See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/338999594

Design of Fixed-off-time Controlled PFC Pre-regulators with the L6562

Technical Report · November 2003

DOI: 10.13140/RG.2.2.11452.80005

CITATIONS READS

2 1,105

1 author:

Claudio Adragna
STMicroelectronics
60 PUBLICATIONS   642 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Claudio Adragna on 03 February 2020.

The user has requested enhancement of the downloaded file.

 AN1792
APPLICATION NOTE
DESIGN OF FIXED-OFF-TIME-CONTROLLED
PFC PRE-REGULATORS WITH THE L6562
by C. Adragna

Beside Transition Mode (TM) and fixed-Frequency Continuous Conduction Mode (FF-CCM) operation
of PFC pre-regulators a third approach is proposed that couples the simplicity and affordability of TM
operation with the high current capability of FF-CCM operation: it is a peak current-mode control with
fixed-OFF-time. After showing advantages and limits of this technique, in both its basic and advanced
implementation, design equations will be given and a practical design will be illustrated and evaluated.

Introduction
Two methods of controlling Power Factor Corrector (PFC) pre-regulators based on boost topology are cur-
rently in use: the Fixed-Frequency (FF) PWM and the Transition Mode (TM) PWM (fixed ON-time, variable
frequency). The first method employs average current-mode control, a relatively complex technique requiring
sophisticated controller IC's (e.g. STMicroelectronics' L4981A/B) and a considerable component count. The
second one uses the simpler peak current-mode control, which is implemented with cheaper controller IC's
(e.g. STMicroelectronics' L6561, L6562), much fewer external parts and is therefore much less expensive.
With the first method the boost inductor works in Continuous Conduction Mode (CCM), while TM makes
the inductor work on the boundary between continuous and discontinuous mode, by definition. For a given
power throughput, TM operation then involves higher peak currents as compared to FF-CCM (see figure
1). This, also consistently with the above mentioned cost considerations, suggests the use of TM in a lower
power range, while FF-CCM is recommended for higher power levels.
This criterion, though always true, is sometimes difficult to apply, especially for a midrange power level, say
around 150-300W. The assessment of which approach gives the better cost/performance trade-off needs to
be done on a case-by-case basis, considering the cost and the stress of not only power semiconductors and
magnetics but also of the EMI filter: at the same power level, the switching frequency component to be fil-
tered out in a TM system is twice the line current, whereas it is typically 1/3 or 1/4 in a CCM system.

Figure 1. Line, inductor, switch and diode currents in: a) FF-CCM PFC, b)TM PFC.

Inductor current

0 00 0 000000000000000000000000000 0 0 0 0
peak envelope
ILpk

00 00 00 00 00000 00 00 00 00 00 00 00 00 00 00 00000000000000 00 00 00 0 0
Low frequency Switch current Diode current
Inductor current inductor current

00 00
Low frequency

00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0 0 0
inductor current

Switch current

00 00 00
00000000000000000000000000000000
Diode current

00 ON
00000000000000000000000000000000 ON
Switch Switch
OFF OFF

a) FF-CCM control b) TM control

November 2003 1/30

AN1792 APPLICATION NOTE

In this area where the TM/CCM usability boundary is uncertain, a third approach that couples the simplicity
and affordability of TM operation with the high current capability of CCM operation can be a solution to the
dilemma.
Generally speaking, FF PWM is not the only alternative when CCM operation is desired. FF PWM modu-
lates both switch ON and OFF times (their sum is constant by definition), and a given converter will operate
in either CCM or DCM depending on the input voltage and the loading conditions. Exactly the same result
can be achieved if the ON-time only is modulated and the OFF-time is kept constant, in which case, how-
ever, the switching frequency will not be fixed anymore (see figure 2). This is referred to as "Fixed-OFF-
Time" (FOT) control. Peak-current-mode control can still be used.

Figure 2. Basic waveforms for Fixed-frequency PWM (a) and Fixed-OFF-Time PWM (b)

T ON' TOFF' T ON' T OFF' TON'

Gate
drive
signal

TON TON TON

TOFF TOFF
T SW T SW

a) Fixed Frequency

TON' TON' TON'

TOFF T OFF
T SW' TSW'
Gate
drive
signal

TON TON T ON
TOFF TOFF
TSW TSW

a) Fixed OFF-TIme

The concept of FOT control is not new [1], [2] but, to the author's knowledge, it has never been applied to
PFC pre-regulators to allow CCM operation. In addition to the many advantages and the few drawbacks
that this control technique brings to PFC pre-regulators and that will be highlighted in the following sec-
tions, an important point is that FOT control does not need a specialized control IC. A simple modification
of a standard TM PFC controller operation, requiring just few additional passive parts and no significant
extra cost, is all that is needed.

Operation of a FOT-controlled PFC pre-regulator and its practical implementation

Figure 3 shows a block diagram of a FOT-controlled PFC pre-regulator. An error amplifier (VA) compares
a portion of the pre-regulator's output voltage Vout with a reference VREF and generates an error signal
VC proportional to their difference. VC, a DC voltage by hypothesis, is fed into an input of the multiplier
block and there multiplied by a portion of the rectified input voltage VMULT. At the output of the multiplier,
there will be a rectified sinusoid, VCSREF, whose amplitude is proportional to that of VMULT and to VC,
which represents the sinusoidal reference for PWM modulation. VCSREF is fed into the inverting input of a
comparator that, on the non-inverting input, receives the voltage V CS on the sense resistor Rsense, pro-
portional to the current flowing through the switch M (typically a MOSFET) and the inductor L during the
ON-time of M. When the two voltages are equal, the comparator resets the PWM latch and M, supposed
already ON, will be switched off.

2/30
 AN1792 APPLICATION NOTE

Figure 3. Block diagram of an FOT-controlled PFC pre-regulator.

Vpk

0 L D Vout
0
M
Iin R1 Cin Co

Vin =
88 to 264 EMI FILTER
VAC R2 Rsense

Vcs 0

TIMER Timer Start

(Fixed OFF-time generator)

Driver

S Q
PWM
+ latch
PWM R Q
- VREF
0
0 V CSREF +
VMULT
Multiplier VA
Vc -

0
Frequency
VREF
compensation

As a result, VCSREF determines the peak current through M and the inductor L. As V CSREF is a rectified
sinusoid, the inductor peak current will be enveloped by a rectified sinusoid as well. The line current Iin
will be the average inductor current, that is the low frequency component of the inductor current resulting
from the low-pass filtering operated by the EMI filter.
The PWM latch output Q going high activates the Timer that, after a predetermined time TOFF has elapsed,
sets the PWM latch, thus turning M on and starting another switching cycle. If T OFF is such that the induc-
tor current does not fall to zero, the system will operate in CCM.
It is apparent that FOT control requires nearly the same architecture as TM control, just the way the OFF-
time of M is determined changes. It is not a difficult task to modify externally the operation of standard TM
PFC controller so that the OFF-time of M is fixed. As a controller we will refer to the L6562 [4], which is suit-
able for a few hundred watts power applications because of its gate drive capability and its high noise im-
munity.
The circuit that implements FOT control with the L6562 is shown in figure 4 along with some relevant
waveforms. During the ON-time of M the gate voltage VGD is high, the diode D is forward-biased and the
voltage at the ZCD pin is internally clamped at VZCDclamp ≈ 5.7V. During the OFF-time of M VGD is low,
the diode D is reverse-biased and the voltage at the pin decays with an exponential law:
–t
---------
RC
V Z CD = V Z CDcla mp e ,

until it reaches the triggering threshold (VZCDtrigger ≈ 1.4V) that causes the switch to turn on. The time need-
ed for the ZCD voltage to go from VZCDclamp to VZCDtrigger will define the duration of the OFF-time TOFF:
V ZCDc lam p
T O FF = RC ln ------------------------------- ≈ 1.4 ⋅ RC . (α)
V Z CDtri gge r

3/30
AN1792 APPLICATION NOTE

As a practical rule, it is convenient to select a capacitor first and then to calculate the resistor needed to
achieve the desired TOFF. As the gate voltage VGD goes high the resistor Rs charges the timing capacitor
C as quickly as possible up to VZCDclamp, without exceeding clamp rating (IZCDx =10 mA). Then it must
fulfill the following inequalities:

V GDx – V ZCDc lam p – V F V G D – V Z CDcla mp – V F

- < Rs < R ------------------------------------------------------------
-------------------------------------------------------------- - , (β)
V ZCDcla m p V ZCDc lam p
I ZCDx + -----------------------------
R

where VGD (assume VGD = 10V) is the voltage delivered by the gate driver, VGDx = 15V its maximum
value, and VF the forward drop on D.

