Note: Descriptions are shown in the official language in which they were submitted.
POWER FACTOR I~PROVING AR}~AN&EMENT
Tecllnical Field
The invention relates generally to power supplies and particularly to
the power factor ~f power converters.
5 Back~round of the Invention
A typical power converter -- a rectifier, generally followed by a power
conditioner -- which operates off of ~.C. lines directly rectifies the voltage and
stores energy on large input capacitors. As a consequence, it draws current fromthe A.C. lines in narrow but large current pulses, thereby yielding a poor power10 factor.
The power factor (P.F.) is the preferred method of measuring the
efficiency of power passing through a point in a power distribution system. The
power factor is the ratio of the aYerage power, or true power, measured in watts,
to the apparent power, measured in volt-amperes, drawn by a circuit. I~ is
15 expressed as follows:
P.F actual input power
(input RMS voltage) x (input RMS current)
Typically, the power factor measure is applied to A.C. distribution systems in
which the voltages and currents are substantially sinusoidal, though usually not in
phase. In such systems, the power factor is simply calculated as the cosine of the
20 phase angle between the current and the voltage.
Power is distributed most efficiently when the actual power delivered
to a load equals the product of the input RMS voltage and current, i.e., when the
power factor equals 1. However typical power ~actor values for power converters
range from about .75 to less than .5.
Low power factor is compensated for by high current drawn by the
converter in order to supply sufficient power to a load. Undesirable consequences
of low power factor include (a) increased impedance losses, (b) the need for
larger-capacity and more robust A.C. power distribution system components (e.g.,circuit breakers, transformers, and wiring) that are capable of handling the power
30 converter's high RMS current demands, (c) the need for larger and more robustrectifier diodes, storage capacitors, and wiring in the power converter to handle the
power surges, and (d) greater difficulty and expense of meeting safety (e.g.
Underwriter's Laboratories) safety requirements. Not only is the cost of power
distribution increased thereby, but the cost of the power itself may be increased,
~`
.
130A~48
because some users must pay not on the basis of the actual power consumed but
on the basis of apparent power consumed. Clearly, then, it is desirable to improve
the power factor of power converters.
The prior art does provide circuits for power factor compensation.
S For example, W. Shepherd and P. Zand suggest certain such circuits in Ch. 11 otheir book ~y Flow and Power Factor in Nonsinusoidal Circuits, Cambridge
University Press, 1979. Most are shunt circuits, both linear and nonlinear. These
shunt circuits tend to be only marginally effective in improving the power factor.
Several series-compensation schemes are also described, but they a~e complex,
10 costly, and often unreliable active networks that supply the harmonic frequency
content of the rectified load current so that the A.C. lines have to supply only the
current at the fundamental frequency.
^ A compensation circuit that avoids the disadvantages of those
proposed by Shepherd and Zand is presented by G. J. Scoles in Ch. 18 of the
15 Handbook of Rectifier Circuits, John Wiley & Sons, 1980. He describes a "tuned
bridge rectifier" in which an L-C circuit is placed in series between the A.C. lines
and the rectifier, which L-C circuit is specifically tuned to resonate at the A.C.
Iine frequency. This circuit has serious drawbacks of its own, however. With
variations in load on the rectifier, the rectified voltage varies significantly -- on the
20 order of 30% between no load or light load and norrnal load (of about 72~ ~~~which is an unacceptably high variance for most applica~ions. At high loads, andparticularly during start-ups, overloads, or short-circuits, the voltages and the
current peaks across both the inductor and the capacitor of the L-C circuit get very
large and are likely to damage both the inductor and the capacitor unless very
25 robust, and hence very expensive, components are used that can wi~hstand the
surges and avoid breakdown.
A beKer solution than those hitherto proposed by the art is therefore
required for improvement of power factor of power converters.
Summarv of the Invention
The invention is directed to solving these and other problems of the
prior art. According to the invention, an arrangement for improving power factorof a rectifier drawing power from a power source having a substantially sinusoidal
voltage with an operating frequency, comprising: in combination, first inductivemeans and captive means connected in series with each other and for connection
. ~
- 2a -
-~ 13q~
in series between a rectifier drawing power from a power source and the power
source having a substantially sinusoidal voltage and tuned to resonate at a
frequency higher than the frequency of the voltage to improve the power factor of
the rectifier.
., ~..
;.. ~.. , ~ ~ .
~31~
Preferably, the resonant frequency falls within a range of frequencies from 1.4
times the source frequency to 3.5 times the source frequency. A highly-desirableresonant frequency has been found to be about two times the source frequency.
Empirically derived formulas for the desirable values of the inductors and
5 capacitors are as follows
L ~ f el28 x P~ microhenries, and
C ~ 104 e-6.38 xPF farads where
PF is the desired power factor,
f is the input power frequency in hertz, and
10 R is the load impedance in ohms.
The arrangement reduces the RMS current drawn by the apparatus,
thereby improves the power factor of the A.C.-powered apparatus, and hence
eliminates or lessens the problems caused by a low power factor, described above.
