<div dir="ltr">The last diagram is most instructive.<div><br></div><div>If you turn off one of the PCs, you increase the load across the remaining PCs on that side of the N</div><div><br></div><div>One of the problems to be dealt with up at Sullys place will be upgrading N.</div><div><br></div><div>If most of your load is 240, you dont use the N much and its not a problem.</div><div>Also older buildings get away with it.</div><div><br></div><div>The problem comes with modern code and electrical circuit demand.</div><div>Current code calls for an outlet every 6ft around the perimeter of a room, and every 4ft in bath, kitchen, or wet areas.</div><div>Its tempting to pop another breaker in the panel (which is probably already at capacity) and run those extra circuits.</div><div>Now that you have the ability to draw a lot more at 120, you start overloading the N.</div><div><br></div><div>Imagine a house built in the 30s, knob and tube wiring in the attic, with the N running right down the ridge of the house.</div><div>The wiring is gorgeous.</div><div>Ove course additional circuits have been added, each time tapping into that single N for the whole house.</div><div>These wires all lead to an old fuse panel.</div><div><br></div><div>At some point in time, the single meter box that connected to the fuse box via conduit was replaced.<br><div>The new panel Includes breakers and the new meter.<br></div><div>A run of conduit connects this box to the old fuse box.</div><div>The fuses in the old fuse bax have been removed and there is a rats nest of wire nuts connecting wires that once connected to </div><div>fuses to wires leading out to the bew breaker panel.</div><div>The new panel is a 100 amp panel.</div><div>There are no 240 circuits in the panel.</div><div>The N connecting the boxes is a #12 wire.</div><div><br></div><div>What could go wrong? (I can hear you scream from here)</div><div><br></div><div>Once again, in an AC circuit, the other end of the circuit will always be 180 from one end to the other.</div><div>This is IN PHASE!</div><div><br></div><div>An often used labeling scheme is H0 & H1.</div><div>When we refer to phases, we speak of phases as seen in the panel.</div><div>H0 & H1 are IN phase at the panel because one is the return of the other.</div><div><br></div><div>It IS OK to refer to the other hot as "Other leg". Thats quite acceptable.</div><div><br></div><div>I would say that this "phase" concept is something that escapes more than hald the population.</div><div>AC is weird stuff, very weird.</div><div>If you dont understand it, Im not going to call you stupid. It can be a pain to wrap your head around.</div><div>I WILL tag you on it for using the whord phase in the wrong place, but i wont call you a dummy for it.</div><div>Phases are not too hard to get unless you throw in that N and THATS where it starts getting wierd.</div><div><br></div><div>There is also somehting called a "stinger" lead which happens in "Open Delta" wiring and Im going to leave that out of the discussion.</div><div>I will point out that 277vac is NOT stinger, however, and leave it at that.</div><div><br></div><div>The economy was starting to come back in 2010. Private IP space had bought us time, but this time we were REALLY out of IPv4 addresses.</div><div>I had taken a 3 month job with a startup networking company that was embracing IPv4 at a time when ATT was hoping it would go away.</div><div>Asia had totally burt its last IP block and the rest of the world wasnt far behind.</div><div>Being an asian company, the aisian telecom companies discovered out IPv6 & app level switches.</div><div>(This is where I learned to eat fruit on the bottom yogurt with chopsticks.)</div><div>We went into that hockey stick curve just as I started there.</div><div><br></div><div>I ended up staying 2 years, as the demand for computing resources climbed faster than we could satisfy it.</div><div>It was an interesting time that included literally exploding chiller units, electrical panels hot enough to burn you.</div><div><br></div><div> We had already expended into 2 more buildings in the complex to spread out our electrical load.</div><div>I had installed thermal monitoring in all the labs, and wrote scripts to alert me about alarms, and shot the data to rrdtool.</div><div><br></div><div>It was my birthday in 2011. I came in the main building and the "B" side was dark.<br></div><div>Main building was 2 offices next door to either with a couple doors between the suites.