Figure 4. Circuit implementing FOT control with the L6562 and relevant timing waveforms.
5.7V
Pin 5
M 1.4V
GD
7
TOFF TON T OFF TOFF
S without Cs
L6562 4
CS
Internal
PWM latch
5 signals R
Rsense
(see fig. 3)
ZCD Rs
Q

D
1N4148 Pin 7
Cs
R C
VREFCS
Pin 4

When working at high line/light load the ON-time of the power switch becomes very short and the resistor
Rs alone is no more able to charge C up to VZCDclamp. The speed-up capacitor Cs is then used in parallel
to Rs. This capacitor will cause an almost instantaneous charge of C up to a level, after that Rs will com-
plete the charge up to VZCDclamp. It is important that the steep edge caused by Cs does not reach the
clamp level, otherwise the internal clamp of the L6562 would undergo uncontrolled current spikes (limited
only by the dynamic resistance of the 1N4148 and the ESR of Cs) that could overstress the IC. Cs must
then be:
V ZCDc lam p
Cs < C --------------------------------------------------------------- , (γ)
V GDx – V ZCDc lam p – V F

Implications of FOT control for CCM-operated PFC pre-regulators

Essentially, the aim of FOT control in PFC pre-regulators is to allow CCM operation, and hence high power
capability, but with the same complexity level of a TM-operated PFC stage. This goal can be achieved
since the properties of FOT control enable the use of the simpler "peak current-mode" control rather than
the more complex "average current-mode" needed by the FF-CCM approach.
In short, peak current-mode FOT control provides an unconditionally stable inner current loop with no gain
peaking and shows "dead-beat control" characteristics. Referring to reference [1] for a detailed explana-
tion of these properties, it is useful to see how they affect the characteristics of FOT-CCM PFC pre-regu-
lators.
1) Simple control, low part count. The absence of the sub-harmonic instability typical of FF-systems at duty
cycles greater than 50% (see figure 5), makes FOT control very convenient in a PFC boost stage, where
the duty cycle can theoretically reach 100%. FF-CCM using peak-current-mode results in an unstable sys-
tem as long as the instantaneous line voltage is below half the regulated output voltage (condition for duty
cycle >50%) with no slope compensation, unless the system is designed to run in DCM (Discontinuous

4/30
 AN1792 APPLICATION NOTE

Conduction Mode) under these conditions. This is quite a limitation. This is why average-current-mode is
usually preferred for FF-CCM operation, despite the penalty of an increased circuit complexity. Addition-
ally, FOT control does not require the use of an auxiliary winding on the boost inductor as does the TM
approach. If the control IC can be powered by an external source (e.g. the transformer of a cascaded DC-
DC converter) the inductor can be made with a single winding with some saving in its cost.
2) Dynamic behavior improvement. The optimum response of the inner current loop tends to limit inductor
current ringing resulting from load changes. However, it has little impact on the performance of the outer
voltage control loop, still largely dominated by the low bandwidth needed to achieve a high PF.

Figure 5. FF vs. FOT control at D>50%: instability (FF) and stability with critical damping (FOT).
perturbed inductor current perturbed inductor current
steady-state inductor current
Inductor steady-state inductor current Inductor same slope
current programmed value
current programmed value
Dmax · Tsw

∆Ι 2

∆Ι 0 Tsw
T OFF T OFF
∆Ι 1

time time

FIXED-FREQUENCY CONTROL FIXED OFF-TIME CONTROL

3) Reduced EMI emissions. Variable frequency operation is inherent in FOT concept: any variation of load
current or input voltage is compensated by the feedback control with a variation of the switch ON-time.
Thus, the switching frequency of a PFC pre-regulator is modulated, at a modulation rate twice the mains
frequency, by the input voltage swinging all the way from zero to the peak. The result is a spread-spectrum
action that reduces the peak energy of the noise generated and simplifies the ability to comply with EMI
regulations.
4) DCM and CCM always live together, at least at nominal load. Figure 6 shows typical current waveforms:
at two points along the sinusoid the inductor current ripple during one switching cycle equals the peak val-
ue in that cycle. This is the boundary between CCM and DCM operation: there will be DCM around the
line voltage zero-crossings and CCM around the top of the sinusoid. As the power is reduced, the region
of DCM operation will get larger until it takes up the entire line cycle.

Figure 6. Typical current waveforms along a line voltage half-cycle of a FOT-controlled PFC stage.

Inductor current peak envelope

Inductor current ripple
Inductor average current

DCM DCM
CCM

0 0.52 1.05 1.57 2.09 2.62 3.14

q

5) Stress due to boost diode's reverse recovery and MOSFET's capacitive loss is alleviated. CCM opera-
tion requires the use of ultra-fast recovery diodes. In the DCM portion, however, (typically, about 30% of
the line cycle) the recovery of the boost diode is not invoked, hence the related losses in the diode itself
and those induced in the MOSFET are reduced. Additionally, because of the DCM portion, the capacitive
losses at MOSFET turn-on due to the discharge of the drain capacitance are decreased as well. Actually,
during a small part of the DCM portion the MOSFET is soft-switched.

5/30
AN1792 APPLICATION NOTE

6) Line current distortion is not negligible. It will be shown that as long as the system operates in CCM the
line current waveform is a portion of a sinusoid but, as it enters DCM around the line voltage zero cross-
ings, the shape changes, causing a distortion of the line current (see the dash-dot line in figure 6). How-
ever, it will be shown that its harmonic contents is still comfortably compliant with EN61000-3-2 standards
on harmonic current emissions. Thus the practical impact of this drawback is very limited.
7) Trade-off between operating frequency and line current distortion. It will be shown that, in order to limit
line current distortion at high line, the OFF-time must be selected greater than a minimum value. This
could prevent the use of a switching frequency high enough to have a relatively small inductor size. How-
ever, a variant of the basic FOT control will be presented that allows the designer to overcome this issue.

FOT-CCM PFC: large-signal characteristics.

The large-signal characteristics of a FOT-CCM PFC boost pre-regulator will be now discussed from the qual-
itative point of view. For details on the FOT-CCM PFC large signal model derivation, please refer to [3]. The
quantitative results of the approximate analysis described in [3], provided in tables 2, 3 and 4, are the basis
for a design procedure that will be outlined in another section.
It is well-known that in a boost converter, during the MOSFET's OFF-time TOFF (which is a given design pa-
rameter), the inductor demagnetizes and releases the energy stored during the ON-time. If in a switching cy-
cle TOFF is long enough to completely discharge the inductor within that cycle, there will be DCM operation,
otherwise there will be CCM operation. As previously said, if the PFC boost is operated with FOT, DCM and
CCM operation alternate in a line half-cycle. The phase angle θT ∈ [0, π/2] where the operation changes
from DCM to CCM will be referred to as the "transition angle". Considering waveform symmetry, operation
will change from CCM back to DCM at the supplementary phase angle π - θT. In each line half-cycle there
will be CCM operation for θT < θ < π - θT and DCM operation for θ < θT and θ > π - θT. In a given converter
(i.e. for given Vout, L and TOFF), the transition angle θT depends on the operating conditions (line voltage
and output load). Figure 7 shows schematically the inductor current, also pointing out how it is split between
the MOSFET M and the diode D. The low-frequency component, that is the line current Iin(θ), is shown too.

Figure 7. Inductor, switch and diode currents in a CCM-FOT-controlled PFC stage.

CCM Inductor current

peak envelope
ILpk
DCM
00 00 00 00 00 DCM

00 00 00 00 00
Low frequency
inductor current

00 00 00 00 00
Switch current
00 00 00 00 00 Diode current

00 00 00 00 00
000 00
ON
Switch OFF
T OFF T OFF
θT π − θT

During the DCM portion of the line half-cycle MOSFET's ON-time TON is constant and, as TOFF is constant
by definition, the switching frequency and MOSFET duty cycle will be constant as well. Their value depend
on the peak input voltage and the output load. During the CCM portion they change with both the peak
and the instantaneous line voltage but not with the load. The switching frequency is maximum on the top
of the sinusoid, reaches the minimum at the DCM/CCM boundary and does not change any more in the
DCM portion.

6/30
 AN1792 APPLICATION NOTE

Table 1. List of basic symbols used in tables 2 to 4

Symbol Parameter
TOFF Fixed MOSFET OFF-time
L Boost inductor’s inductance
Vpk Peak line voltage ( 2 times the RMS value Vac)
Vout Pre-regulator’s output regulated voltage
Pout Pre-regulator’s output power
Pin Pre-regulator’s input power (= Pout / η , η = efficiency)
ILpk Peak of inductor current sinusoidal envelope
k Vpk/Vout ratio
Γ Conventional zero-voltage-input inductor current ripple: Vout·TOFF/L

Table 2. Timing quantities of FOT-controlled boost PFC pre-regulators

Symbol Parameter DCM CCM
θT Transition angle (DCM ⇒CCM boundary) –1 Γ
sin -----------------------
I Lpk + kΓ

TON(θ) MOSFET ON-time LI Lpk

--------------  --------------
1 - – 1 T

kVou t k sin θ OF F

D(θ) MOSFET duty cycle 1 - ksinθT 1 - ksinθ

fsw(θ) Switching frequency k - k
------------- sin θ T -------------- sin θ
TO FF T O FF

δfsw(θ) Switching frequency modulation depth 1 – sin θ T

2 ------------------------
1 + sin θ T

TFW(θ) Inductor demagnetization time k sin θ -

-----------------------
TOFF
T
1 – k sin θ O N
DL(θ) Inductor current circulation duty cycle 1 – k sin θ 1
---------------------------T
1 – k sin θ

Table 3. Inductor and line input current in FOT-controlled boost PFC pre-regulators
Symbol Parameter DCM CCM
∆IL(θ) Inductor current ripple ILpksinθ Γ(1 - k sinθ)
IL(av)(θ) Inductor average current
1
2
ILpk
--- ----------------------- sin θ
------------------------
I Lpk
Γ 
+ --- k sin θ – Γ
---
2 ILpk + kΓ 1 – k sin θ 2 2

Iin(θ) Line input current

1
2
ILpk
--- -----------------------
sin θ -
--------------------------
 I Lpk
Γ  Γ
+ --- k sin θ – --- sgn ( sin θ )
2 2
2 I Lpk + kΓ 1 – k sin θ

As shown in figure 7, the inductor current will be a series of rising (during TON) and falling (during TOFF)
ramps whose peaks are enveloped by ILpksinθ. In a single switching cycle, the current will always be tri-
angular: in the DCM portion the triangles start from zero and touch zero before the end of the cycle, in
CCM the triangles are superimposed on top of a current pedestal. The inductor current ripple is minimum
on the top of the sinusoid and maximum at the transition DCM⇔CCM. Note that this property gives a prac-
tical meaning to the transition phase angle θT; its sine provides the ratio of the maximum current ripple
amplitude to the inductor peak current, the so-called "ripple factor" Kr, a parameter typically used in the

7/30
AN1792 APPLICATION NOTE

design phase to specify how deep in CCM the system is required to operate:

∆I Lp k ( m ax )
Kr = sin θ T = ---------------------------
- . (1)
I Lp k

Line current is made by a sinusoidal portion (during CCM, θT < θ < π - θT), shifted downwards during pos-
itive half-cycles and upwards during negative half-cycles by Γ/2, in-phase with the line voltage, that is
joined to line voltage zero-crossings through non-sinusoidal segments corresponding to the DCM portion.
These non-sinusoidal segments result in a distortion of the line current, thus it is not possible to achieve
unity power factor even ideally, unlike as with TM and FF-CCM techniques.

