Howdy,

You're correct. The problem is mainly the result
of many single phase capacitor input power supplies
distributed across a three phase system.

In a capacitor input power supply, the line current
takes the form of narrow spikes, because the rectifiers
only conduct when the line voltage is higher than the
capacitor voltage. Many SMPS have a capacitor input
bulk supply at the front end. So the power factor is
low and there is much harmonic distortion which
includes triplens 3rd 6th 9th... harmonics. also called
zero sequence harmonics (they do not rotate.) As they
add in the neutral of a three-phase four wire system
difference between their half cycles is a DC component.
You can see that half-wave input would be the worst
as Paul mentioned.

The solution reduces THD and increases power factor.
For single phase input power supplies is either a
large inductor which causes the current to look more
like a rectangle. Or active PFC which is usually a
boost converter with no input capacitor and a control
circuit that causes the envelope of the current
pulses follows the shape of the line voltage. Most
often the active PFC is used but in low power supplies
I've seen the big inductor approach taken. There are
also other approaches like "valley fill" which force
current to be drawn when then line is less than
that needed to charge the capacitor. I don't see
these used much.

Before PFC was common this was a huge problem in places
with many crappy PC and VDT terminal power supplies (like
big office buildings full of VT100's.) Sometimes the
neutral conductor which was sized smaller than the phases
because they expect little imbalance would burn up.
Sometimes the DC component would saturate big transformers
and they would fry.

Triplens can be blocked by ungrounded-wye or delta
transformer connections which keeps them out of the rest
of the facility but they still do their dirty work on
the secondary side. So it's better to not draw distorted
line current. With three phase input power supplies
a simple boost converter without the PFC control circuit
improve the PF to about 0.96 and reduce the THD to eh,
maybe 25% or less. The current in phase leg will be
rectangular.

This explanation could be too simple. And as always
some of it might be wrong! But does it answer your
question?


73
Grumpy



"Henry Kolesnik" wrote in
:

Hey Ken & Grumpy
I started a new thread from the 7.5 VAC xfmr, because the subject has
drifted. You guys mentioned a DC component in pole pigs. Sometime ago
I heard that if an AC circuit had too many switching power supplies on
it, that would cause the power company all kinds of problems. What were
or are the problems and what was the evidence and what did the power
companies do to solve the problem?