Application Note AN6032
FAN4800 Combo Controller Applications
www.fairchildsemi.com
General Description
This application note shows the step-by-step process to
design a high-performance supply. The equations shown in
this document can also be used for different output voltages
and total power.
The complete power supply circuits shown in Figures 6 and
7 demonstrate the FAN4800’s ability to manage high output
power while complying with international requirements
regarding AC line quality. The PFC section provides
380V
The output of the converter delivers +12V at up to 8.4 amps.
The circuit operates from 85 to 265V
tions switching at 100kHz.
to a dual-transistor current-mode forward converter.
DC
with both power sec-
AC
D2
D1
1N5406
L1
1
2
3
4
5
6
7
8
D9
MBRS
140
R12
71.5k
IEAO
I
AC
I
SENSE
V
RMS
SS
V
DC
RAMP1
RAMP2
ISL9R460P2
FQPF9N50
NOT USED
U1
FAN4800
PFC OUT
PWM OUT
Q1
41.7k
DC I
R6
R10
6.2k
AC INPUT
85 TO 265Vac
R5A
1.2
RAMP1
R5B
1.2
D12
1N5401
D13
1N5401
I
F1
3.15A
0.68uF
SENSE
R5C
1.2
BR1
4A, 600V
C26
100nF
0.1uF
KBL06
R2A
453k
R2B
453k
C3
R3
110k
C2
0.47uF
R4
15.4k
C19
C18
1uF
470pF
C1
R5D
1.2
R31
100
R1B
500k
R1A
500k
10nF
C11
R27
75k
C30
330uF
25V
R21
Q1G
22
10nF
R28
240
C7
C6
1.5nF
VEAO
V
V
GND
LIMIT
C4
16
15
V
FB
14
REF
13
CC
12
11
10
9
The PFC Stage
Powering the FAN4800
The FAN4800 is initialized once C12 is charged to 13V
through R
age on C
well-regulated 13V for the FAN4800 from its secondary
winding. T
N
SEC
bypassing with low-ESR ceramic or film capacitors on V
and V
is achieved when D
peak AC line voltage before the boost switch Q
on. This ensures the boost inductor current is zero before
100uF
450V
C5
C12
10uF
35V
C17
220pF
D8
MBRS
140
R7A
178k
R7B
178k
and R28. PFC switching action boosts the volt-
27
to 380V via L1’s inductance. T2 then supplies a
5
’s primary-to-secondary turns ratio (N
2
) is 18.8:1. For proper circuit operation, high-frequency
is provided. Orderly PFC operation upon start-up
REF
R8
2.37k
V
FB
MBRS
C31
1nF
quick charges the boost capacitor to the
2
D3
RGF1J
Q2G
R30
4.7k
D7
MMBZ5245B
Q3G
R19
220
MMBZ5245B
V
CC
C16
1uF
R20A
2.2
0.1uF
FQPF
6N50
FQPF6N50
R9
1.1k
D4
C13
Q2
RGF1J
Q3
R20B
2.2
R11
845k
C14
1uF
D5
T2
D6
RGF1J
RAMP2 / DC I
Q4
MMBT3904
V
REF
10nF
C8
68nF
R17
33
C25
0.1uF
T1B
R14
33
C20
1uF
R15
3
T1A
D10
140
C15
10nF
VDC / +380V
LIMIT
R13
10k
C10
15uF
R16
10k
C9
PRI GND
1
V
DC
/
PRI
CC
is turned
Figure 1. The PFC Stage
© 2006 Fairchild Semiconductor Corporation www.fairchildsemi.com
FAN4800 Rev. 1.0.2 • 12/07/06
AN6032 APPLICATION NOTE
PFC action begins. The value of the regulated voltage on C
must always be greater than the peak value of the maximum
line voltage delivered to the supply.
