Re: Variac Safety

[John Denker <jsd@MONMOUTH.COM>]

At 09:25 PM 11/30/00 -0500, Chuck Britton wrote:
> Old, two wire corded variacs are (CAN) be killers.

That is a bold statement.

At 11:24 AM 12/1/00 -0500, Ludwik Kowalski wrote:

> Please correct me if I am wrong in identifying what can possibly
> happen or not happen.

Let me add my voice to Ludwik's question.  I, too, am mystfied by most of
what has been said in this "variac" thread.

In particular, as far as I can tell, variac safety is comparable to
extension-cord safety.  If you abuse a variac in certain ways, you will
have trouble, but abusing an extension cord in the same ways leads to
comparable trouble.

1) A variac is marginally more dangerous than an extension cord, because it
can marginally step up the voltage.  However I don't see that this fact by
itself would support the bold condemnations we've been hearing.

Ludwik continued:
> 1) I suppose that the grounded input wire is connected to the
> body of the variac only and that the remaining four wires (input
> and output) are floating. In fact, one of the output wires is always
> at the same low potential as one of the input wires. Another output
> wire is at a higher potential, as chosen, for example, between 10
> and 150 V.
>
> 2) Suppose the variac is used to control light from a metallic lamp.

I'm happy that Ludwik's variac is connected that way.

Note that an _unpolarized_ two-prong input cord on the variac would lead to
a different conclusion:  one could easily have _both_ of the two output
wires being greatly different from ground.

Why is that significant at all?  Because fixtures such as table lamps have
polarized plugs, and the low side of the line is connected to the large
threaded connector that contacts the vastly-more-accessible part of the
bulb.  Consequently, if you slip while screwing a bulb into a turned-on
socket, you contact only a few volts, not 100+ volts.

Connecting a lamp via an extension cord that has an unpolarized plug runs
the risk of having the more-accessible node be hot.  Connecting a lamp via
a variac with a similar unpolarized plug runs the same risk.

Household appliances other than screw-in lamp fixtures usually do not have
polarized cords at all.  And in the US two-phase system, 220-volt
appliances often don't have a neutral wire:  just hot(black), hot(red), and
ground(green).  So I'm not particularly scandalized that the variac could,
if operated suboptimally, have two hot outputs.

At 11:48 PM 11/30/00 -0800, Leigh Palmer wrote:
> I always want to do some fiddling with the output.

I see danger in fiddling with the output of a variac -- but AFAICT the
danger comes from the fiddling, not from the variac.

Again: fiddling with the output of an extension cord incurs comparable dangers.

Leigh continued:

> A Variac can be made safer to use by connecting it to an isolation
> transformer (1:1) which, in turn, is connected to the mains.

Again:  The same could be said of an extension cord.

At 04:20 PM 11/30/00 -0800, Leigh Palmer wrote:
> I did arcweld a nice hole in the grounding
> lug of a signal generator

The variac was not the proximate cause of this problem.  The hot output(s)
of the variac should not have been anywhere near that grounding lug.


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

Bottom line:  A little knowledge is a dangerous thing.

True fact:  A transformer _may_ be hooked up so as to provide isolation
which gives protection against a certain class of circuit faults.

True fact:  An isolation transformer does not provide protection against
all faults.

True fact:  A variac is an autotransformer.  It does not provide any
isolation.  But neither does an extension cord.