Surge suppressors (not wiring)
mathog at mendel.bio.caltech.edu
Mon Nov 4 09:26:50 PST 2002
> Jim Lux wrote:
> > At 10:13 AM 11/1/2002 -0800, David Mathog wrote:
> >Are my selection criteria unreasonable? All I want in a strip is:
> >6-8 sockets
> >big MOVs (high joule rating - longer expected lifetime)
> >metal case with some holes in it (for zeroU mounting,
> > I'd accept a plastic case if there was some way to
> > mount it in the rack that didn't involve glue or duct tape)
> >disconnect sockets on MOV failure
> >15/20A breaker (fuse would be ok).
> There's a lot more to surge suppression than MOVs and
> Joule ratings...
> There was an article in IEEE Spectrum a few years back on
> this (at least 5 years).
Absolutely there's more to it than MOVs. The problem is
that you basically can't find out much about the guts of
most commercial surge suppressors, nor can you find any
relevant test data. The only recent review I've been able
to locate was in Consumer Reports (1/2000, p49). There
they tested a bunch of strip suppressors for home use.
Unfortunately, these strips are not optimal for use in a lab
because they only turn out a "protected" light when the
protection is gone, rather than disconnecting the load,
which is orders of magnitude better. That is, when some
odd bit of home grown equipment two doors down partially
fails and starts sending 1 spike per second into the
building power I'd really rather have the power strip on the
first floor detach the load when it fails on the 157th spike,
rather than losing all of the nodes plugged into it on the
the 158th. Besides, most of the time one can't even easily
see the "protected" light on zeroU mounted surge strips.
Anyway, the top rated model was a TrippLite Isotel 6 Ultra,
which is essentially the same as an Isobar 6 Ultra plus modem/FAX
protection. This was rated excellent both for performance
Good luck though finding a review on something like an EDCO AC-RACK:
Which is pretty close to my desired products specs (except
joules, which seems low at 900) and does have the "remove
load on suppression failure" feature. So on paper it looks good.
Be a bit more comforting if some independent group had tested
mathog at caltech.edu
Manager, Sequence Analysis Facility, Biology Division, Caltech
More information about the Beowulf