On Monday, March 22, 2021 at 5:53:13 PM UTC-4, Clare Snyder wrote:
On Sun, 21 Mar 2021 16:22:03 -0700 (PDT), DerbyDad03
wrote:
On Sunday, March 21, 2021 at 4:33:18 PM UTC-4, J. Clarke wrote:
Please forgive me--this is a bit of a rant.

I went to put a new fixture at the top of the stairs--the old one is
physically too small to hold anything bigger than a 60w incandescent
(by physically I mean you can't put the globe on if anything bigger is
in it, and that includes CF and LED that are larger than 60W
equivalent and in that location a 60 just isn't enough light.

Well, went to kill the power to the circuit and discovered chaos.

First:
Turned the switch off
Checked socket with a voltage sensor
Still voltage on the circuit
Par for the course in this house
Switch is in the neutral leg
Add to list of stuff to fix.

Next:
Screwed adapter into socket and plugged in tracer
Traced signal in breaker panel
Not one, but _two_ breakers showed signal
Turned both off
Went back upstairs
Checked for voltage again
No voltage--good
Flipped switch--checked again
Voltage--not good
Back to the panel
Identify third breaker, turn off
Now no voltage

Replaced fixture, turned breakers back on, everything works, I didn't
die.

But now I have the real mystery--how is this effing light managing to
be connected to not one, but _three_ separate branch circuits?


It's possible that you are dealing with an Edison circuit, also known as
a "shared neutral" or "multi-wire branch circuit". Granted, Edison circuits
usually only involve 2 breakers, but I think I could imagine one with 3.

In a typical Edison circuit a cable with 2 hots and a single neutral has the
2 hots tied to separate breakers. That 3 wire cable runs to a junction box
from which a pair of 2-wire (hot and neutral) cables emerge. Those 2 circuits
share the single source neutral from the 3 wire cable.

If you only turn off one breaker, you will often get some voltage that bleeds
back through the neutral at the supposed-to-be-dead fixture. If you used a
meter as opposed to a voltage sensor, I'll bet that it's not the full 120 VAC.

Now - although this is not something that I have ever seen - expand that to
a 4 wire source cable (3 hots and a single neutral) to the first junction box
and you've brought that 3rd breaker into the picture.

I don't know if a "super-Edison" (I made that up) would be allowed by code,
but I can certainly imagine someone grabbing some 12/4 Romex and running
one.


Could only occur on a 3 phase circuit - and BY LAW all edison circuits
MUST use a "tied breaker" - and in a fused panel a "ganged pull-out"
which makes it IMPOSSIBLE to shut off power on one side only.

By *law* or by *code*?

I did mention both code & 3 phase circuits earlier. Of course, we don't know
if code was followed.


Using a "tracer" is fraught with possibilities - including "induced
cross-talk"

As noted earlier, a meter reading would certainly tell us more.

The REAL test is if the light functions with any ONE of the 3
circuits enabled singly - or only with one of the three. Are all of
the 3 on one side of the panel? If so check if all 3 neutrals are tied
together - in which case they may all be connected to each other, but
not properly connected to the incoming neutral

He also should confirm that the neutral is actually switched or if it's just a
unmarked white conductor used as part of a switch leg. Unmarked would
be a code violation, at least in the US.