Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Re: electrical safety



At 05:20 PM 12/1/00 -0500, Michael Edmiston wrote in part:

> John Denker's comments about extension cords is valid.

OK :-)

we know the extension cord is plugged in (unsafe) or it is not plugged in
(safe).

OK.

But the point he is missing is .... With the old variacs we look at it
and see the dial is turned down and/or the switch is turned off and we
think we are safe, ...

There are multiple dangerous fallacies there, as will be discussed below.

> but we are not.

That's for sure.

Your variac-powered equipment has a 50-50 chance of being "live" with
respect to ground (full line voltage) even when the dial is turned to zero
if you have a non-polarized variac cord/plug.

Quite true.

Worse, this could still happen even if we turn the switch off instead of
just turning the dial down, because some old-fashioned variacs have
single-pole switches instead of double-pole switches.

Indeed, if there is a single-pole switch, it is almost certainly not in the
leg that feeds the "zero" tap on the autotransformer. So throwing the
switch and setting the dial to zero are equivalent; neither is a backup
for the other.

================

Note that (once again) the variac is mostly analogous to an extension
cord. The problem is 0% in the autotransformer, 1% in the unpolarized
plug, and 99% in the fallacious notions and dangerous procedures that were
mentioned above. Specifically:

If I open the chassis of, say, my stereo amplifier, plug it in
(unpolarized), and turn it OFF, there is at least (!) a 50-50 chance that
there is 110+ volts running around in there. (Actually for my stereo there
is a 100% chance, for reasons we need not discuss right now.)

The fact that the dial, the switch, and the unpolarized plug are on the
variac rather than on the device you are messing with is a distraction, a
red herring.

So, when fiddling with electrical equipment, the rules are:
1) If you value your life, UNPLUG the equipment before grabbing any wires.
2) Don't trust any OFF switch to be "off" enough for repair work. It is
only meant to stop the normal functioning of the device.
3) Especially don't trust any OFF switch if there is an unpolarized plug
anywhere in the picture.
4) For heaven's sake don't trust the OFF switch on device "A" (the
variac) to de-energize device "B" (whatever you are working on).
5a) Don't trust the "neutral" side of a polarized socket to be wired to
the neutral side of the line. Mis-wired sockets are quite common.
5b) Even if the sockets are polarized correctly, don't assume that
"neutral" is at a negligible voltage. Unless you are colocated with the
grounding strap on the power company's transformer, Ohm's law guarantees
that there will be _some_ voltage on that wire. Several volts (with a very
low source impedance) can produce quite exciting fireworks.
5c) To express points (5a) and (5b) in another way: when grabbing a
wire, always assume that BOTH sides of the line are hot. Treat a
single-phase circuit (black : white : green) as just as dangerous as a
two-phase circuit (black : red : green). Then, if one side turns out to be
"not quite as hot" as the other side, it's a bonus -- not something you
stake your life on.
6) You don't need to worry about rules (2,3,4, and 5) if you follow rule
(1). Unplug the thing already!
7) Even after the thing is unplugged, you still have to be careful,
because there might be stored energy. I've seen systems with enough stored
energy to fry a bison. Before grabbing something, check it (and everything
near it) with a voltmeter and/or probe it with a grounded wand.
8) Normally you should keep one hand in your pocket, and should not grab
two things at once. If for some special reason you can't comply with this
rule, a whole new set of rules applies, and I don't feel like going into it
right now.
9) Don't work alone.
10) No rings, no watches, no bracelets, no necklaces, et cetera.
11) Wear sensible shoes. Barefoot is no good. Wet shoes is no good.
12) Every lab should have a "panic button" near the door that shuts off
power to all the circuits in the room. Make sure the students know about it;
if you get hurt, you need them to rescue you.
13) Be especially careful when working on equipment built by
students. It could be miswired.
14) For heaven's sake, teach your students these rules. If they don't
learn anything else this week, it will be a good week. If they aren't
advanced enough to comprehend these rules, they shouldn't be messing with
circuits of this kind.