In message , Steve
writes
Andrew Gabriel wrote:

Power factor correction capacitors only work to correct lagging
current. Low power factor caused by modern electronic appliances
does not fall into this category (there's no phase shift), and
can't be corrected by factor correction capacitors.


I'm a bit confused by that.

AIUI, power factor can be lagging current (inductive) or leading
current (capacitive). But there has to be a phase shift - either way -
for there to be a power factor other than unity.

Steve
In simple linear circuits you are correct.
But electronic products often have switch mode power supplies*. The
input to these is (was) a bridge rectifier feeding a capacitor straight
off the mains, so the capacitor is charged to around 340V DC (i.e. the
mains peak voltage). High frequency switching circuits then feed the
power through a transformer to the low voltage secondary circuits. But
the point for this discussion, is that the input capacitor is only
"topped up" for a brief period near the peak of the mains voltage cycle.
Hence quite a large, but narrow (~2ms) current peak from the mains. This
non-sinusoidal current has a higher than expected RMS level, and also
contains a lot of harmonics - which are delivering no useful power.
For example, a switched mode power supply drawing 100W of real power may
well draw 0.75A rms from the mains. 0.75*240 = 180VA, and power factor
can also be defined as real power (watts) divided by apparent power
(VA). In this case 100/180, or 0.55 power factor.

Now all these harmonic currents were starting to adversely affect power
distribution networks, so legislation** appeared about 10 years ago to
restrict the harmonic content of the power drawn by higher power
products (computers, TVs, anything electronic above 75W basically).
Hence switch-mode power units with "power factor correction". For lower
power stuff, (100W or so) a large inductor in the input can be used to
spread the width of the current pulse and thus reduce its harmonic
content. Otherwise a variety of electronic switching circuits can be
used to force the product to draw a near sine wave of current from the
mains. Power factors of 0.95 are easily achievable, 0.98-0.99 is
typical.

*Even wallwarts and other 50Hz transformered products exhibit the same
problem - there's still a diode-capacitor on the secondary of the
transformer for the DC output. But this lower power stuff falls out of
the scope of the legislation.
**Was EN60555, now EN61000-3-2.


--
steve