noise Re: [Beowulf] Re: Setting up a new Beowulf cluster
Robert G. Brown
rgb at phy.duke.edu
Thu Feb 14 06:04:33 EST 2008
On Wed, 13 Feb 2008, Jim Lux wrote:
> That's what the big Tesla Coil or quarter shrinker is for.
Oh, yeah. Let's put a great big tesla coil right in with all those
computers! Wait, I hear it now...
(...that's the sound of all those itty bitty gaps on a circuit board or
NIC arcing at the same time...;-)
OK, funny story time, sort of. Stop me if you've heard this one.
My kids in E&M get to do an extra credit project for a 1/3 of a letter
grade promotion at semester's end, and maybe a decade ago I had a
student who wanted to build a tesla coil for his project and I said,
sure, cool, go for it. So off he went and with whatever web browsers
were around and pre-google alta vista found some howto sites for
building coils, and a few weeks later ran down a neon sign transformer,
built a saltwater-aquarium-wine-bottle capacitor array, assembled a
fan-quenched spark gap, and hand-wound the coils and added a toroid on
We still had our "old" lab rooms for the intro courses -- no computers,
stained lab benches and tables a big lead sink and gas and air nozzles
in the central bench(es) up front. Imagine old wooden (oak) chassis lab
equipment in glassed cabinets around the walls, a huge beam balance with
brass weights that was probably worth a kilobuck as an antique on top of
a tall cabinet in the back, that sort of thing. So my student rolls his
creation on a big cart into this, and I and the class all gather to
Naturally, we turn off all the lights and darken the shades the better
to see the lightning. Student hooks it all up, flicks the switch on the
neon sign transformer to power it all up, and bzzzzaaappppppp -- the
spark gap starts going off like a machine gun and footlong purple
lightning starts zapping off the top toroid, impressive as all hell.
And every fluorescent light in the room goes on.
And they were turned OFF, remember. They were "on" being driven by the
radiated RF power coming off of the thing with no other source, just
like Tesla dreamed.
In addition, as I walked around the room, I noted that pretty much every
metal gap a millimeter or less was arcing. Little arcs zapping across
the fixtures in the sink, the bolts on the tables, no doubt across the
wires holding up the drop ceiling. I could imagine arcing occurring
across my teeth if I grinned just right.
After a few minutes of harmless fun and demos (which involved yours
truly taking a 100+ kV "hit" straight in through the >>glass<< of a
fluorescent tube that I inadvertently waved too near the toroid and
drawing down the fire to pass through me to ground through my rubber
soled shoes, which amused the heck out of the kids but which was NOT fun
for me) we powered it off and it went into class history as one of the
coolest projects ever.
Three years ago, a second round of students wanted to build one, and did
so using 1F caps that you can apparently now buy over the counter --
back when the first one was built I used to tell students that a 1F
capacitor would end up being the size of a bench or good sized filing
cabinet, but this is no longer true. In the meantime, all the lab rooms
were gutted and rebuilt, and each workstation has its own computer. The
entire building is now filled with computers. The computers are now all
unshielded twisted pair networked, not thinwire ethernet. If I were to
turn on a tesla coil inside the building ANYWHERE (unless it were inside
a faraday cage, of course), I'd probably blow $10,000 worth of
equipment, as a tesla coil is sort of a steady state EMP bomb or solar
flare on a table.
We demo'd this one OUTDOORS in the parking lot, figuring that the
building steel would act as enough of a cage to protect the interior,
with a surge protector inline to help keep the primary power cable from
carrying back too much of an RF harmonic onto the building wiring. I
was a bit worried about the cars nearby -- if you drive an arc at e.g.
the cap on a gasoline supply it can be a bad thing -- but cars tend to
have metal on the outside and again cage off their guts. No worries --
or at any rate none of the cars exploded or blew their starter coils.
But putting one in a server room, with all of those wires strung around
in loops and connected to electronics that really hates high voltage
even at very low current -- that's just plain funny...:-)
(P.S. -- every few years I have to explain to one student or another
that no, they are NOT permitted to build an EMP bomb for their project.
They are driven by high explosive and -- however much fun it would be --
where and how would we test it? Without, of course, bringing out the
mob with pitchforks and torches afterwards...)
Robert G. Brown Phone(cell): 1-919-280-8443
Duke University Physics Dept, Box 90305
Durham, N.C. 27708-0305
Book of Lilith Website: http://www.phy.duke.edu/~rgb/Lilith/Lilith.php
Lulu Bookstore: http://stores.lulu.com/store.php?fAcctID=877977
Beowulf mailing list, Beowulf at beowulf.org
To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf
More information about the Beowulf