Table 4. Approximate energetic relationships in FOT-controlled boost PFC pre-regulators

Symbol Parameter Approximate value
ILpk Inductor peak current 2Pin 4 – kπ
≈ ----------------- + ---------------- Γ
kVout 2π
Ipk Line peak current 2Pin- + 4 – π-Γ
≈ ---------------- -----------
kVout 2π
Iin(rms)(θ) Line RMS current 2 Pin
≈ -----------------
kVout
IQ(rms) MOSFET RMS current Pin
≈ ----------------- 2 – 16k
----------
kVou t 3π
ID(rms) Diode RMS current Pin 16k
≈ ----------------- ----------
kVout 3π
ICo(rms) Output capacitor total RMS current Pin 16k-
≈ ----------------- --------- –1
kVout 3πk
ICo,H2 Peak-to-peak low-frequency current ripple Pin kΓ
≈ ------------- + -------
Vout 3π

It is worth reminding that the accuracy of the approximate energetic relationships of table 4 is quite good
at maximum load for low values of the parameter k, that is at low line voltage, but worsens at high line and
as the power throughput is reduced. Since in the design phase current stress is calculated at maximum
load and minimum line voltage, their accuracy is acceptable for design purpose. An exact description, ac-
curate under all operating conditions, requires the use of the exact model [3], which is quite difficult to treat
without using an automatic calculation tool, such as MathCad®.
To give the reader a better idea on how the various quantities change within a half-line cycle as well as
how peak and RMS current values change under different operating conditions, a series of 3-D plots gen-
erated with MathCad® are provided (figures 8 to 14).
All of these plots refer to a pre-regulator designed for Vout=400V and a maximum inductor current ripple
equal to 40% of the maximum inductor peak current (Kr=0.4) when supplied with 88Vac (k=0.311,
⇒
θT=24°) at rated load Pout0 ( Pin = Pin0). Frequency values are normalized to that on the top of the si-
nusoid @88Vac. TON is normalized to TOFF. Currents within a line half-cycle are normalized to the peak
inductor envelope ILpk @88Vac. Peak or RMS values are normalized to their respective maximum values
@88Vac.

8/30
 AN1792 APPLICATION NOTE

Figure 8. Normalized switching frequency vs. phase angle: a) Full load Pin0; b) Pin0/2; c) Pin0/10
fsw(θ) fsw (θ)
fsw (θ)
fsw0 fsw0
fsw0
3 3
3

2 2
2

1 1
1

π π 88 π
88 88
θ θ θ
Vin Vin
0 0 Vin 0
264 264
264

a) b) c)

Figure 9. Normalized MOSFET's ON-time vs. phase angle: a) Full load Pin0; b) Pin0/2; c) Pin0/10
TON (θ) T ON (θ) TON (θ)
TOFF TOFF T OFF
4
6
2
3
4
2
1
2
1

88 88 88

Vin Vin Vin

π π
θ π θ
264 0 264
0 θ 264 0

a) b) c)

Figure 10. Inductor current conduction angle vs. phase angle: a) Full load Pin0; b) Pin0/2; c) Pin0/10

DL (θ) DL (θ) DL (θ)

1 1 1
0.8
0.8 0.8
0.6
0.6
0.4 0.6
0.4
0.2 0.4
0.2
0
88 88 88

Vin Vin Vin

π π π
264 0 θ 264 θ 264 θ
0 0

a) b) c)

Figure 11. Line current vs. phase angle: a) Full load Pin0; b) Pin0/2; c) Pin0/10
Iin(θ) Iin(θ) Iin(θ)
ILpk0 ILpk0 ILpk0
0.8 0.4 0.08

0.6 0.3 0.06

0.4 0.2 0.04

0.2 0.1 0.02

0 0 0
88 88 88

Vin Vin Vin

π π π
264 0 θ 264 θ 264 θ
0 0

a) b) c)

9/30
AN1792 APPLICATION NOTE

Figure 12. Inductor current ripple vs. phase angle: a) Full load Pin 0; b) Pin0/2; c) Pin0/10
∆IL(θ) ∆IL(θ) ∆IL(θ)
ILpk0 ILpk0 ILpk0

0.4 0.4 0.2

0.3 0.3

0.2 0.2 0.1

0.1 0.1

0 0 0
88 88 88

Vin Vin Vin

π π π
264 0 θ 264 θ 264 θ
0 0

a) b) c)

Figure 13. a) Normalized Inductor peak current; b) Normalized line peak current; c) Transition angle

ILpk Ipk
ILpk0 ILpk0 θT (°)
1.0 90
0.8 0.8 80
70
0.6 0.6
60
0.4 0.4 50
0.2 40
0.2
30
0 88 0
88 88
Pin 0
Vin Vin Vin Pin
Pin 0 Pin Pin 0
264 1 264 1
Pin0
264 1

a) b) H c)

Figure 14. a) Normalized line RMS current; b) Power Factor; c) Total Harmonic Distortion

Iin(rms)
Iin(rms)0 PF
1.0 1.0 THD%
0.8 60
0.6 0.95

0.4 40

0.2 0.90 0
20
88 0
0 Pin
88 88
Vin Pin Pin0
Vin Pin
Pin 0 Pin0 Vin
264 1 264 1
264 1

a) b) c)

The pictures of figure 11 show clearly the distortion of the line waveform, quite limited at low line, more
and more accentuated as the line voltage increases and the output power decreases. This is confirmed
by the PF and THD% plots of figure 14.
It is important to evaluate the harmonic content of this current waveform, at least at nominal load, in order
to compare it with the limits envisaged by regulations. The Japanese JEIDA-MITI standard is considered
at low input voltage (100Vac) and the European EN61000-3-2 at high input voltage (230Vac).
To do so, Fourier analysis needs to be done on the waveform Iin(θ). The bar diagrams of figure 15 show
the worst-case harmonic contents of the line current ("odd counterpart" of the average inductor current
shown in figure 11), along with the limits envisaged by the above mentioned norms, showing plenty of mar-
gin.

10/30
 AN1792 APPLICATION NOTE

Figure 15. Harmonics of line currents against the class-D limits of: a) JEIDA-MITI; b) EN61000-3-2.
Amplitude (% of fundamental) Amplitude (% of fundamental)

Vin = 120Vac Vin = 264Vac

JEIDA-MITI class D limit k = 0.424 EN61000-3-2 class D limit k = 0.933
θ T = 31° θT = 40°

3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39
# Harmonic # Harmonic

a) b)

FOT-CCM PFC small-signal analysis.

A small-signal model for any PFC pre-regulator, valid for control, line and load variations at frequencies
sufficiently lower than the line frequency (quasi-static approximation), is illustrated in figure 16. This model
is generally applicable, regardless of the control technique used: the control affects the values of the pa-
rameters g2, j2 and r2 and not the circuit configuration.
The values for FOT control can be determined considering Pout=Pin and expanding the large-signal mod-
el given in [3] in a three-dimensional Taylor series to the first order around the operating point.

Figure 16. General low-frequency small-signal PFC model.