In particular, the arrangement generally improves the efficiency of the A.C.-
15 powered apparatus; decreases the apparatus' power losses, particularly inequivalent series resistance (ESR) of any input capacitors, rectifier diodes, line
filters, and power cords of the apparatus; reduces the capacity requirements of the
A.C. power distribution system components by lowering the apparatus' peak
power demands; and reduces the strain on, and hence the cost and the capacity
20 requirements of, any rectifier diodes, storage capacitors, and wiring of the
apparatus. Because the applicability of progressively more stringent safety
requirements is generally dependent upon maximum RMS currents that may exist
in the apparatus, the power factor improvement makes it easier and less expensive
to meet applicable safety requirements. In addition, the arrangement provides
25 inrush and short-circuit current protection.
Furthermore, the power factor-improving arrangement with unchanged
component values is often suited for use with a variety of input power
frequencies, particularly with both 50 and 60 Hz, and hence it does not require
redesign or change to accommodate the power frequency differences that exist
30 between nations. The arrangement causes only a slight drop in the output
voltage--approximately 4.5%. It also tends to keep input current and voltage in
phase. DuIing overloads or short circuits, the arrangement autornatically limits the
current to levels which are unlikely to cause damage or nuisance events such as
fuse blowing, and during start-ups the network also automatically limits to an
35 acceptable value the inrush cunent that would otherwise be drawn by a typically-
''`~` ,:
~L3~4L4~L8
large rectifier capacitor, and hence eli~unates the need for additional "soft start"
circuitry that is typically used with such rectifiers. Moreover, the a~angement can
be constructed of only passive and well-known components -- inductors and
capacitors -- and hence it is simple and reliable.
In accordance with another aspect of the invention there is provided
a power supply comprising: means for connecting to a power source having
substantially sinusoidal alternating voltage with an operating frequency; means
coupled to the connecting means ~or rectifying the alternating voltage; and a
circuit comprising inductive means and capacitive means connected in series witheach other, the circuit serially interyosed between the connecting means and therectifying means and tuned to resonate at a frequency higher than the frequency
Oe the alternating voltage to improve the power factor of the power supply.
These and other advantages and features of the present invention will
become apparent from the following description of an illustrative em~odiment of
the invention taken together with the drawing.
Brief Description of the Drawin~
- FIG. 1 is a block diagram of a system embodying an illustradve
exarnple of the present invention;
FIG. 2 is a circuit diagrarn of the power converter of the system of
FIG. 1;
FIG. 3 is a circuit diagram of the basic configuration of the power
factor compensator of the system of FIG. 1; and
FIGS. 4-lO are circuit diagrarns of alternative conSgurations of the
power factor compensator of the system of FIG. l.
Detailed Description
FIG. 1 shows an illustrative system lO embodying an illustrative
embodiment of the invention. System lO comprises a power supply l l, and a
load 12 connected to the power supply. Illustratively, system lO is a computer
system, load 12 represents the computer itself, and power supply l l is the power
supply that drives the computer.
Further as indicatçd in FIG. l, system lO is plugged into a source of
A.C. power 13. Illustratively, power source 13 includes conventional building
power lines 14 and a wall socket lS, and system lO is plugged into socket lS by
rne~ns of a power cord 16.
,;~
. .~..
14441~
Power supply 11 includes a power converter 17. Power converters are
known in the art, but for purposes of completeness of discussion, an illustrative
power converter is shown in FIG. 2. It comprises a conventional rectifier
bridge 21 made up of four diodes 22, a storage capacitor 23, and a conventional
5 power conditioner 24 such as a DC-AC or a DC-DC converter.
Power supply 11 also includes a power factor compensator 18 which
is interposed between power converter 17 and power source 13, and which
functions to improve the power factor o power converter li, and hence of power
supply 11. Whereas a conventional power supply -- one without compensator 18
10 -- typically draws current from power supply 13 in narrow but very large current
~3~4~4~3
pulses, the use of compensator 18 substantially widens, and reduces the magnitude
of, the current pulses without significantly affecting the phase relationship of the
current and the voltage.
The basic structure of compensator 18 is shown in FIC~. 3: it
5 comprises an inductor 30 and a capacitor 31 connected in series with power
converter 17 and power source 13. Safety considerations dictate that inductor 30and capacitor 31 be connected in series with the power leg of the connection
between supply 13 and converter 17, leaving the ground connection a straight-
through connection. However, from a functional standpoint, inductor 30 and
10 capacitor 31 may be connected instead with the ground leg of the connection, or
inductor 30 may be connected with one leg of the connection while capacitor 31
may be connected with the other leg, as shown in FIG. 4.
Further permutations on the basic configuration are possible: a
plurality of inductors 70, 71 and capacitors 80, 81 may be used, whose combined
15 values in the circuit of compensator 11 reduce to the value of inductor 30 and
capacitor 31 connected as shown in F~G. 3. One such possible arrangement is
shown in FIG. 5, where one inductor-capacitor pair 70, 80 is connected in serieswith the power leg of the connection between supply 13 and converter 17, and
another inductor-capacitor pair 71, 81 is connected in series with the ground leg.
