In recent years, regulation in terms of electrical efficiency, electromagnetic noise, and
reliability of the electronic apparatus, in all the application fields, are pushing to use power
factor correction in almost all applications; from low power range to high power range, and
from industrial to domestic applications.
For this reason the design of a new PFC architecture is required to satisfy these new
priorities. The PFC must work from a few watts to thousands of watts and also use cheap
devices and materials. The design described in this paper is able to cover most of the
common domestic and industrial applications.
A few examples of the application fields are: air conditioning, inverters (for fans and pumps),
welding machinery, and industrial battery chargers.
Figure 11.Scope acquisition: startup with a connected load of 1200 W . . . . . . . . . . . . . . . . . . . . . . . 25
Figure 12.EMI tests after the design and assembling of the filters . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4/27 Doc ID 15500 Rev 1
AN2951PFC working mode general description
1 PFC working mode general description
Essentially two methods of controlling power factor corrector (PFC) preregulators are well
known and used. Fixed frequency PWM average current control and fixed-on-time variable
frequency control.
The first method is a relatively complex control that requires a controller (e.g. L4981) which
is usually sophisticated and expensive. With this method the current on the inductor is
continuous and oscillates around the sinusoidal semiperiod value with a predetermined and
fixed ripple.
In the second method the current on the inductor is discontinuous, reaching zero and
starting to increase again on each switching period. It is cheaper but cannot be used for high
power range applications due to the large peak current it causes during its action according
to the load.
Other than these two methods, a third may be approached. Instead of maintaining the ontime fixed, such as TM PFC, the Toff is kept constant and the Ton is free to be changed in
order to modulate the power drained from the source according to the load. This is called
the “fixed-off-time modulation method”. Obviously, it is also a variable frequency control with
some advantages, in particular the spread of noise energy conducted on the net that gives a
lower noise power density in respect to a fixed frequency modulation, which simplifies the
design and realization of the main filter required to match the EMC regulation.
The interesting point is that this kind of modulation can be obtained using a simple and
cheap controller designed for TM (transition mode), e.g. the L6562 or L6563 by
STMicroelectronics.
This method, with a complete theoretical explanation, is well described and depicted in the
AN1792 application note by STMicroelectronics; “Design of fixed-off-time controller PFC
preregulators with L6562”. We suggest first reading the antecedent AN1792 application note
before continuing to read. Starting from this point we refer to the AN1792 to study in depth
the design of a 3 kW FOT PFC preregulator.
Figure 2.Line, inductor, switch and diode current on TM and CCM control
!-V
Doc ID 15500 Rev 15/27
Practical implementation of an FOT-controlled PFC preregulatorAN2951
2 Practical implementation of an FOT-controlled PFC
preregulator
Figure 3.Block diagram of a fixed-off-time architecture
!-V
The power conversion topology is classically based on a boost converter modulated to drain
a current that follows the shape of the main AC voltage; in order to have a current as
sinusoidal as possible, and in phase with the mains,assuring a high power factor and low
THD (total harmonic distortion).
The architecture is based on a fixed-off-time generator, a multiplier, a PWM generator, and a
voltage amplifier with frequency compensation to implement the output voltage control
against the load variations.
To modulate the current on the inductor a current reading is required. The information
regarding the peak current can be retrieved by reading, at the end of the Ton period, the
voltage drop on a sense resistor in series with the power switch.
The input at the multiplier (Vmult) is responsible of the “shaping” action dictated by the input
AC voltage.
The amplitude of the sinusoidal shaping action that directly imposes the reference to the
PWM generator is given by the voltage loop control through the error amplifier and the
frequency compensation.
In practice, the PWM block fixes the Ton time. As soon as the current on the power switch
reaches the value fixed by the output multiplier level (Vcsref), the output of the comparator
resets the flip-flop and turns off the power switch.
At the same time the output of the comparator triggers the timer and the counting of the offtime starts. At the end of the fixed-off-time period the timer block sets the flip-flop and the
power switch is again switched on.
These functions can be implemented by using a TM controller such as the L6562 or L6563.
6/27 Doc ID 15500 Rev 1
AN2951Fixed-off-time using the L6562
3 Fixed-off-time using the L6562
The circuit to implement FOT control with the TM L6562 driver is shown in Figure 4.
