscillator frequencies ;-). But don't expect miracles, you end 
up with a few synchronized cards that only happen to not skip samples 
with respect to each other. Compared to decent signal input they are 
still cheap cards, and hog the CPU for their individual servicing.

    Chris






More information about the time-nuts mailing list