3b2's and elec. codes

[Lerxt]

Summary: software-controlled power down may violate electrical codes
Line eater: yes
Apology: My apologies if you are seeing this for the second time.  I tried
to post it yesterday, but it never appeared in the newsgroup, so my gateway
machine appears to have thrown it on the floor, but I'm not certain.

I have been following the discussion about the 3b2's which switch their own
power on and off via software-controlled relays or triacs.  (Did anyone in
netland ever see the Apple Lisas that also did this?  When you turned the
power off, a box appeared on the screen that said "The Lisa is preparing to
shut down", followed a couple of minutes later by another box saying "The
Lisa is shutting down", after which the screen would slowly and dramatically
fade to black, and then a relay would click and the light in the power switch
would go out.)  I got to thinking about some things that I picked up about
electrical codes from my father (that's what happens when you have an EE for
a father :-)), and this occurred to me: those 3b2's and Lisas might be
violating the electrical codes in a number of cities.

Here's the deal: the National Electrical Code, which a number of cities and
states have adopted as their local code, specifies that any electrical device
must have at least *two* mechanisms for shutting off the power (switch,
circuit breaker, plug that can be pulled from an outlet, etc.).  These
cutoffs must meet these criteria:

1.  A cutoff must be within 1000 feet of the device that it controls.

2.  The cutoff must be visible from the device and vice versa.

3.  The cutoff must be "readily accessible".  The phrase is open to inter-
pretation, but generally it means that the cutoff must be clearly labeled,
in a freely accessible area (i.e., not in a locked room), the cutoff must
not be locked, taped, or screwed down, and must not be inside the cabinet
of the device that it controls.

4.  The cutoff must be "reliable", which means that it is capable of
interrupting the load and that the probability of the cutoff failing to
work when activated is low.

There are some exceptions to the two-cutoff requirement for low-amperage
120-volt devices; the 3b2 may qualify as a low-amperage device which
admittedly makes this argument moot.  Still, most minis (Vax 11/7xx and 86xx,
for example) do not qualify and are subject to the two-cutoff provision. 

Now, on most computers, the power switch on the front panel is one cutoff,
and either the line cord, a back-panel circuit breaker, or a nearby branch
circuit breaker is the second.  However, I don't believe that the power
switch on the 3b2 meets the criteria for a cutoff, since software can delay
the power interruption or even disable it altogether.  If there is no readily
accessible circuit breaker on the machine (breakers inside the cabinet don't
count), then there is only one cutoff: the line cord.  Big deal, you say; all
you have to do is pull the plug out.  Well, what if the line cord overheats,
or the insulation cracks? Do you know where the circuit breaker for that
outlet is?  Can you even get to it, or is it in a locked electrical area?  At
our site, we have two upstairs machine rooms whose breakers are in another
area.  Since the plugs are mostly in either the ceiling or the floor, we had
to install "panic buttons" that shut off power to the room to meet the two-
disconnect requirement.

Is the 3b2 UL-certified, and does it have a circuit breaker on the back
panel?  If not, then I'd be a bit surprised that it would pass UL inspection.
Please not that this is not a flame against the 3b2 in general, just against
this particular feature.

The moral of the story is: software-controlled power disconnect is a code
loser.  If you're designing a computer, and you're thinking about putting
this feature in, DON'T!

---
It's been said by many a wise philosopher that when you die and your soul
goes to its final resting place, it has to make a connection in Atlanta.

Dave Cornutt, Gould Computer Systems, Ft. Lauderdale, FL
UUCP:  ...{sun,pur-ee,brl-bmd}!gould!dcornutt
 or ...!ucf-cs!novavax!houligan!dcornutt
ARPA: wait a minute, I've almost got it...

"The opinions expressed herein are not necessarily those of my employer,
not necessarily mine, and probably not necessary."