20 Another such possible arrangement is shown in FIG. 6, where two inductor-
capacitor pairs 70, 80 and 71, 81 are connected into a leg of the connection
between supply 13 and converter 7 in parallel with each other. Each of these L-Cpairs may then be tuned to a different frequency.
Significantly, in a departure from the prior art, inductor 30 and
25 capacitor 31 are tuned, i.e., their relative values are selected, to resonate at a
frequency not equal to, but higher than, the frequency of power source 13.
Desirably, the selected resonant frequency falls within a range of frequencies
from 1.4 times the source frequency to 3.5 times the source frequency.
Depending on the weighting of different parameters such as 1) the amount of
30 improvement desired in the power factor, 2) the amount of allowable drop caused
by compensator 18 in rectified voltage, 3) the peak current limit for overloads,4) the maximum allowable voltage across capacitor 31, and 5) the maximum
current allowable through inductor 30, there is a range of acceptable resonant
frequencies, and hence a range of acceptable values of inductor 30 and
35 capacitor 31. Experimentally, it has been found that very good performance with
respect to the above-mentioned parameters is obtained with a resonant frequency
- .
-~ 13~
of about two times the power source 13 frequency. Because power source 13
frequencies of both 50 Hz and 60 H~ are common, a resonant frequency of
about 110 Hz serves quite well for both power source frequencies. Empirically-
derived formulas for the desirable values of inductor 30 and capacitor 31 are as5 follows:
L ~ 3f8R c12.8 x PF microhenries, and
C ~ 104 e-6.38 xPF farads, where
PF is the desired power factor,
f is the source fre~uency in hertz, and
10 R is the load 12 impedance in ohms.
As the formulas suggest, compensator 18 need not be a finely-tuned
circuit; rather, there is a significant tolerance allowed on the values of inductor 30
and capacitor 31 which still meet the basic power factor objectives. Of course,
inductor and capacitor values selected from the tolerance range yield a slightly15 lower than the desired P.F. value. Examples of reasonable tolerances for the
values are +30% for the capacitance value and +30% for the inductance value.
The acceptable tolerances allow use of relatively-cheap components for
constracting compensator 18, and permit the interchangeable use of a particular
compensator 18 for both 50 Hz and 60 Hz power supply systems.
Of course, a plurality of smaller capacitors connected either in parallel
or in series may be used instead of capacitor 31. However, from a cost
standpoint, it may be preferable to use a step-up transforrner in conjunction with a
single capacitor to achieve the equivalent of the desired capacitance. This
arrangement is shown in FIG. 7. As shown, the primary winding of a step-up
transformer 40 is connected in series with inductor 30, in place of capacitor 31.
capacitor 41 is connected across the secondary winding of transformer 40.
Assuming transformer 40 to be a l:N step-up transformer, the capacitance of
capacitor 41 is C2, where C is the capacitance of capacitor 31 that has been
replaced by transformer 40 and capacitor 41. The value of N may be either
30 greater than 1, for a step-up voltage transformer, or less than one, for a step-up
current transformer.
It may further be economically advantageous to combine inductor 30
and transforrner 40 of FIG. 7 into a single magnetic structure. Such combined
structures are known in the art. An illustrative example thereof in the context of
~ ,...., . .; . , .
L4~L8
compensator 18 is shown in F~G. 8. The arrangement uses a transformer 60
having a single folded winding. The power input of compensator 18 is coupled to
the input lead of transformer 60 and to one lead of capacitor 41. The output lead
of transformer 60 is connected to the other lead of capacitor 41, and power output
5 of compensator 18 is coupled to the tap of transformer 60.
Compensator 18 does result in a small voltage drop at the outpu~ of
converter 17. If necessary, this drop may be offset in several ways. One way,
illustrated in FIG. 9, uses a step-up transformer 50. In this arrangement, the
output of capacitor 31 is connected to a tap of transformer 50. One end of the
10 winding of transformer 50 is connected to the input of inductor 30. The other end
of the winding is connected to the return line, or ground. Another way, illustrated
in FIG. 10, uses an inductor 60 instead of transformer 50. Inductors 60 and 30
are not magnetically coupled. The value of inductance of inductor 60 is large
relative to that of inductor 30, illustratively on the order of 50 times the value of
15 inductor 30. One end of inductor 60 is connected to the output of capacitor 31
and to the input of inductor 30, and the other end is connected to the return line,
or ground.
Of course, it should be understood that various changes and
modifications to the illustrative embodiment described above will be apparent to20 those skilled in the art. For example, a non-linear inductor may be used to good
advantage and allowed to partly saturate during overload, to reduce the peak
voltage across the capacitor. Such changes and modifications can be made
without departing from the spirit and the scope of the invention and without
diminishing its attendant advantages. It is therefore intended that all such changes
25 and modifications be covered by the following claims.
,