At 09:49 AM 2/13/2004 -0500, you wrote:
>Hi All. I'm building a little signal splitter/repeater/isolator for a
>250KBps signal (no, not DMX sadly, I'd feel more comfortable with that).
>
>For the original, I won't have any isolation between the input and
>output. My idea is to have one receiver chip feeding all four inputs of
>a quad buffer/driver chip, and each output then feeds its own
>transmitter chip. Question is, can I tie all four inputs together like
>that? I don't think there's a problem. Second, would a 74lsXX (don't
>know the number, I'm not near a catalog at the moment) handle the
>250Kbps stream ok?
>
>For the slightly more advanced version, I want to have isolation between
>the input and each output to avoid ground loops in larger systems. My
>thought is to add an optoisolator to each output of the buffer chip in
>the above circuit. First off, would a garden variety 4N25 (from memory,
>might be a wrong part number) or similar work at those speeds?

Not a hope in hell. Even working into an unrealistic 100 ohm load
with their lousy 20% CTR, and preventing saturation, they are not
guaranteed to be fast enough (that's 500uA of output signal in a reasonable
design, so you'd need to add a comparator to the output to detect
the 50mV signal- not nice). The cheapest parts that will work with
logic-level outputs are probably the industry standard 6N137 and similar.
You can get duals too, in a DIP-8.

>Second,
>to truly avoid loops, each transmitter chip would have to be on its own
>power supply to avoid connecting the grounds of all the chips together,
>correct? Is there a semi-easy way to do this? At the moment the only
>thing I can think of are independant 120V-5V switching power supplies
>for each chip, or some sort of DC-DC convertor for each chip. This will
>get expensive and large very quickly. Is there any other way?

No, there isn't. Transformers (mains frequency or DC-DC converter
high frequency) can have multiple secondaries, but take
care wrt interwinding capacitance and voltage rating (often not even
specified).

The only alternative techniques for getting power across galvanic
isolation (piezo transformers, mechanical motion, photovoltaics etc.)
are generally unsuitable.

Best regards,

Spehro Pefhany --"it's the network..."            "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.