^i
2

^ j 2 ^vC r2
g2 k Co Ro

PFC pre-regulator output port

After a symbolism change and some algebraic manipulations it is possible to arrive at the following result:
2
∂i Pin - + kΓ ∂i 2 k K M K P Vo ut ∂i 2
1 = ---------------- kΓ - ,
Pin - – ----------------
g2 = ------2- = ---------------- - = ---------------------- ---------------------- , – ----
------- , j 2 = --------- - = ---------------
∂k kVout 4 ∂V c 2 R sense r2 ∂Vou t Vou t
2 πVou t

where VC is the control voltage (the output of the error amplifier of the L6562, pin #2 COMP), KM the multi-
plier gain, KP the divider ratio of R1 and R2 and Rsense the current sense resistor (see figure 3).
KM [dimensionally, 1/V] is actually a function of the operating point (an intentional non-linearity vs. VC is
introduced in the multiplier of the L6562 to partly compensate for the gain changes with the input voltage:
in particular, j2 is proportional to k2), but the value specified in the datasheet can be used with good ap-
proximation.
Ro is the incremental load resistance. The nature of the load driven by the PFC pre-regulator determines
its value. If the load is of resistive type, Ro coincides with the static resistance value:

11/30
AN1792 APPLICATION NOTE
2
∂Vou t Vo ut Vo ut
Ro = ----------------- = ------------- = ---------------- .
∂Iou t Iou t Po ut

In the important case of a DC-DC converter, which can be regarded as a constant power load, the low-
frequency incremental load resistance is negative:

2
∂Vou t Pout Vo ut
Ro = ----------------- = – -------------- = – ----------------
∂Iou t Io ut
2 Po ut

The control-to-output transfer function is:

v̂ou R o// r 2
------------t = j 2 ⋅ ---------------------------------------- ,
v̂ c 1 + sC oRo // r 2

while the line-to-output transfer function is:

v̂ou R o// r 2
------------t = g 2 ⋅ ---------------------------------------- .
kˆ 1 + sCo Ro// r 2

In the case of constant power load (PFC stage supplying a cascaded DC-DC converter), unlike TM and
FF-CCM systems, where the parallel combination of r2 and Ro tends to an open circuit, thus placing the
pole of the transfer functions at the origin, in FOT-CCM systems the combination maintains a finite value:

πVout
Ro // r 2 = ----------------- ,
kΓ

so that the pole is still at a finite frequency that moves with the line voltage:

1 kΓ k T OFF
f p = --------------------------------------- = -------------------------------- = ---------------------- .
2π ( Ro // r 2 )Co 2
2π Vou tC o 2π L Co
2

The DC gain of the control-to-output transfer function is:

πkK M K P Vo ut 2 πkK M K P Vou tL

G 0 = v̂out
------------ = j2 ( Ro // r 2 ) = ----------------------- ------------------- = ----------------------- ------------------------------------- .
v̂ C 2 Γ R se ns e 2 T OFF Rsen se
s=0

FOT-CCM PFC pre-design considerations.

Operating frequency and TOFF selection. When specifying the operating frequency of a FOT-controlled
PFC stage, some care must be taken of the resulting minimum ON-time of the power switch TONmin. As it
can be easily derived from table 2, this occurs on the top of the sinusoid at maximum line voltage and its
value is:

1 – k ma x
T ONm in = ---------------------- T OFF . (2)
k m ax

If one wants to ensure the operation described by the large-signal model considered so far over the full
input voltage range, this value has to be greater than the minimum ON-time that the L6562 is able to han-
dle (350 ns max.) plus the MOSFET's turn-on and turn-off delays (assume 150ns total). In wide-range
mains applications this means a minimum value of 6.96µs for TOFF and a maximum operating frequency
(i.e. on the top of the sinusoid) at minimum line voltage not greater than 45 kHz, which could be quite a
limitation, especially in view of a small-size inductor (see next paragraph). Actually, if the condition
TON>TONmin, with TONmin given by (2), is not met in an area around the top of the sinusoid, there will be a
temporary unbalance of the volt·second across the inductor, so that the cycle-by-cycle average inductor

12/30
 AN1792 APPLICATION NOTE

current will grow. If the system is designed so that this occurs at high line, where the inductor current is
considerably lower than its maximum values (less than one half), there is no risk of core saturation.
The most conspicuous consequence will be an increased line current distortion: the energy in excess
around the top of the sinusoid will be balanced by a corresponding reduction across the zero crossings,
so that the current waveform will get narrower and more peaked. As a result, at high line the harmonic
contents will be higher than expected. This increased distortion may be still acceptable, provided the
class-D limits of the EN61000-3-2 standards are not exceeded at rated load and Vin=230Vac, as pre-
scribed by the standard, with some safety margin. However, a slight modification of the control technique
can significantly improve this situation (see "An improvement of FOT control" section) and allow compli-
ance with the standard over a wider load range.
Ripple current (Kr) selection. The selection of the ripple current is a trade-off between inductor size, ease
of EMI filtering and line current distortion. Low values of Kr, which result in low inductor current ripple and
a lighter input line filter for EMI compliance, lead also to a lower line current distortion because the CCM
region inside a line half-cycle is extended, hence making the sinusoidal portion of Iin(θ) more and more
dominant. On the other hand, however, a low Kr results in a larger inductance L and, in the end, in a larger
magnetic core. Typical values for Kr range from 0.25 to 0.40.
Inductor selection. If a ferrite core is to be used, due to the moderate switching frequency, the maximum
flux density Bmax will be typically limited by core saturation and not by core losses. This also means that
transformer's power losses will be located mostly in the windings. Referring to commonly used Area-Prod-
uct formulae [5], a first-cut estimate of the minimum required core size is:
1.31
 1 – k m in Kr Pin 0 T OFF
AP m in ≈ 186  ---------------------------- ⋅ --------------------------- [cm4] .
 k m in Kr B m ax 

This formula accounts also for the tolerance of the current limit threshold of the L6562 (see next section).
At a higher operating frequency core losses might be dominant, but usually the losses related to diode
recovery force to select a switching frequency where core losses do not dominate yet. As to core losses,
there is no worry about the frequency increase with the line voltage pointed out by figure 8 (there is a factor
of three): at high line the inductor ripple, and then the flux swing, becomes very small (typically, 10 times
smaller).
In case the system is operated at a moderate switching frequency, it is worthwhile trying to use ferrous
alloy powder toroidal cores like Kool-Mµ from Magnetics or -xx from Micrometals, especially if working with
quite low Kr values. Typically, core losses will be higher than in a ferrite core and the resulting total effi-
ciency will be not as high as, but in some cases this could be a tolerable penalty.
"Tracking boost" configuration. In some applications it may be advantageous to regulate the output volt-
age of the PFC pre-regulator so that it tracks the RMS input voltage rather than at a fixed value like in
conventional boost pre-regulators. This approach, commonly referred to as "tracking boost" or "follower
boost", brings some benefits, such as reduction of the MOSFET's RMS current and of transformer's size,
and some drawbacks, such as increased diode current, increased low-frequency output ripple and requir-
ing the downstream converter to accommodate a larger input voltage range.
Basically, the tracking boost approach reduces the range of the parameter k increasing its minimum value
kmin. This stated, the benefit of core size reduction becomes apparent if one looks at the above given AP min
formula. Also the frequency change is reduced and the limits concerning TONmin for a full FOT operation
are pushed upwards.
For instance, if the regulated output voltage @Vin=88Vac is set at 250V (and at 400V @Vin=264Vac), kmin
will be 0.498 (instead of 0.311), thus the maximum switching frequency @Vin=88Vac for a full FOT oper-
ation can be as high as 70 kHz; @Vin=88Vac the MOSFET current is reduced by 11% and its conduction
losses by about 22%, whereas the DC and RMS diode current are increased by 60% and 27% respective-
ly; the 2·fL output ripple will be typically increased by about 60%; the minimum AP required for the inductor
core, for the same Kr and Bmax, is reduced by 50%.

13/30
AN1792 APPLICATION NOTE

FOT-CCM PFC design procedure.

The starting point is the electrical design specification, which we assume is as shown in table 5. Actually,
a number of requirements is not directly related to the FOT control technique and will be used exactly like
in the design of standard TM or FF-CCM PFC pre-regulators. Of course, in this context special emphasis
will be given only to the points directly related to FOT control.

Table 5. Basic electrical specification of a FOT-CCM PFC stage

Parameter Symbol
Line voltage range Vin(RMS)min ÷ Vin(RMS)max
Minimum Line frequency fLmin
Regulated output voltage Vout
Rated output power Pout0
Maximum 2fL output voltage ripple ∆Voutpk-pk
Hold-up time TH
Min./max./mean switching frequency (@Vin=Vin(RMS)min, Pout=Pout0) fswmin/fswmax./fsw(m)
Estimated efficiency η
Max. inductor current ripple-to-peak ratio (@Vin=Vin(RMS)min, Pout=Pout0) Kr (= sin θT)
Ambient temperature range Tambmin÷Tambmax

Based on this information and on the results presented in the previous sections, the recommended step-
by-step design procedure is the following:
1) Calculate the range of k (kmin ÷ kmax) associated to the line voltage range:

Vin ( RM S ) m in Vin ( RMS ) m ax

k m in = 2 ⋅ ---------------------------------- , k m ax = 2 ⋅ ----------------------------------- .
Vou t Vou t

2) Calculate the required TOFF from the specification on the switching frequency. If fswmax is specified use:

k m in
T OFF = ----------------
- ;
f swmax

if fswmin is specified use:

k m in
T OFF = Kr ---------------- ;
f swmin

if fsw(m) (average value within a line half-cycle) is specified use:

1 + Kr k m in
T OFF = ---------------- --------------- .
2 f s w(m)

3) Calculate Pin 0=Pout0/η and determine Γ:

Pin 0 4 πKr
Γ = -------------------------- -------------------------------------------------- .
k m in Vo ut 2π – Kr ( 4 + πk m in )

14/30
 AN1792 APPLICATION NOTE

4) From the value of Γ calculate the required inductance L of the boost inductor:

Vou t
L = ------------- T OF F .
Γ

5) Calculate the maximum inductor peak current I Lpkmax using the value of Γ found in step 3:

Pin0 4 π ( 1 – Kr k m in ) 1 – Kr k min
I L pk max = -------------------------- -------------------------------------------------- = Γ ------------------------------ .
k m in Vout 2 π – Kr ( 4 + πk m in ) Kr

6) Determine the maximum sense resistor Rsensemax:

1.6
Rse nse ma x = --------------------- ,
I Lp k max

and select a resistor value Rsense < Rsensemax. 1.6V is the minimum value of the pulse-by-pulse
current limiting threshold on the current sense pin of the L6562. Take into account that the value of
this threshold can go as high as 1.8V, hence the inductor must not saturate up to a current equal to
1.8/Rsense.
7) Calculate the current stress of all components, design the boost inductor with any commonly used pro-
cedure, and select the MOSFET and the diode. Use either the output voltage ripple or the hold-up spec-
ification, whichever gives the higher capacitance value, to select the output capacitor.
8) Design the bias component around the controller IC (multiplier setpoint and feedback) using the same
criteria given for TM systems [6], just considering the different small-signal model as to the feedback
design. Finally design the circuit that sets up FOT control:
8a. Select a capacitor C in the hundred pF or few nF. Like in an FF controller, the tolerance of this capacitor
and its temperature drift will affect the resulting timing.
8b. Select the timing resistor value R from (α):
T OFF
R = 0.7 --------------
C

and pick the closest standard value.