</div><div>I had noticed that my alerts had stopped at 4am.</div><div>I fumbled my way to my cube and measured the outlet voltege. 32 vac. How quaint!</div><div><br></div><div>Grabbing a screw driver, I headed to the electrical room.</div><div>Voltages in the breaker panel for the B side were also screwy.</div><div><br></div><div>Then I opened the 480vac junction panel (To the horror of the VP of OPS)</div><div><br></div><div>OK Phase-A to Phase-B 480, cool. Phase-B to Phase-C....</div><div>WTF?!?!?!?! Where is the bussbar to Phase-C ???????</div><div><br></div><div>Oh, I see.... That must be the slag in the bottom of the cabinet....</div><div><br></div><div>I had Cupertino Electric on speed dial on my Crackberry along with the chiller service.</div><div><br></div><div>This was a 400 amp bussbar, by the way.</div><div>(Did you face palm yet?)</div><div><br></div><div><div>By the way, this was not the first time that we had run part of the building off a portable gen set.</div><div></div></div><div><br></div><div>That was the same week that I discovered the rotten roof truss under the 25 ton chiller, RIGHT ABOVE MY F***ING CUBE!</div><div>Thats quite another story.</div><div><br></div></div><div>We eventually got a new building with a huge engineerting lab, enabling the consolidation of the company back into a single building.</div><div>It had a 3 phase 480 feed to the engineering lab at 4000 amps.<br></div><div><br></div><div><br></div><div><br></div><div><br></div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Fri, Apr 17, 2020 at 1:23 AM Michael Paoli <<a href="mailto:Michael.Paoli@cal.berkeley.edu">Michael.Paoli@cal.berkeley.edu</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">> From: "Rick Moen" <<a href="mailto:rick@linuxmafia.com" target="_blank">rick@linuxmafia.com</a>><br>
> Subject: [conspire] 1999 called, and wants its backhoe returned<br>
> Date: Mon, 13 Apr 2020 13:23:54 -0700<br>
<br>
> pipe) along Monte Rosa. In front of 815 Monte Rosa they hit a PG&E<br>
> secondary and shorted one phase to the uninsulated neutral (according to<br>
> PG&E). The result was that the two lines (beyond the transformer)<br>
> coming to our house that should have been +120 and -120 became +85 and<br>
> -155. Electronic stuff on the 155v phase (without a surge protector)<br>
<br>
And might've mentioned it before, once-upon-a-time. Many moons<br>
ago, I dealt with a kind'a similar situation at work.<br>
We'd recently moved into a previously occupied building.<br>
The (our) business included computers (mostly selling PC "clones"),<br>
and servicing thereof, accessories, etc. The previous tenant was a<br>
more industrial customer (some printing operation, if I recall correctly).<br>
Anyway, the room that was used for computer repair, had been repurposed<br>
from whatever it was before. It had workbench-like surfaces all around,<br>
with (nominally) 120VAC outlets a plenty readily available all around the<br>
workbench type working surfaces.<br>
<br>
Well, one day, things went badly. A few things semi-randomly fried on<br>
circuit(s) on there, when there was (otherwise) no particularly good<br>
reason for 'em to do so. Our tenancy being new, I didn't know the<br>
relevant breaker (or maybe even where they were) for the applicable<br>
circuits - and all else in the building seemed perfectly fine (was<br>
likely just one or two circuits or so for that entire room).<br>
So I gave the directive to the techs, to unplug or otherwise disconnect<br>
everything from all circuits in the room as fast as humanly possible.<br>
And so they did. Then we went about diagnosing the problem - as it<br>
still appeared limited to just the circuits in that room. Doing a bit<br>
of checking ... two circuits. But alas, not hangin' very close to 120VAC.<br>
But if one measured between the two hots from the two circuits, a nice<br>
reliable right around 240VAC between 'em. So, me thinks (and was later<br>
confirmed) - flakey neutral ... basically 2 circuits, 180 degrees<br>
out-of-phase (or call it center tapped one phase - for the most part<br>
effectively the same for the end circuits).<br>
<br>
And sometimes this is (or used to be?) a semi-common problem. In more<br>
industrial environments, there might be simple 2-phase like that<br>
(180 degrees out-of-phase), or might be more (3-phase also being relatively<br>
common). In many industrial settings, it is presumed, and often the case,<br>
that the loads across each phase are very well balanced - not only<br>
power/current, but phase relationships, etc. Essentially if they're<br>