On the L6563 driver the V
pin is available which allows, working in TM, to detect the zero
ZCD
crossing of the current on the main inductor. Here we show how it is possible to use this pin
to force the driver to work in a fixed-off-time mode.
During Ton of the power switch, pin 7 (GD gate driver) is high and the D diode is direct
biased. Under this condition, because the output GD is at Vcc (usually 15v), the voltage on
V
is internally clamped to V
ZCD
ZCDclamp
=5.7 V.
During Toff, the GD pin is low, the D diode is reverse-biased and the C capacity can be
discharged by the R resistor following the exponential law:
Equation 1
t
RC
eVV−=
Until the voltage on the V
ZCDclampZCD
pin reaches the triggering threshold (about 1.4 V) the power
ZCD
switch is switched on. Using this passive net the Toff can be set by design. Solving Equation
1 in respect to the time the Toff can be calculated:
Equation 2
V
ZCDclamp
lnRCToff
V
ZCDtrigger
RC4.1
≈=
As suggested with AN1792, it is better to select a capacitor first, and calculate the needed
resistor to set the desired Toff.
The passive net composed by CS and Rs assures that, as soon as the V
capacitor is charged atV
capability of the V
ZCD
pin.
ZCDclamp
as quickly as possible without exceeding the current
goes high the C
GD
Equation 3
VVV−−
−−
FZCDclampGD
V
+
ZCDclamp
R
I
ZCD|
RR
<<
S
V
ZCDclamp
VVV
FZCDclampGD
Cs acts as a speed-up capacitor needed to instantaneously charge C in case of very short
Ton when working at high AC input voltage and light load.
The Cs capacitor should be chosen as:
Equation 4
V
CC−−<
s
The FOT control allows a CCM operation and high power capability but with the circuit
ZCDclamp
VVV
FZCDclampGD
complexity and driver performance generally sufficient on a TM controller.
Doc ID 15500 Rev 17/27
Fixed-off-time using the L6562AN2951
This is possible because a characteristic of an FOT control is the possibility to use a simple
peak current control instead of a more difficult average mode control as required on a
classic CCM control.
For further details and implications of FOT control, please refer to the AN1792 application
note.
Figure 4.FOT control with the L6562 and relevant timing waveforms
!-V
8/27 Doc ID 15500 Rev 1
AN2951Working mode along line period in a FOT PFC
4 Working mode along line period in a FOT PFC
Figure 5.Working mode along line semiperiod
!-V
In Figure 5 a full description of the main characteristics and the evolution of some key
parameters of an FOT,along the semiperiod of the input AC source, is shown. The dotted
line represents the envelope of the peak inductor current. As clearly noted, the shape is
sinusoidal as required, in order to have a high power factor. This is a consequence of the
Ton modulation carried out by following the output of the multiplier Vcsref (see Figure 3).The
dash-dot line represents the inductor average current. A distortion on the average
instantaneous boost current value can be noted around the zero crossing of the line voltage.
Though this aspect is not negligible, it is inevitable. It is due to the working mode
(discontinuous) in this region. A trade off between operating frequency and line current
distortion has also to be taken into account. To limit the line current distortion at high line
voltage a suitably large Toff must be selected; also fixing a limited current ripple on the boost
inductor, a bigger inductor size may be needed. Returning to Figure 5, the continuous line
represents the inductor current ripple along the sinusoidal semiperiod (0 -π).
Some important considerations can be taken into account regarding this matter. Two kinds
of working modes can be recognized during the semiperiod; DCM and CCM. For low input
voltage (near zero crossing of the mains) the converter works in discontinuous mode; this
means that the current reaches zero for a certain time during the switching period. The
behavior along the semiperiod is symmetrical in respect to the centre of the semiperiod
(π/2). Two points can be identified where the current ripple on the boost inductor is equal to
the peak value. Those two points dictate the transition between the discontinuous and
continuous mode, therefore we call the respective angle; the transition angle. If θ is the
transition angle:
Equation 5
DCM
0t
<<
θ
--- -
and
ω
θ−π
ω
π
<<
t
ω
Equation 6
CCM
θ
t
ω
θ−π
<<
ω
The amplitude of the DCM portion, and so the angle, are not fixed but depends on input
voltage and load power. As the power is reduced, the DCM region increases.
Doc ID 15500 Rev 19/27
Loading...
+ 18 hidden pages
You need points to download manuals.
1 point = 1 manual.
You can buy points or you can get point for every manual you upload.