8c. Select the limiting resistor Rs using (β) and the speed-up capacitor Cs using (γ).

A practical FOT-CCM PFC design with the L6562.

As an example, a 375W design, e.g. suitable for a 300W ATX12V PSU will be now described following the
procedure previously given. The electrical spec, listed in table 6, is consistent with ATX12V specification.

Table 6. Electrical specification of the design example of a 375W, L6562-based FOT-CCM PFC stage
Parameter Value
Line voltage range 90 to 265 Vac
Minimum Line frequency 47 Hz
Regulated output voltage 400V
Rated output power 375 W
Maximum 2fL output voltage ripple 20V pk-pk
Hold-up time 17 ms
Maximum switching frequency (@Vin=88Vac, Pout=375W) 100 kHz
Estimated efficiency (@Vin=88Vac, Pout=375W) 90%
Max. inductor current ripple-to-peak ratio (@Vin=88Vac, Pout=375W) 40%
Maximum ambient temperature 50 °C

15/30
AN1792 APPLICATION NOTE

Following the step-by-step procedure:

1) The range of k is:

kmin = 90 2 / 400 = 0.318; kmax = 265 2 / 400 = 0.937.

2) The required OFF-time is:

0.318
T OFF = ------------------------ = 3.18 µs ;
3
100 ⋅ 10
the minimum ON-time will be:
–6 1 – 0.937-
T ONm in = 3.18 ⋅ 10 ----------------------- = 0.21µs ,
0.937

which is shorter than the minimum ON-time handled by the L6562, hence significant additional dis-
tortion of the line current is expected at high line.
3) The expected maximum input power is:

Pin0 = 375/0.9 = 417 W,

and Γ is:

417 4 ⋅ π ⋅ 0.4
Γ = ----------------------------- --------------------------------------------------------------- = 3.85 [A].
0.318 ⋅ 400 2 ⋅ π – 0.4 ( 4 + 0.318 ⋅ π )

4) The required inductance L of the boost inductor is:

–6
400 ⋅ 3.1 ⋅ 10 - = 322µ H
L = --------------------------------------
3.85

that will be rounded up to 330 µH. Using the same value of Kr will give a small additional margin.
5) The maximum inductor peak current is calculated:
1 – 0.4 ⋅ 0.318
I L pkm ax = 3.85 ⋅ ------------------------------------- = 8.4 A .
0.4

6) The maximum sense resistor Rsensemax is:

1.6
Rse nse ma x = -------- = 0.19 Ω ;
8.4

it will be realized with four 0.68Ω, 1W-rated paralleled resistors, for a total resistance of 0.17Ω. The
inductor peak current that the inductor must be able to carry without saturating will be:
1.8
I L pks at = ----------- = 10.6A .
0.17

7) From the formulae in table 4, the MOSFET RMS current is:

417 16 ⋅ 0.318
I Q ( rm s ) = ----------------------------- 2 – -------------------------- = 3.96A ;
0.318 ⋅ 400 3⋅π

the diode RMS current is:

16/30
 AN1792 APPLICATION NOTE

417 16 ⋅ 0.318
I Q ( rm s ) = ----------------------------- -------------------------- = 2.41 A ;
0.318 ⋅ 400 3⋅π

the dissipation on the sense resistor will be 0.17·3.962=2.7W, which justifies the use of four resistors;
the selected MOSFET is the STW26NM50, a 0.12Ω/500V MdmeshTM type from STMicroelectronics,
housed in a TO247 package; the selected diode is an STTH8R06, a 8A/600V Turbo 2 Ultrafast re-
covery rectifier again from STMicroelectronics, housed in a TO220 package. Both of these must be
dissipated to keep their temperature within safe limits.
As for the inductor, assuming a peak flux density Bmax=0.3T, a core with a APmin = 1.91 cm4 is need-
ed and an E42 core has been chosen. The complete inductor spec can be found in the electric circuit
diagram in figure 17.
The output capacitor is determined by the hold-up time requirement. Assuming a minimum voltage of
300V after the line drop, a minimum of 180 µF is needed and a 220µF/450V capacitor will be used.
8) a) A C=560 pF film capacitor is selected.
b) The timing resistor is:
–6
3.18 ⋅ 10
R = 0.7 ----------------------------- = 3975Ω ;
– 12
560 ⋅ 10
the closest standard value (3.9 kΩ) will be selected.
c) The limiting resistor Rs will be such that:
15 – 5.7 – 0.5 – 6 10 – 5.7 – 0.5
------------------------------------------------------- = 768 Ω < R s < 3.9 ⋅ 10 ----------------------------------- = 2600 Ω ;
–3 5.7 5.7
10 ⋅ 10 + -------------------------
–3
3.9 ⋅ 10
since there is no need to stress the clamp a 2.4 kΩ resistor will be chosen. The maximum speed-
up capacitor is:

– 12 5.7
Cs < 560 ⋅ 10 ----------------------------------
- = 362p F ;
15 – 5.7 – 0.5

a 330 pF film capacitor will be used.

The electrical schematic is shown in figure 17. A prototype of this circuit has been realized and evaluated
on the bench. Figures 18 and 19 show a series of diagrams illustrating its performance.

Figure 17. 375W FOT-CCM PFC pre-regulator: electrical schematic.

T1: core E42*21*15, B2 material

D5 D4 NTC
2.0 mm air gap on centre leg, main winding inductance 0.33 mH 1N5406 STTH8R06D 2.5 Ω
Vout = 400V
20 °C /W heat sink Pout = 375 W +
N1: 58 T of 20 x AWG32 (∅ 0.2 mm)
N2: 6 turns of AWG32 N1
R1A R1B D1 C4 R4 R10A
120 kΩ 120 kΩ 1N4148 10 nF 47 Ω 499 kΩ
T1

N2 R10B
DZ1 R5 C5 499 kΩ
R2A 6.8 kΩ
620 kΩ 1N5248B 1 µF
B1
KBU8M + D2
FUSE 8 °C /W heat sink 1N4148
C1 R2B C9
5A/250V
5 °C/W heat sink
STW26NM50

0.47 µF 620 kΩ 2 1 220 µF

8
400V R6 6.8 Ω 450 V
M1

- L6562 7
3
Vac
(88V to 264V) 6 4 C8
CX1 5 D3
CX2 100 nF
0.1 µF 0.33 µF 1N4148 630 V
R8 2.4 kΩ

L3
6 mH C6 330 pF
C3
R3 R9
C2 22 µF
10 kΩ 10nF 3.9 kΩ R11
Note: EMI filter not tested 25V R7A, B, C,D 6.34 kΩ
for EMC compliance
C7 0.68 Ω
560 pF 1W -

17/30
AN1792 APPLICATION NOTE

Figure 18. 375W FOT-CCM PFC pre-regulator: evaluation data.

Eff. [%] Vout [Vdc]

97 395
Pout = 375W Pout = 375W
(*)
96 Pout = 180W 394.5 Pout = 180W

Pout=70W Pout=70W
394
95
393.5
94
(*) 393
93
392.5

92 392
50 100 150 200 250 300 50 100 150 200 250 300
Vin [Vac] Vin [Vac]
(*) Burst-mode operation (*) Burst-mode operation

PF THD [%]
1 80
Pout = 375W Pout = 375W
(*)
Pout = 180W Pout=180W
0.9 Pout=70W 60 Pout = 70W

0.8 40

0.7 20
(*)
0.6 0
50 100 150 200 250 300 50 100 150 200 250 300
Vin [Vac] Vin [Vac]
(*) Burst-mode operation (*) Burst-mode operation

Figure 19. 375W FOT-CCM PFC pre-regulator: conformity to JEIDA-MITI and EN61000-3-2 standards.

Harmonic Current [A] Harmonic Current [A]

10 10
Measurement JEIDA-MITI class D limits Measurement EN61000-3-2 class D limits
3
1
1

0.1 0.3
0.1
0.01 0.03
0.01
0.001
0.003
0.0001 0.001
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39
Harmonic Order [n] Harmonic Order [n]
Vin = 100 Vac, 50 Hz; Pout = 375 W THD = 9.2% PF = 0.996 Vin = 230 Vac, 50 Hz; Pout = 375 W THD = 28.9% PF = 0.957

Figures 20 to 22 show the line current waveforms under the operating conditions considered in the dia-
grams of figure 18 (100%, 50% and 20% of the rated load) at nominal voltage of both US and European
mains. Note: the waveforms are taken with an acquisition method that eleminates high frequency ripple,
just to show the low frequency component, the "odd counterpart" of the average inductor current.
Figure 23 shows a close image of the line current waveform under two different operating conditions: on
the left (a), the waveform is taken at low line, and one can easily recognize the shape as well as the DCM
and the CCM portions as per the theory; on the right (b), the waveform, taken at high line, clearly shows
the distortion due to the condition TON>TONmin, with TONmin given by (2), not met throughout a line cycle.
Note that in the portion where this occurs, the system is running at a fixed frequency fsw= 1/(TONmin+TOFF).
The theoretical current, which assumes it is always TON>TONmin and does not exhibit additional distortion,
is shown in red.

