WEST COAST MAGNETICS Application Notes

ALNT 1440, Revision 1 FLYBACK CONVERTER DESIGN

TYPICAL APPLICATIONS
The flyback converter has a number of advantages
which make it one of the most widely used topologies
for low power (< 200 W) designs. The flyback converter provides DC isolation of up to 5000 volts and it
is much more failure tolerant than a non-transformer isolated topology. When compared to other transformer
isolated topologies like the forward converter or push-pull
converter the major advantage is simplicity. It is very
easy to have more than one output on a single transformer and the outputs can be both positive or negative
in voltage. Additionally very few components are
needed to build a flyback converter (only one switching
device is required and output inductors are not needed).
Input voltage range is typically 3 to 1 and special designs can even accommodate ratios as high as 6 to 1.
The flyback also has some disadvantages which
will favor other topologies in many applications. The
input and output current ripple is high compared to other
topologies. Cross regulation also is generally high with
values on the order of +/-5% being typical. These problems can be addressed by adding other components to
the supply however with the addition of more components the essential advantage of the flyback design
(simplicity and cost) is lost. Therefore it may be preferable to look at a different topology when output ripple
and cross regulation are very important.
Due to the high ripple currents the flyback is less
efficient than other designs and therefore is generally
not used for high power applications. Also the flyback
can have voltage spikes on the primary which will tend
to stress other components and can lead to reliability
problems if the design and build process is not controlled carefully.
Despite these disadvantages the flyback topology
is very widely used for low power switch mode designs
and a large selection of switching transistors are available. With careful and thorough design, and good
manufacturing control the flyback transformer will perform efficiently and reliably.

OPERATION DESCRIPTIONGENERAL
The basic circuit for a flyback transformer with a
single output is shown in Figure A. The operation of the
flyback can be described by breaking the operational
cycle into two parts: the on-time and the off-time of the
switch. During the on-time, a constant voltage is applied across the primary winding of the transformer

WEST COAST MAGNETICS

which is equal to Vin minus Vt. During this time the

current will ramp in a linear fashion with a slope equal
to (VinSVt)/Lpri where:
Vin = DC input voltage
Vt = Switching transistor voltage drop
Lpri = Transformer primary winding inductance.

Np

Vin

Ns

Vout

Vt

Figure A
TYPICAL FLYBACK
REGULATOR CIRCUIT

When the power switch turns off the input voltage

drops to zero and the energy stored in the core flies
back to the secondary of the transformer and to the
load. The voltage on the secondary of the transformer
during the flyback time is then determined by the turns
ratio, (note that the actual power supply output voltage
will be lower than the transformer secondary voltage
due to the diode drop). The flyback period will continue
until the core is depleted of energy and the output current drops to zero (discontinuous operation) or until the
power switch is turned on again (continuous operation).
Because the transformer never operates with both
windings conducting at the same time it can be modeled as two separate inductors. Looking at the transformer
in this fashion it is apparent that the designer can choose
the primary to secondary turns ratio based on factors
besides the voltage ratios and in fact a 2:1 turns ratio
can create 3.3 volts just as easily as 5 volts from the
same input voltage. On and off times are rarely equal
due to variations in input and output voltages and the
turns ratio.

OPERATIONAL DESCRIPTIONDISCONTINUOUS
Most flyback designs are discontinuous because
the transformer is typically smaller and because the
power supply is more stable. The power switch voltage

1-800-628-1123 www.wcmagnetics.com

continued

Application Notes
ALNT 1440, Revision 1 FLYBACK CONVERTER DESIGN
and input and output currents are shown for the discontinuous mode of operation in Figures B, C, and D.
Transformer
Current

I peak

Transformer
Primary
Voltage

ts = 1/f

Vin - Vt

ts = 1/f

Time

Figure E
Discontinuous Operation
Time

Figure B

Transformer
Current
Transformer
Primary
Current

I peak
Ioc

I peak

ts = 1/f
Time

ts = 1/f

Time

Figure F
Continuous Operation

Figure C

It should be noted that with a discontinuous design it is possible to deliver more power to the load by
reducing the primary inductance of the transformer. This
will also reduce the size of the transformer. The degree
to which this is possible will be governed by reliability
concerns as high peak currents will cause the power
supply semiconductor devices to fail. The disadvantage
of the discontinuous mode is that peak currents are higher
than in the continuous mode of operation.

Output
Diode
Current

I peak

ts = 1/f

Time
Figure D

In the discontinuous mode of operation the output current drops to zero before the power switch turns
on and the current ramp starts up. In this case the secondary current ripple starts from a zero base on each
cycle and the core only stores energy during the switch
on-time and during the flyback period. Every cycle will
have a small dead time in which nothing in the transformer is energized. It is this shut down time which is
unique to the flyback and which allows the flyback to
regulate over a wide range of input voltages and over
a wide range of output currents.
The discontinuous mode of operation can be distinguished from the continuous mode by examining the
output current waveform across the transformer
(Figures E and F).