all equal loads, and well shaped sinusoidal loads - even if V/A shifted<br>
a bit - so long as they're all shifted the same bit, a "convenient"<br>
interesting bit happens. In such case, the current on the neutral, in<br>
that nominal case, is very much zero (notwithstanding some noise and<br>
imbalances). Still pretty close to zero - much less than the hot legs,<br>
anyway. So, what - at least was - common practice, the neutral would be<br>
somewhat undersized, relative to the hots - as it would be expected to<br>
"always" carry less currents than the hots. I might be mistaken, but I<br>
think older electrical codes (and/or practice) actually allowed that /<br>
and/or was relatively commonly done. Well, that's sort-kind'a all<br>
fine and dandy ... until the loads aren't well balanced, and there's<br>
quite a bit more current on the neutral - can be as high as at least<br>
any one of the hots (or approximately that), at least in worst case<br>
unbalanced loading. Well, that could be a really bad thing if the<br>
neutral was undersized. The other related thing, is earlier well balanced<br>
load case, the neutral could've gone decades or more, without carrying<br>
hardly any current. Then the loads change (go from huge industrial<br>
motors, to dozens of switch mode power supplies for PCs - which have<br>
rather messy not-very-sinusoidal load characteristics), neutral current<br>
can go way up ... where that part of the electrical really hasn't been<br>
"exercised" like that before. So issues such as latent defects,<br>
could go from what was a hidden issue, to a bad issue and major<br>
failure. In our case, it was one of the neutral connections,<br>
which probably for decades before that hadn't shown as an issue,<br>
as it hadn't carried enough current to cause it to fail,<br>
but with heavier currents on the neutral, the not-so-great<br>
connection failed to an intermittent open ... which caused the<br>
two individual 120VAC circuits, relative to their now oft floating<br>
neutral to vary wildly anywhere from 0VAC to about 240VAC ... so some stuff<br>
got fried.<br>
<br>
Hmmm, maybe teensy diagram helps.<br>
So, let's say this is my (crude ASCII) drawing of secondary of a center<br>
tapped transformer. From center to each side, 120VAC, or 240VAC between the<br>
two end sides of the secondary:<br>
|| C------------------<br>
|| C<br>
|| C<br>
|| C<br>
|| C------------------<br>
|| C<br>
|| C<br>
|| C<br>
|| C------------------<br>
<br>
With a bunch of PCs on the 120VAC circuits, we have something roughly<br>
like this:<br>
|| C------------------+--+--+-----<br>
|| C | | |<br>
|| C PC PC PC ...<br>
|| C | | |<br>
|| C------------------+--+--+-----<br>
|| C | | |<br>
|| C PC PC PC ...<br>
|| C | | |<br>
|| C------------------+--+--+-----<br>
So far so good.<br>
Note also in the above, were it something other than PCs,<br>
or otherwise full balanced loads on each of the two legs,<br>
the neutral current would be (approximately) zero.<br>
Then the neutral goes (intermittent) open:<br>
|| C------------------+--+--+-----<br>
|| C | | |<br>
|| C PC PC PC ...<br>
|| C | | |<br>
|| C-----X X-----+--+--+-----<br>
|| C | | |<br>
|| C PC PC PC ...<br>
|| C | | |<br>
|| C------------------+--+--+-----<br>
Now things are ugly. The PCs are essentially in series across 240VAC.<br>
And to the extent their loads across the two legs aren't highly balanced,<br>
they can be getting anywhere from almost 0VAC to about 240VAC. Not good.<br>
And, that was before most power supplies were auto-ranging. Yes, (some)<br>
things fried.<br>
<br>
Anyway, that's a problem. Likewise if the neutral shorts<br>
(fully, partially, or intermittently) to one of those two hots.<br>
<br>
Anyway, depending where the fault happens, this can happen to an<br>
individual (pair of) circuit(s), entire house/building, or even a<br>
neighborhood - really depends on the scope and coverage of the<br>
circuit.<br>
<br>
<br>
_______________________________________________<br>
conspire mailing list<br>
<a href="mailto:conspire@linuxmafia.com" target="_blank">conspire@linuxmafia.com</a><br>
<a href="http://linuxmafia.com/mailman/listinfo/conspire" rel="noreferrer" target="_blank">http://linuxmafia.com/mailman/listinfo/conspire</a><br>
</blockquote></div><br clear="all"><div><br></div>-- <br><div dir="ltr" class="gmail_signature"><br>R "Texx" Woodworth<br>Sysadmin, E-Postmaster, IT Molewhacker<br>"Face down, 9 edge 1st, roadkill on the information superdata highway..."<br></div>