18/30
 AN1792 APPLICATION NOTE

Figure 20. 375W FOT-CCM PFC pre-regulator: line current waveforms @ Pout=375W.

Ch1: Input voltage (after the bridge rectifier) Ch1: Input voltage (after the bridge rectifier)
Ch2: Line current (@Vin=115Vac, Pout=375W) Ch2: Line current (@Vin=230Vac, Pout=375W)

Figure 21. 375W FOT-CCM PFC pre-regulator: line current waveforms @ Pout=180W.

Ch1: Input voltage (after the bridge rectifier) Ch1: Input voltage (after the bridge rectifier)
Ch2: Line current (@Vin=115Vac, Pout=180W) Ch2: Line current (@Vin=230Vac, Pout=180W)

Figure 22. 375W FOT-CCM PFC pre-regulator: line current waveforms @ Pout=70W.

Ch1: Input voltage (after the bridge rectifier) Ch1: Input voltage (after the bridge rectifier)
Ch2: Line current (@Vin=115Vac, Pout=70W) Ch2: Line current (@Vin=230Vac, Pout=70W)

19/30
AN1792 APPLICATION NOTE

Figure 23. Line current: a) at low line; b) at high line, with the additional distortion due to violation
of the condition TON>TONmin for some part of the line cycle (in red the theoretical curve).

Vin=115Vac Vin=230Vac
Pout=180W Pout=375W

ONminON
T
<T
CCM
DCM FF DCM
CCM DCM CCM DCM CCM

CCM

CCM

CCM

CCM
a) b)

An improvement of FOT control

The most severe limitation of FOT control seems to be the trade-off between operating frequency and dis-
tortion of the line current waveform pointed out in the "Operating frequency and TOFF selection" sub-sec-
tion. As a matter of fact, if TOFF is selected for an operating frequency above 60-70 kHz on the top of the
sinusoid at minimum line voltage, the harmonics of the line current may exceed the class-D limits of
EN61000-3-2 at maximum line voltage (264Vac). Actually the standard prescribes the measurement of
the harmonic current emissions to be done at the nominal voltage (230 Vac) and rated load but sometimes
the PSU is specified to stay within the limits under a range of operating conditions, e.g. throughout the
input voltage range and/or from rated load down to a minimum value. In this case a trade-off could be dif-
ficult or even impossible to find.
A simple modification of the standard Fixed-Off-Time technique can overcome this possible issue. The
idea behind is to make TOFF a function of the line voltage, so that at high line it is long enough to ensure
that the condition TON>TONmin, with TONmin given by (2), is met throughout the entire line cycle. Probably
the simplest way to do so is shown in figure 24a, where TOFF is made a function of the instantaneous line
voltage. The circuit of figure 24b makes TOFF a function of the RMS line voltage thanks to the peak-holding
effect of T1 (which acts as a buffer) along with Ra and Ca whose time constant is significantly longer than
a line half cycle. In the following, reference will be made only to the circuit of figure 24a.
With the addition of R2 and T, as long as the voltage on the ZCD pin during TOFF is above VMULT+VBE, C
is discharged through R1 and R2, following the law:
t
– -----------------------------
R1 ( R1//R2 ) C R1
V Z CD ( t ) = V ZCDc lam p – ---------------------- ( V M UL T + V BE ) e + ---------------------- ( V M ULT + V BE ) ;
R1 + R2 R 1 + R2

as VZCD(t) falls below VMULT+VBE, T is cut off and C is discharged through R1 only, so that its evolution
from that point on is described by:
t
– ------------
R1 R1 C
V Z CD ( t ) = ---------------------- ( V MUL T + V BE )e .
R 1 + R2

20/30
 AN1792 APPLICATION NOTE

Figure 24. Improved FOT control; TOFF changes with: a) the instantaneous, b) the RMS line voltage.
Rectified Rectified
input voltage input voltage
Vcc
M M
GD 8 GD
7 7

MULT
3
L6562 4
CS MULT
3
L6562 4
CS

5 Rsense 5
Rsense
ZCD Rs T1 R2 ZCD Rs
T
BC557 BC547
R2 D T2 D
1N4148 BC557 1N4148
Cs Ra Ca Cs
R1 C
68kΩ 1 µF R1 C

a) b)

In this way, once fixed the multiplier operating point (that is, the VMULT /Vin ratio), with a proper selection
of R1 and R2 it is possible to increase TOFF with the line voltage so that at maximum line voltage it is al-
ways TON>TONmin. It is easy to see that TOFF is now a function of the instantaneous line voltage. We will
refer to this technique as "Line-modulated Fixed-Off-Time" (LM-FOT).
This modification, though simple, introduces profound changes in the timing relationships, with a positive
influence on the energetic relationships. From the control point of view, modulating T OFF is a feedforward
term that modifies the gain but does not change its characteristics. Then all of the properties of the stan-
dard FOT control are maintained.
Due to the highly non-linear nature of the T OFF modulation introduced by T and R2, its effects will be dis-
cussed only qualitatively and the quantitative aspects will be provided graphically for a specific case.
Figure 25a shows the dependence of T OFF upon the time constant τ = (R1//R2)·C, TOFF = K2·τ, for differ-
ent values of the (R1, R2) divider ratio K1 = R1/(R1+R2). In this diagram and the following ones, it is as-
sumed VBE = 0.55V. As long as it is VMULT+VBE < VZCDtrigger, that is VMULT < 0.85V, there is little effect
on TOFF, while the effect becomes considerable for higher values of VMULT. Figure 25b shows how TOFF
(normalized to the value on the top of the sinusoid at minimum line voltage) changes along a line half-cycle
at minimum and maximum line voltage for the particular case of K1=0.9. The multiplier peak voltage
@Vin=88Vac is assumed to be 1V, resulting in TOFF@Vin=264Vac 2.5 times longer.

Figure 25. LM-FOT: a) TOFF modulation with the circuit of fig. 25a; b) TOFF change in a line half-cycle.

20 2.5
K1=0.95 Vin=264Vac K1=0.9
R1 2
K1 =
15 R1+ R2
K1=0.9 1.5
K2 10 TOFF
K1=0.85 TOFF @88Vac Vin=88Vac
1

5
K1=0.8 0.5

0 0
0 0.5 1 1.5 2 2.5 3 0 0.52 1.05 1.57 2.09 2.62 3.14
V MULT θ

a) b)

Figures 26 to 32 are the analogous of figures 8-14: they refer to the same representative system, with the
addition of TOFF modulation. The value of TOFF on the top of the line voltage sinusoid at minimum mains
is the same as in the system of figures 8-14, so that the switching frequency is the same in that point.
Quantities are normalized with the same criteria.

21/30
AN1792 APPLICATION NOTE

At low line the switching frequency dos not change much, however, the noticeable point is that, since TOFF
gets shorter next to the line voltage zero-crossings, the CCM portion in a line half cycle is enlarged. The
result is that peak currents will be lower and all RMS currents will be closer to those of a FF-CCM system.
Also line current shape will get closer to a perfect sinusoid, then both PF and THD will improve (see figure
32).
At high line the switching frequency, which stays close to the minimum around the top of the sinusoid, is only
about 20% higher than that at minimum line (instead of 3 times). Efficiency will definitely benefit from that.
Around zero-crossings, the frequency increases about 2.5 times, but the power handled in those regions is
not high, then the erosion on the efficiency improvement will be minimum. The minimum ON-time goes from
7.1% to 17.8% of the normalized TOFF value but the CCM portion gets narrower: the THD will improve be-
cause it is now easy to have TON>TONmin, hence eliminating the related additional distortion, but the achiev-
able value is slightly worse. However, diode reverse recovery-related losses will be further reduced, in favor
of efficiency. Line current wave shape changes little. Note that the maximum inductor current ripple changes
little as well but in a line half-cycle it does not occur on the CCM-DCM boundary any more.