>
V2V
C5 in(rms_max)
>×
V 1.414 265
()()
C5
>
V 375V use 380V
C5
1
Because the FAN4800 uses transconductance amplifiers, the
loop compensation networks are returned to ground (see the
FAN4800 datasheet for the error amplifier characteristics/
advantages). This eliminates the interaction of the resistive
divider network with the loop compensation capacitors, permitting a wide choice of divider values chosen to minimize
amplifier offset voltages due to input bias currents. For reliable operation, R
and R7B must have a voltage rating of at
7A
least 400 volts.
Calculate the divider ratio (R
+
RRV
7A 7B C5
R2.5
8
+
R R 380
7A 7B
R2.5
8
+
RR
7A 7B
R
8
7A+R7B
)/R8 by:
=−
1
=−
=
1
151
2
Selecting the Power Components
The FAN4800 PFC section operates with continuous inductor current to minimize peak current and to maximize available power. The boost inductor value found by setting ΔI, the
peak-to-peak value of high-frequency current, is typically
10% to 20% of the peak value of the maximum line current.
2P
in(max)
I
in( peak _ max)
P
in(max)
where I
in(peak_max)
at low line, V
age, P
O(max)
ciency. Value I
is a peak value of input current occurred
in(rms_min)
is the maximum output power, and η is effi-
in(peak_max)
the specified percentage rate. I
mum current.
Δ
IdII
II
L(max) in( peak _ max)
Duty cycle D and switching frequency f
selection.
=
V
in( rms _ min)
P
O(max)
=
η
is RMS value of minimum line volt-
defines value of ΔI, where dI is
is the inductor maxi-
L(max)
=×
in( peak _ max)
Δ
=+
I
2
influence inductor
S
4
5
5
−
V2V
Oin(rms_min)
=
)
The boost diode D
D
=
L
1
=
=
=
3.128mH use 3.0mH
V
O
×
D2V
in( rms _ min)
×
fI
Δ
S
−⋅⋅
V2V V
()
O in( rms _min) in( rms _ min)
⋅⋅⋅
VfdIP
OS O(max)
−⋅⋅⋅
380 1.414 85 85 0.95
()()
{}
380 1 10 0.15 100
()
and switch Q1 are chosen with a reverse
1
5
⋅× ⋅ ⋅
()
()()
2
2
()( )
voltage rating of 500V to safely withstand the 380V boost
potential. The maximum Q
Equation 8 and the maximum Q
RMS current is obtained by
1
peak current is calculated
1
by Equation 9.
42V
=−
I2I
Q1rms in( rms _max)
2P 4 2V
O(max) in( rms _ min)
=−
ηπ
V23V
in( rms _ min) O
1.414 100 4 1.414 85
()()
=−
1.06 A
II
2.025 A
0.95 85 2 3 3.1416 380
()()
=
=+
Q1peak in( peak_ max)
2P
O(max)
=
V
η
in( rms
1.414 100
()()
=+
0.95 85
()()
=
1
23V
1
⋅⋅⋅
⋅⋅⋅
I
Δ
2
V2V 2V
()
+
_min) O S 1
⋅
⋅
in( rms _ min)
π
O
()()
1
()()
−⋅
Oin(rms_min)in(rms_min)
⋅⋅
VfL
−⋅⋅⋅
380 1.414 85 1.414 85
()()
{}
⋅× ⋅×
380 1 10 3 10
()
()( )
()()
53
The boost diode average current can be calculated by:
=
II
D1avg O(max)
P
O(max)
=
V
O
100
==
0.26 A
380
The boost capacitor value is chosen to permit a given output
voltage hold-up time in the event the line voltage is suddenly
removed.
2P t
≥
C
5
O(max) HLD
22
VV
C5(NOM ) C5(MIN )
−
where:
t
= hold-up time (sec)
HLD
V
= minimum voltage on C5 at which the PWM stage
C5(min)
can still deliver full output power
7
)
8
9
−
10
)
11
)
© 2006 Fairchild Semiconductor Corporation www.fairchildsemi.com
FAN4800 Rev. 1.0.2 • 12/07/06 2
AN6032 APPLICATION NOTE
A key advantage of using leading/trailing-edge modulation
is that a large portion of the inductor current is "dumped"
directly into the load (PWM stage transformer) and not the
boost capacitor. This relaxes the ESR requirement of the
boost capacitor. For reference, Equation 12 should be used as
a starting point when choosing C
’s maximum ripple current
5
rating (at 120Hz).