Even when the design is based on a discontinuous mode of operation it is not unusual for a flyback to
revert to continuous mode of operation when the input
voltage drops too low. In the continuous mode energy
must be stored in the core.

OPERATIONAL DESCRIPTIONCONTINUOUS
If efficiency and reliability are a driving concern,
and the flyback topology must be used it is better to
operate in the continuous mode. Lower peak currents
will not only improve the reliability of the power supply,
they will reduce switching losses and improve the efficiency of the power supply.
In the continuous mode of operation the output
current never drops to zero (Figure F). The current ripple
essentially rides on a pedestal of continuous DC current, thereby creating a strong DC bias in the core. The

WEST COAST MAGNETICS

2

continued

Application Notes
ALNT 1440, Revision 1 FLYBACK CONVERTER DESIGN

It is very important to note that operation in the

continuous mode introduces a control problem for the
designer. Because of a negative zero in the small signal
control to output voltage transfer function located in the
right half plane of the frequency plane the flyback converter is subject to unstable operation in the continuous
mode. While this problem rarely leads to converter failure it does create problems with stability and transient
response.

on-time of the power switch is very short and it is not

possible to source enough output current to drive the
switch into a saturated state before turning it off again.
This creates a big reduction in efficiency and can lead
to a runaway failure mode.
High Input Voltage
Primary Current

converter designer must allow for this DC bias and determine the magnetizing force under high load conditions.
Adequate margin must be allowed to insure against saturation due to the magnetizing force.

Therefore when designing a continuous mode converter compensation should be introduced to minimize the
problem. Typically this problem is addressed by adding of
current mode control to the circuit topology.

DESIGN INPUT VARIABLES

In order to minimize the design cycle time it is
preferable to fully specify the transformer design variables before undertaking a design. A complete set of
design variables would include the following:
Efficiency: The target design efficiency of the entire power supply is utilized in order to generate a
maximum loss value for the transformer. By estimating
losses from each component in the power supply a loss
budget can be generated for the transformer. This loss
value (expressed in Watts) then becomes a part of the
iterative transformer design process. Alternatively the
transformer may be designed around an allowable temperature rise specification.

Low Input
Voltage

I peak

Time
Figure G
PRIMARY CURRENT vs TIME
for Differing Input Voltages

Output Load: Amperage and voltage should be

specified for each output. Typically the amperage is
specified as a DC value. This value must be converted
to RMS AC by the designer for loss calculations. Maximum or design load conditions must be known to
complete a design. It is also beneficial to know minimum load conditions as this will effect cross regulation
and will influence aspects of the transformer design.
Output Diode(s): A flyback converter will require
a diode for each output. The voltage drop across each
diode is required in order to accurately specify the primary to secondary turns ratio from the DC transfer
function. The voltage drop will vary with output current
and typically a design value is chosen at high load to
allow some margin.

Minimum Input Voltage: The minimum voltage is very

important as most flyback designs are based on this variable and therefore it should be specified very carefully. If
an AC voltage is used as a design variable, specifications on the rectification circuit are required.

The output diode drop is also the primary factor

effecting load regulation. Therefore an accurate prediction of load regulation will require specification of
the relationship between current throughput and diode
voltage drop for each diode.

Maximum Input Voltage: The maximum voltage is

important because it will effect the switch on- time. At
maximum line input voltage the switch on-time will be at
a minimum. At high input voltages the on-time of the
PWM can become so short that the special control techniques are required. Typically the flyback can
accommodate a 3:1 input voltage range. For greater
input voltage ranges special control circuits will allow
input voltage ranges in excess of 6:1.

The final specification which is important is an

estimate of diode losses. The diode loss value is used
as part of the power supply loss budget and can effect
the transformer design efficiency.

Most flyback topologies utilize a fixed frequency

current mode type of control. Under these conditions
the controller attempts to maintain a constant peak current over the range of input voltages. See figure G. The
problem that arises is that at high input voltages the

Other Design Variables: The maximum duty cycle

and the voltage drop across the switching transistor are
also required to optimize the flyback design output inductor is used it may be necessary to include the inductor
voltage drop in the input/output transfer function to determine the turns ratio.
Output Inductor: Output inductors are not needed
with a flyback converter but they are often used to minimize output ripple. The starting point for the design of

WEST COAST MAGNETICS

3

continued

Application Notes
ALNT 1440, Revision 1 FLYBACK CONVERTER DESIGN
the output inductance will be a calculation of the desired inductance value. This inductance value will be
derived from a choice of how much attenuation of
ripple is desired. Then the design will take into account the level of DC to predict magnetizing force,
AC input to predict flux density and the loss budget
for the component.
Switching Transistor: The selection of a switch for
the flyback regulator must be made carefully because
it will define the requirements for much of the power
supply circuit.

