On 1/6/2018 1:21 AM, wrote:
On Fri, 5 Jan 2018 16:32:19 -0500, "Mr. Luddite"
wrote:

On 1/5/2018 2:06 PM, wrote:
On Fri, 5 Jan 2018 12:28:51 -0500, "Mr. Luddite"
wrote:

On 1/5/2018 11:37 AM, wrote:
On Fri, 5 Jan 2018 05:34:25 -0500, "Mr. Luddite"
wrote:

I agree the harmonic problem is just on wye but that is the most
common configuration, 208 or 480.
I think the problem first showed up on 480/277 wye systems where they
had a building full of 277v electronic ballasts.
Then it started showing up in those 208 systems that were feeding 120v
lines in cubicles through the internal wiring with a 5 wire feeder as
PCs started replacing terminals with ferro power supplies and space
heaters.

Delta is always funny stuff, depending on where or even if you land
the ground. Corner grounded is probably the easiest to confuse the
novice. It will look just like single phase until you get your meter
out. (2 pole breakers, white wires etc)


Not familiar with that. The majority of the systems we built were
designed to run on 480v, 3 phase. All the main breakers, fused
disconnects, motor starters, etc. were three pole. Ground was run
separately. We used a dedicated control transformer in each power
cabinet to generate whatever the low voltage control wiring was ... 120v
in the early days but later 24 volt. IIRC, the control transformer on
the 120v systems was a 480v/240v step down with a center tap, providing
a neutral.




Did you ground your 480 delta at all? If so where did the ground land?

I understand your control circuit voltage can be 120 if you want. That
is just a class 1 control circuit. Most people are familiar with class
2 but they are both controlled by the same article in the NEC.
Class 1 just looks like regular line voltage circuits because it is
not voltage or current limited like class 2 and 3.
The center tapped 240 delta is commonly called "red leg" or "wild
leg"because the corner between the center tapped windings will be 208v
to ground. "Red" is a misnomer tho because the NEC requires the wire
to be orange. That is very common in places where they want discount 3
phase and have a significant amount of single phase loads like those
small industrial bays. The PoCo can do it with 2 transformers,
generally the wild leg will be on a much smaller one. One "winding" is
actually open. hence another name, "Delta Veep".

http://gfretwell.com/electrical/red%...ansformers.jpg

99.99% of the time, if you see 3 transformers on the pole, it will be
wye. but I have seen one place in Key West where they had red leg
delta with 3 transformers. The only tip off was one was bigger than
the other two and it was confirmed by analysing the wiring.
http://gfretwell.com/electrical/Tran...%20_breath.jpg

Yeah my wife thinks I am crazy too, taking pictures of transformers.



Ground for the 480 Delta system service was run separately back to the
physical (metallic) ground at the panel, usually via a 6 awg wire.
Ground was not taken from any of the legs of the 3 phase Delta the way
you have described.

Many of the systems we built also had a RFI ground consisting of wide,
copper flashing to two, 8' copper rods driven through the floor and into
the ground. The rods were about 10-15 feet apart and we tried to get
close to 1 ohm resistance between them. This often required a copper
sulfate solution to be poured into the rod holes.

The reason for the exotic grounding had nothing to do with safety. Many
of the systems utilized a RF transmitter running at 13 Mhz. The load
for the RF transmitter was a plasma (ionized partial pressure gas)
generated within the vacuum chamber.

We had specially designed "matchboxes" that allowed load impedance
matching of the 50 ohm transmitter output to the very low impedance of
the plasma. Other systems used an electron beam operating at 10,000
volts. A stream of electrons are emitted from a filament and focused
magnetically onto various metals or dielectrics that were vaporized by
the beam and deposited on optics in very carefully controlled thicknesses.

Anyway, there were often some arcs and sparks within the chamber with
either of these deposition methods that would raise hell with some of
the very sensitive measurement instruments. The elaborate grounding,
using a wide conductor (flashing) is much better at RF and EMI quenching
of the arcs. A simple, round ground wire has too much inductive
reactance. It's only purpose on these systems was to serve as a safety
ground. Sometimes we'd have to chase our tail for a while however
because the use of both grounding systems sometimes generated a ground
loop which only magnified the RFI problem.

So you were running ungrounded delta. That is fairly rare and usually
only for systems that are very intolerant of faults, like a glass
plant where a power failure makes the whole place pretty much trash.
I assume you had ground fault indicators since the first ground fault
is "free". I have never actually seen ungrounded delta here. The other
option is impedance grounding that grounds the system via a resistor,
just to stabilize the voltage, not to provide any fault protection.

You are right that there is no exception that allows the non current
carrying parts of the system to be ungrounded. They are really trying
to change that terminology to "bonded" to avoid confusion with the
"grounded conductor" that we normally call the neutral and the
"grounding electrode" that is your physical connection with earth.
Article 250 may be the most misunderstood article in the whole NEC,
hence the one I spent the most time studying. They are trying to
redefine some of the terms to make it easier to understand because
"ground" is such an all encompassing word that it misses the intent
about half the time and when you actually start testing, it isn't even
the same voltage from one place to another when you actually test with
stakes in the dirt.




I don't think the configuration I described is all that rare. In fact
it's pretty common for industrial machine tool wiring for equipment that
requires significant power. There are literally thousands of systems
similar to the ones my company built, designed manufactured and
installed by other companies in the USA and world wide, going back to
the 1950's.



I don't know. Like I said I have never seen ungrounded delta but I
know the guys to ask.
I am still wondering if it was impedance grounded or corner grounded.
They do not need to bring the grounded conductor out to distribution
equipment, in fact we never did in computer rooms. It stopped at the
service disconnect.

Ah .. I think I see the confusion. Yes, the secondary of a 3 wire, 480v
delta service will usually have one leg tied to *earth* ground but it's
not used as a current carrying conductor. The primary side of a 3 wire
service transformer is not grounded. There is also a 4 wire delta
service that includes a neutral from the service side.

You are talking the service side. I am talking the load side.

The equipment we built was powered by the 3 wire, 480v legs for power to
the various motors and heaters. A safety ground is run back to the
service panel that is tied to earth ground. It's why we used a
separate transformer to generate the single phase, 120v and 24v control
voltages used for the instruments and control switches.