Figure 26. Normalized switching frequency vs. phase angle: a) Full load Pin0; b) Pin0/2; c) Pin0/10
fsw(θ) fsw (θ)
fsw (θ)
fsw0 fsw0
fsw0
6
3
4
4
2
2
2
1

88 88 88
π π π
Vin Vin Vin
θ θ θ
264
264 0 264 0 0

a) b) c)

Figure 27. Normalized MOSFET's ON-time vs. phase angle: a) Full load Pin0; b) Pin0/2; c) Pin0/10
TON (θ) TON (θ) TON (θ)
TOFF TOFF TOFF

6
2
3
4
2
1
2
1

88 88 88
π π π
Vin Vin Vin
θ θ θ
264 264 264
0 0 0

a) b) c)

Figure 28. Inductor current conduction angle vs. phase angle: a) Full load Pin0; b) Pin0/2; c) Pin0/10

DL (θ) DL (θ) DL (θ)

1 1

0.8 0.8
0.8
0.6 0.6
0.6
0.4
0.4
0.4
0.2
0.2
88 88 88
Vin π Vin π π
Vin
θ θ θ
264 264 264
0 0 0

a) b) c)

22/30
 AN1792 APPLICATION NOTE

Figure 29. Line current vs. phase angle: a) Full load Pin0; b) Pin0/2; c) Pin0/10
Iin(θ) Iin(θ) Iin(θ)
ILpk0 I Lpk0 ILpk0
0.8 0.4 0.08

0.6 0.3 0.06

0.4 0.2 0.04

0.2 0.1 0.02

0 0 0
88 88 88

Vin Vin Vin

π π π
264 0 θ 264 θ 264 θ
0 0

a) b) c)

Figure 30. Inductor current ripple vs. phase angle: a) Full load Pin0; b) Pin0/2; c) Pin0/10
∆IL(θ) ∆IL(θ) ∆IL(q)
∆IL0 ∆I L0 ∆IL0
0.2
0.4 0.4
0.3 0.15
0.3
0.2 0.1
0.2
0.1 0.1 0.05

0 0 0
88 88 88
Vin π π Vin π
Vin
θ θ θ
264 264 264
0 0 0

a) b) c)

Figure 31. a) Normalized Inductor peak current; b) Normalized line peak current; c) Transition angle

ILpk I pk
ILpk0 ILpk0
1.0 0.8
0.8 θT (°)
0.6
0.6 90
0.4
0.4 60
0.2
0.2
30 0
88 88
Pin
0 0
Vin 0 Vin Pin0
Pin Pin
88
Pin0 Pin0 Vin 1
264 264 1
1 264
a) b) H c)

Figure 32. a) Normalized line RMS current; b) Power Factor; c) Total Harmonic Distortion

Iin(rms)
Iin(rms)0 PF
1.0 1.0 THD%
0.8
0.6
0.95
40
0.4
0.2
20 0
0.90
88 88
Pin
0 0 88 Pin0
Vin Vin
Pin Pin
Vin
Pin0 Pin0
264 1 264 1 1
264
a) b) c)

23/30
AN1792 APPLICATION NOTE

The starting point for the design of the circuit of figure 25a is the pair of the desired values for TOFF on the
top of the line voltage sinusoid at minimum (TOFF@Vin=Vin(RMS)min) and maximum line
(TOFF@Vin=Vin(RMS)max). Let ρ their ratio: ρ = TOFF@Vin=Vin(RMS)max / TOFF@Vin=Vin(RMS)min. With
these data the design can be done with the aid of the diagrams in figure 33. The diagrams, which assume
an input voltage range 88-264 Vac, features two sets of curves, ρ curves and K2 curves, each of them
plotted for a particular value of the peak multiplier voltage VMULTpk @88Vac.
Figure 33. Diagrams for the design of the circuit of figure 24a (valid for Vacmin=88V, Vacmax=264V).
4 2
VMULTpk = 0.7
K2 curves VMULTpk = 0.7 0.8V
2.4
ρ curves
0.8V 0.9V
3.5 2.8
0.9V
1.0V 1.0V 3.2
3 3.6

ρ P2 4
K2
2.5 4.4
P1
4.8

2 5.2
5.6

1.5 6
0.8 0.810.820.830.840.850.860.870.880.89 0.9 0.910.920.930.940.95
K1

The suggested step-by-step design procedure is the following:

1) Determine the multiplier setpoint using the same criteria given for TM systems [6] and take note of the
resulting VMULTpk @88Vac value. It should be between 0.7 and 1V.
2) Draw an horizontal line located at the desired value of ρ (on the left vertical axis) as long as it intercepts
the curve relevant to the value VMULTpk previously determined in P1. The abscissa of P1 gives the nec-
essary value of the parameter K1.
3) From P1 draw a vertical line as long as it intercepts the K2 curve relevant to the same value of VMULTpk
in P2. The ordinate of P2 (on the right vertical axis) gives the necessary value of K2.
4) Calculate the time constant τ =(R1//R2) C necessary to achieve the desired TOFF@88Vac:
T OFF @ 88Vac
τ = ----------------------------------------
- .
K2
5) Select a capacitor C in the hundred pF or few nF, and determine the required resistance value:
τ
R' = R1//R2 = ---- .
C
6) Determine R1 and R2:
R' -
R1 = ---------------- R' .
R2 = ------- (23a, b)
1 – K1 K1
7) Select the limiting resistor Rs according to the following inequalities:
V GDx – V ZCDc lam p – V F ( V GD – V ZCDc lam p – V F )R 1R2
--------------------------------------------------------------- < Rs < --------------------------------------------------------------------------------------------------------
- ,
V ZCDcla m p ( V ZCDcla m p – V BE )R1 + V ZCDc lam p R2
I ZCDx + -----------------------------
R1
and the speed-up capacitor Cs using (γ).
The 375W design will now be modified to implement LM-FOT. The target is to keep TOFF = 3.18 µs
@90Vac and to get TOFF = 8 µs @265Vac, thereby ρ = 8/3.18 ≈2.5. The input voltage range, although
different from, is very close to that covered by the diagrams of figure 33, then these curves are still appli-

24/30
 AN1792 APPLICATION NOTE

cable. Following the above given step-by-step procedure:

1) From the part values in the schematic diagram of figure 17, VMULTpk @90Vac=1.02V, then the curves
for VMULTpk =1V will be considered.
2) In figure 33 the points P1 and P2 are shown; the resulting values for K1 and K2 are 0.891 and 4.17 re-
spectively.
3) The required time constant is:
–6
3.18 ⋅ 10 –6
τ = ---------------------------- = 0.76 ⋅ 10 s .
4.17
4) The same capacitor (C=560 pF) will be used, then the associated resistance value will be:
–6
0.76 ⋅ 10 - = 1357 Ω .
R' = ----------------------------
– 12
560 ⋅ 10
5) R1 and R2 will be respectively:
1357 - = 12450Ω
R1 = ----------------------- 1357- = 1523 Ω ;
R2 = --------------
1 – 0.891 0.891
the standard values R1=12kΩ and R2=1.5kΩ will be chosen.
6) Under the worst-case condition, only R1 will divert some current from the ZCD pin to ground, then the
limiting resistor Rs must be selected according to:
3 3
15 – 5.7 – 0.5 ( 10 – 5.7 – 0.5 )1.5 ⋅ 10 12 ⋅ 10
----------------------------------------------------- = 840 Ω < Rs < --------------------------------------------------------------------------------------------- = 972 Ω ;
–3 5.7 - 3 3
10 ⋅ 10 + ---------------------- ( 5.7 – 0.55 )12 ⋅ 10 + 5.7 ⋅ 1.5 ⋅ 10
–3
12 ⋅ 10
in this case a 910Ω resistor will be chosen. Cs will be the same 330 pF capacitor.
The prototype, modified as per the above guidelines, has been evaluated to check the actual differences
with the former realization. Figures 34 to 38 show a series of diagrams illustrating its performance.

Figure 34. 375W LM-FOT-CCM PFC pre-regulator: evaluation data.

Eff. [%] Vout [Vdc]

98 394
Pout = 375W Pout = 375W

97 Pout = 180W 393.5 Pout = 180W

Pout=70W Pout=70W
96 393

95 392.5

94 392

93 391.5

92 391
50 100 150 200 250 300 50 100 150 200 250 300
Vin [Vac] Vin [Vac]

PF THD [%]
1 50
Pout = 375W Pout = 375W
0.98 Pout = 180W Pout=180W
40
0.96 Pout=70W Pout = 70W

0.94 30

0.92 20
0.9
10
0.88

0.86 0
50 100 150 200 250 300 50 100 150 200 250 300
Vin [Vac] Vin [Vac]

25/30
AN1792 APPLICATION NOTE

Comparing these results with those in figure 18, the full-load efficiency does not change significantly at low
line, but improves by as much as 1% @230Vac and 1.5%@265Vac, getting close to 98%; also half-load effi-
ciency exceeds 97% @230Vac; full-load PF at maximum line voltage increases from 0.8 to 0.91 thanks to a
THD% reduction from 60% to 41%; even more significant is the THD% reduction at light load, which makes
confident that EN61000-3-2 compliancy will be achieved not only at rated load but also at light load.
Figure 35. 375W LM-FOT-CCM PFC pre-regulator: conformity to JEIDA-MITI & EN61000-3-2 standards.
Harmonic Current [A] Harmonic Current [A]

Measurement JEIDA-MITI class D limits Measurement EN61000-3-2 class D limits

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39
Harmonic Order [n] Harmonic Order [n]

Vin = 100 Vac, 50 Hz; Pout = 375 W THD = 3.4% PF = 0.999 Vin = 230 Vac, 50 Hz; Pout = 375 W THD = 16.9% PF = 0.985

Figure 36. 375W LM-FOT-CCM PFC pre-regulator: harmonic emissions at half load (180W).
Harmonic Current [A] Harmonic Current [A]

Measurement JEIDA-MITI class D limits Measurement EN61000-3-2 class D limits

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39
Harmonic Order [n] Harmonic Order [n]

Vin = 100 Vac, 50 Hz; Pout = 180 W THD = 3.7% PF = 0.999 Vin = 230 Vac, 50 Hz; Pout = 180 W THD = 20.5% PF = 0.976

Figure 37. 375W LM-FOT-CCM PFC pre-regulator: harmonic emissions at light load (70W).
Harmonic Current [A] Harmonic Current [A]

Measurement JEIDA-MITI class D limits Measurement EN61000-3-2 class D limits

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39
Harmonic Order [n] Harmonic Order [n]

Vin = 100 Vac, 50 Hz; Pout = 70 W THD = 6.6% PF = 0.998 Vin = 230 Vac, 50 Hz; Pout = 70 W THD = 18.7% PF = 0.964

Figure 38. 375W LM-FOT-CCM PFC pre-regulator: harmonic emissions at 265Vac.

Harmonic Current [A] Harmonic Current [A] Harmonic Current [A]

Measurement EN61000-3-2 class D limits Measurement EN61000-3-2 class D limits Measurement EN61000-3-2 class D limits

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39
Harmonic Order [n] Harmonic Order [n] Harmonic Order [n]

Vin = 265 Vac, 50 Hz; Pout = 375 W THD = 41.4% PF = 0.908 Vin = 265 Vac, 50 Hz; Pout = 180 W THD = 45.7% PF = 0.883 Vin = 265 Vac, 50 Hz; Pout = 70 W THD = 31.6% PF = 0.919

26/30
 AN1792 APPLICATION NOTE

As to this point, the harmonic analysis shown in figures 35 to 38 confirms that. In particular, figure 38
shows that the emissions are within the limits even at light load and maximum line voltage. From the total
distortion point of view, half-load seems to be the worst-case condition.

Figure 39. 375W LM-FOT-CCM PFC pre-regulator: line current waveforms @ Pout=375W.

Ch1: Input voltage (after the bridge rectifier) Ch1: Input voltage (after the bridge rectifier)
Ch2: Line current (@Vin=115Vac, Pout=375W) Ch2: Line current (@Vin=230Vac, Pout=375W)

Figure 40. 375W LM-FOT-CCM PFC pre-regulator: line current waveforms @ Pout=180W.

Ch1: Input voltage (after the bridge rectifier) Ch1: Input voltage (after the bridge rectifier)
Ch2: Line current (@Vin=115Vac, Pout=180W) Ch2: Line current (@Vin=230Vac, Pout=180W)

Figure 41. 375W LM-FOT-CCM PFC pre-regulator: line current waveforms @ Pout=70W.

Ch1: Input voltage (after the bridge rectifier) Ch1: Input voltage (after the bridge rectifier)
Ch2: Line current (@Vin=115Vac, Pout=70W) Ch2: Line current (@Vin=230Vac, Pout=70W)

27/30
AN1792 APPLICATION NOTE

Figures 39 to 41 show the line current waveforms under the operating conditions considered in the dia-
grams of figure 34 (100%, 50% and 20% of the rated load) at nominal voltage of both US and European
mains. It is interesting to compare these waveforms with those in figures 20 to 22. The effect of the line
modulation of TOFF is conspicuous.

Figure 42. Line current a) nearly sinusoidal at low line; b) without additional distortion at high line.

Vin=115Vac Vin=230Vac
Pout=180W Pout=375W

CCM DCM CCM DCM CCM

a) b)

This effect is clearly pointed out in figure 42 as well. It shows a close image of the line current waveform
under the same two operating conditions as in figure 23. On the left (a), the waveform is taken at low line,
half-load and one can easily recognize that it is very close to a sinusoid (its THD is 3.8%); this result is
even better than the theoretical prediction because the theoretical CCM-DCM boundary falls in a region
where the effect of the THD optimizer circuit of the L6562 becomes apparent (TON is forced to be longer
than the value commanded by the control loop). As a consequence, in that region the peak inductor cur-
rent is higher and the actual CCM-DCM boundary is virtually located at or extremely close to zero-cross-
ings.
On the right (b), the waveform is taken at high line, full-load and in this case the waveform closely follows
the theoretical one, unlike that in figure 23b, thus proving the effectiveness of a longer TOFF.
Considering that at low line the system works almost entirely in CCM, so that the line current looks very
much like that of a FF-CCM system, it is possible to face the design of the power stage with the same ap-
proach used in an FF-CCM type. The significant difference is the point where inductor current ripple ampli-
tude is maximum: whereas it is well-defined with FF-CCM, it is not with LM-FOT and it is not possible to
find a simple design formula that relates the inductance value L to the maximum desired ripple. As a rule
of thumb, it is possible to refer to the ripple amplitude on the top of the sinusoid at minimum line voltage
and determine L so that the ripple amplitude is 75% of the maximum desired.
This stated, the recommended step-by-step design procedure of an LM-FOT-controlled PFC is the following:
1) Calculate the range of k (kmin ÷ kmax) associated to the line voltage range:
Vin ( RM S ) m in Vin ( RMS ) m ax
k m in = 2 ---------------------------------- , k m ax = 2 ----------------------------------- .
Vou t Vou t

2) Calculate the required TOFFmin from the specification on the maximum switching frequency (on the top
of the line voltage sinusoid) fswmax at minimum line voltage:
k mi n
T OFFm in = ------------------
- .
f s w max

28/30
 AN1792 APPLICATION NOTE

3) Calculate Pin 0=Pout0/η and determine the maximum line peak current Ipk:

2Pin0
lp k m ax = -------------------------
- .
k min Vo ut

4) Determine the ripple amplitude on the top of the sinusoid at minimum line voltage, assuming it is 75% of
the maximum specified, related to Kr:
6Kr 12 Kr Pin0
∆I Lpk = ------------------- lpk ma x = ------------------- -------------------------- .
8 – 3Kr 8 – 3 Kr k m in Vout

5) Determine the required inductance L of the boost inductor:

2
Vo ut 8 – 3Kr Vou t
L = ( 1 – k m in ) -------------- T OFFm in = ------------------- ---------------- k m in ( 1 – k m in )T OF Fm in .
∆I L pk 12Kr Pin0

6) Calculate ILpkmax:
1 8 16 Pin 0
I L pkm ax = lpk ma x + --- ∆I Lpk = ------------------- lpk ma x = ------------------- -------------------------- .
2 8 – 3 Kr 8 – 3 Kr k m in Vout

7) Determine the maximum sense resistor Rsensemax:

1.6
Rse nse ma x = -----------------
l Pkm ax

and select a resistor value Rsense < Rsensemax. 1.6V is the minimum value of the pulse-by-pulse
current limiting threshold on the current sense pin of the L6562. Take into account that the value of
this threshold can go as high as 1.8V, hence the inductor must not saturate up to a current equal to
1.8/Rsense.
8) Calculate the current stress of all components, design the boost inductor with any commonly used pro-
cedure, and select the MOSFET and the diode. Use either the output voltage ripple or the hold-up spec-
ification, whichever gives the higher capacitance value, to select the output capacitor.
9) Design the bias component around the controller IC (multiplier setpoint and feedback) using the same
criteria given for TM systems [6], just considering the different small-signal model as to the feedback de-
sign. Finally, once specified TOFF@Vin=Vin(RMS)max > 7 µs, design the circuit that sets up LM-FOT con-
trol according to the given procedure.

Conclusions
Fixed-Off-Time control of PFC pre-regulators has been discussed and analyzed, highlighting its merits
and drawbacks. Simplified large-signal and small-signal models have been described and a design pro-
cedure has been provided for its basic implementation. A prototype built following this procedure and uti-
lizing the L6562, a cheap 8-pin Transition-Mode controller IC with some simple additions, has been
evaluated on the bench and the results have been presented.
An improved version, denominated "Line-modulated Fixed-Off-Time" (LM-FOT) control, has been intro-
duced and its benefits over the basic FOT control have been shown and proven by the bench evaluation
of the prototype modified accordingly. In particular, from these results it is possible to conclude that an LM-
FOT-controlled PFC pre-regulator, operated in CCM, at low line performs exactly like a conventional FF-
CCM type. Thereby, the power stage can be designed with the same procedure and using the same rela-
tionships.
The good performance that the system has shown, especially in the improved LM version, validates this
approach, which couples the simplicity and cost-effectiveness of TM operation with the high-current capa-
bility of CCM operation. FOT control can justifiably be added to these two popular techniques.

29/30
 AN1792 APPLICATION NOTE

REFERENCES
[1] R. B. Ridley "A New Continuous-Time Model for Current-Mode Control with Constant Frequency,
Constant On-Time and Constant Off-Time, in CCM and DCM", IEEE Power Electronics Specialists
Conference Record, San Antonio, Texas, pp. 382-389, 1990.
[2] D. H. Venable "Current Mode Control", Venable Technical Paper #5, www.venableind.com
[3] C. Adragna "Fixed-Off-Time Control of PFC Pre-regulators", 10 th European Conference on Power
Electronics and Applications, EPE2003, Toulouse, France, paper 382.
[4] "L6562, Transition-Mode PFC Controller", Datasheet, www.st.com
[5] Lloyd H. Dixon, Jr. "Filter Inductor and Flyback Transformer Design for Switching Power Supplies",
UNITRODE Power Supply Design Seminar Manual, 1994 (SEM-1000)
[6] "L6561, Enhanced Transition Mode Power Factor Corrector", AN966, www.st.com

Information furnished is believed to be accurate and reliable. However, STMicroelectronics assumes no responsibility for the consequences
of use of such information nor for any infringement of patents or other rights of third parties which may result from its use. No license is granted
by implication or otherwise under any patent or patent rights of STMicroelectronics. Specifications mentioned in this publication are subject
to change without notice. This publication supersedes and replaces all information previously supplied. STMicroelectronics products are not
authorized for use as critical components in life support devices or systems without express written approval of STMicroelectronics.

The ST logo is a registered trademark of STMicroelectronics.

All other names are the property of their respective owners

© 2003 STMicroelectronics - All rights reserved

STMicroelectronics GROUP OF COMPANIES

Australia - Belgium - Brazil - Canada - China - Czech Republic - Finland - France - Germany - Hong Kong - India - Israel - Italy - Japan -
Malaysia - Malta - Morocco - Singapore - Spain - Sweden - Switzerland - United Kingdom - United States
www.st.com

30/30

View publication stats