Re: is x10.com dead?

[Marc F Hult <MFHult@xxxxxxxxxxxxxxxxxxxxx>]

On Wed, 01 Mar 2006 19:26:15 GMT, "Jeff Volp" <JeffVolp@xxxxxxx> wrote in
message  <rzmNf.483603$qk4.311180@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx>:

>"Marc F Hult" <MFHult@xxxxxxxxxxxxxxxxxxxxx> wrote in message
>news:1onb02h43ejn59qd3qlc7l3voo52qclla4@xxxxxxxxxx
>> On Wed, 01 Mar 2006 16:43:27 GMT, "Jeff Volp" <JeffVolp@xxxxxxx> wrote in
>> message  <PakNf.482883$qk4.116787@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx>:
>>
>> How on earth (pun intended) does being coupled to power lines that
stretch
>> from your house to the power plant afford you with "some measure of
>> protection" ?
>>
>> It is precisely because RS-485 is readily isolated using optical couplers
>> that RS-485 is an excellent and common choice for what you want to do.
>
>The ADI modules do not include opto-isolation.  I had considered using
>opto-isolated SSRs at the output of the SECU16, but the Rain8 was a simpler
>solution.  Opto-isolators are only good for a couple KV, and I'm concerned
>about high induced voltage in a relatively high impedance circuit.  That's
>what took out part of our alarm system at the other house.  The bolt itself
>hit a tree across the driveway from the house and blasted a hole in the
>ground near the trunk.  The powerlines were underground, and nothing at
>that house was ever effected by transients on the AC line. A similar
>condition exists here.
>
>If a close strike does induce high voltage in the switched 24V line, that
>could take out the irrigation controller.  But the powerline and whole
>house surge protector should dissipate the induced energy better than that
>485 link back to the Ocelot.
>
>Like I said earlier, all bets are off if there is a direct strike.
>
>Jeff


Interesting -- a whole 'nuther topic. Our problems have been a differential
between the power line and ground. Whether that was exacerbated by a
code-compliant but inadequate grounding rod remains an open question. These
problems are all different in detail and yours does seem different than
mine.

Your point about conventional optoisolators providing only several KV of
protection is well taken and important. A while back I researched
high-isolation opto-couplers for DMX512 ( which uses RS-485) and purchased
some IC's for constructing an improved isolation system. They are still in
the parts bin so they aren't actually helping yet.

That's also part of why my PC 's are now connected by fiber which provides
nominally 'infinite' isolation rather than by CAT5. Twice I had the ethernet
ports blown out. The audio system is 'protected' by an air-gap relay in
series with an SSR http://www.econtrol.org/power_conditioning.htm.  The HA
and web servers are on an isolated chassis with a separate supply for DC-DC
converters. Seems to have "kept the wild pink elephants away" for a while
.... I do feel better for having done _something_.

Marc
Marc_F_Hult
www.ECOntrol.org