Dave Tweed and I were finally granted US patent 7,411,378 last week on our
power factor correction circuit.  The basic concept is to use digital
computation to know what currents are flowing in various places so that
there is no need to measure them.  Our circuit is controlled digitally, and
the controller only receives measurements of the voltages, not any currents.
There are various ways to measure current, but it's always a bit
inconvenient.  Our scheme gets around this requirement.

The first place this concept is applied is in controlling the AC input
current such that it stays proportional to the input voltage.  This is the
power factor correction (PFC) function.  The details are in the patent, of
course, but since it's been near lawyers it's not the easiest to follow.
The writeup at http://www.embedinc.com/pfc explains this application of the
concept in a more friendly manner.

The other application that the writeup above does not go into is regulating
the output voltage of the PFC controller.  Since the AC input current must
be kept proportional to the AC input voltage, the output voltage can't be
tightly controlled like it would be in other types of switching power
supplies.  The traditional approach (which is also used in the write at the
link above) filters the output voltage feedback signal to try to remove the
frequecy components due to the necessary AC input current ripple, but
otherwise pass as much as possible so that changes in the load can be
accomodated as quickly as possible.  This is of course a tradeoff with no
ideal solution.  There will be some response to the ripple caused by PFC
(thereby defeating a portion of the PFC), and the response to load changes
will be slowed.

Our solution is to use voltage measurements and computing power to determine
what the load is doing independent of the ripple caused by the PFC.  One way
to think about it is that since we know what charge is being dumped onto the
output capacitor, the ripple caused by the AC input waveform can be computed
and thereby explicitly removed from the feedback signal.  The remaining
feedback signal is due solely to load variations, and can be reacted to
quickly without defeating the PFC.  This allows for tighter output voltage
control and/or a smaller output capacitor.  One explicit equation for this
is given in the patent, although others are possible.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist