Path: utzoo!attcan!uunet!crdgw1!rpi!batcomputer!eacj
From: eacj@batcomputer.tn.cornell.edu (Julian Vrieslander)
Newsgroups: sci.electronics
Subject: Re: Opto-Isolator Advice Needed (I could use some help, too)
Keywords: Opto-Isolator CMOS TTL
Message-ID: <9820@batcomputer.tn.cornell.edu>
Date: 1 Mar 90 21:30:36 GMT
References: <5020@scolex.sco.COM>
Reply-To: eacj@tcgould.tn.cornell.edu (Julian Vrieslander)
Organization: Cornell Theory Center, Cornell University, Ithaca NY
Lines: 36

In article <5020@scolex.sco.COM> deanr@sco.com (Dean Reece) writes:
>
>I'm building a circuit to control various functions on my audio/video
>equipment by computer.  I realize that isolating my circuitry from the
>existing A/V stuff is a really good idea, so I need some advice on the
>use of opto-isolators.
>
>So... I'm asking for your favorite Opto-Isolator circuits to do the
>above.  If you've got a known solution or a good reference, please

By coincidence, I am also starting a project that may need opto-isolators.
In my case, the goal is to send 20 digital lines (data, clock, control)
from a Macintosh II nubus card to a D/A convertor card outside the
computer.  Hopefully, the opto-isolation will reduce noise in the
analog output signal.  This is for a high performance sound synthesis
system: 25k to 100k sample rates, 16-bit sampling, 90+ dB S/N ratio.

I have seen products that have up to 4 opto-isolator channels in one
DIP package; this seems like a way to simplifiy life when 20 lines must
be transmitted.  Are there any "standard", or recommended models among
these?  Any tricks or pitfalls to know about when rates of up to 100 kHz
are being sent?  Any particularly useful Application Notes worth reading?

I would also appreciate hearing any tips for eliminating noise *without*
going to opto-isolators.  The main requirement is to reduce spurious
stuff at the low level end of the D/A convertor's dynamic range.  When we
wire up the external box containing the convertor, we will first try it
with a direct ribbon-cable connection.  Then we will look to see what sort
of clock and P.S. garbage is reaching the analog output.  If needed, we
will try resistors between grounds, taking the ribbon cable through a
ferrite common mode choke, etc.  Any other ideas?
-- 
Julian Vrieslander 
Neurobiology & Behavior, W250 Mudd Hall, Cornell University, Ithaca NY 14853    
UUCP: {cmcl2,decvax,rochester,uw-beaver}!cornell!batcomputer!eacj
INTERNET: eacj@tcgould.tn.cornell.edu     BITNET: eacj@CRNLTHRY