I
O( C5 )
=
I
C5_rms
I2I
()
eak C5 _ rms
2
=⋅
12a
3. Select the value of (R
)
multiplier output current without saturating the output.
The maximum output current of the multiplier is
k
M
=
k
2
V
rms
=
kkV
Min(rms_min)
=⋅
0.35 85
()()
=≈
2528.75 2529
1A+R1B
2
2
) that permits the greatest
16
16a
228.57μA.
Selecting the Power Setting Components
The maximum average power delivered by the PFC stage is
set using the following procedure:
1. Find the resistive divider ratio that results in the voltage
at the V
pin being equal to 1.14V at the lowest line
RMS
k 2V V 0.625
+≥
RR
()
1A 1B
+≥
RR
()
1A 1B
+≥
R R 989.38k use 1M
()
1A 1B
in( rms _ min) EAO(max)
0.35 1.414 85 6 0.625
()( )()( )
()
228.57 10
⋅⋅−
228.57 10
ΩΩ
−
−
6
×
−
6
×
17
voltage. The voltage at this pin must be well filtered, yet
able to respond well to transient line voltage changes.
π
R1.14
4
=
R
22V
TOT
⋅
in( rms _ min)
13
The resistor and capacitor values in the typical example
were found empirically to offer the lowest ripple voltage
and still respond well to line voltage changes. Should a
ratio be required that is greatly different from that found
4. Select the value of the current sense resistor to complete
the calculations for the power setting components.
⋅−⋅
R k V 0.625
MULO M EAO( max)
R||R||R||R
5A 5B 5C 5D
R||R||R||R
5A 5B 5C 5D
R || R || R || R 0.452 use 0.3
5A 5B 5C 5D
≤
3.5 10 2529 6 0.625 0.95
()
≤
≤
(
PRR
O(max) 1A 1B
3
×⋅ − ⋅
ΩΩ
+
()
()( )()
100 1 10
()
()
×
η
18
6
in Equation 13, adjust the filter capacitor values according to Equations 14 and 15.
C
=
3
2f R R R R
⋅+⋅+
π
()()
12A2B 34
⎛⎞
1
+
⎜⎟
⎜⎟
RR RR
()()
2A 2B 3 4
⎝⎠
C
=
2
R
TOT
RR
⋅
4TOT
+⋅+
2f R
⋅
π
24
14
15
where:
R
= multiplier output termination resistance (3.5kΩ).
MULO
Voltage Loop Compensation
Maximum transient response of the PFC section, without
instability, is obtained when the open-loop crossover frequency is one-half the line frequency. For this application,
where:
f
= 15Hz, f2 = 23Hz
1
R
= R2A + R2B + R3 + R
TOT
4
the compensation components (pole/zero pair) are chosen so
that the closed loop response decreases at 20dB/decade,
crossing unity gain at 30Hz, then immediately decreasing at
40dB/decade. The error amplifier pole is placed at 30Hz and
an effective zero at one-tenth this frequency, or 3Hz. Find
2. Find the constant of proportionality kM of the multiplier
gain k in Equation 16a. To obtain "brownout" action
the crossover frequency (G
reference, Equation 20 finds the power stage pole and Equation 21 finds the power stage DC gain.
= 1) of the power stage. For
PS
below the lowest input voltage, the maximum gain of the
multiplier must be used when finding k
gain (0.35) occurs when the V
RMS
. The maximum
M
input of the multiplier
is 1.14V. Equation 16 is the general expression for the
multiplier gain versus the line voltage.
)
)
)
© 2006 Fairchild Semiconductor Corporation www.fairchildsemi.com
FAN4800 Rev. 1.0.2 • 12/07/06 3