For the transformer design the voltage drop across

the transistor should be specified to determine the correct turns ratio.
The other critical design variable which comes from
the PWM is the maximum duty cycle, typically 45% to 50%.
The current rating for this component will depend
on a number of variables including mode of operation,
input voltage and load. The maximum voltage is also
very critical and can be determined from the transformer
specifications and the snubbing method.
As discussed before a large voltage spike will
destroy the PWM. This can be avoided by the addition of a snubber across the primary winding of the
transformer. A typical snubber schematic is shown in
Figure H. However this will add cost and complexity,
and will reduce the efficiency of the power supply
significantly. The use of a snubber can be avoided in
most cases by the careful design and controlled manufacture of the transformer.

would be 50 microhenries +/- 10%.

Geometry: A number of core geometeries are
suitable for flyback transformers including pot cores,
E cores, EP cores and RM cores. West Coast Magnetics
401, 402, 403, 404, 405, 406 and 407 series are
ideal for flyback applications. The choice of size will be
a function of frequency, throughput power, allowable
losses (temperature rise), and safety agency requirements.
Core Material: Flyback transformers are typically
constructed from high quality power ferrite materials.
Material permeability should be selected based on frequency of operation. For operation below 500 Khz
most designers will use a core material with a permeability of 2000 to 2500. Permeability varies
significantly with temperature rise and operating flux
density. In general however this will not effect operation of the converter as long as the core is not close to
saturation as the inductance (which controls the mode
of operation) is primarily determined by the air gap.
However temperature rise and operating flux density
will effect core losses and this must be taken into account to ensure reliable operation.
Core Gap: The core gap will be determined by
the number of primary turns and the inductance specification. The designer will verify that the gap is sufficient
to prevent core saturation resulting from DC bias.
Windings: The number of primary and secondary turns, wire type, gauge and winding pattern will
also be specified. The manner in which the windings
are placed on the bobbin is very important.

+
OTHER
Secondary Winding

FIRST
Secondary Winding
Primary Winding

Interwinding
Insulation

Vin

CORE

Vt
-

Bobbin

Figure H

Figure I

SNUBBER SCHEMATIC

FLYBACK TRANSFORMER
Cross Section

FINAL DESIGN PARAMETERS

Inductance: The inductance of the primary winding of the transformer will govern the mode of operation
(discontinuous vs. continuous). Inductance is typically
determined at high output current and low line input
voltage as this is the most severe loading condition. A
typical inductance specification for a flyback transformer

In particular the coupling between the primary

and secondary windings and between the windings and
the core is very important. This will effect the magnitude
of the voltage spikes across the transistor. This coupling
can be measured by testing leakage inductance and
interwinding capacitance. It is controlled by designing
the windings of the transformers carefully. Techniques
used to minimize leakage inductance and capacitance

WEST COAST MAGNETICS

4

continued

Application Notes
ALNT 1440, Revision 1 FLYBACK CONVERTER DESIGN
include the use of multistranded layered windings, copper foil windings, and bifilar windings. It is also effected
by the order in which the windings are placed on the
transformer It can be controlled in a production setting
through the use of automated winding and testing
equipment.
Figure I shows a cross section of a typical flyback
transformer winding used in a low voltage application.
Safety Agency Requirement: Flyback converters may
have special build and test requirements to meet safety
agency requirements. These requirement may include
one of more of the following:
1.) Hipot - expressed in applied VAC or VDC between selected windings and other windings and
between the core and selected windings.

Creepage and clearance - distance between

uninsulated portions of selected windings measured according to UL or IEC guidelines. Similiarly distance
between the core and uninsulated portions of selected
windings.
2.)

C, depending on the application.

It is very important to bear in mind that theoretical temperature rise calculations are not a substitute for
thermocouple measurements under actual load conditions in the design circuit. Thermocouple measurements
are recommended in all cases, especially if safety agency
approval is required.
Core Losses: Core losses can be determined with
a high degree of precision once the core geometry and
material type, operating frequency, number of turns and
primary voltage are known. This information is used to
estimate the operating flux density of the transformer under max load conditions. Flux density and frequency of
operation determine the losses (typically expressed in
mW/cm3).

Flyback transformers operate in the first quadrant

of the hysterisis loop (Figure J). As such, peak operating
flux density is typically limited to 2000 gauss at 50 Khz.
This will vary depending on core geometry with smaller
sizes having an ability to effectively dissipate more power
per unit of volume.

3.) Insulation system - in most cases a UL recognized transformer with a temperature rating of class B
(130 degrees C) and higher will require a UL approved
insulation system.

B
B max
B
B
B

4.) Insulating

materials - many UL and IEC specifications will have specific requirements for the thickness
and number of layers of insulation which will be required between selected windings.

THERMAL CONSIDERATIONS:
Most power supply designs have a goal of compactness, cost effectiveness, and a goal of reliability.
These goals often conflict and the power supply design engineer must balance these considerations during
the design process. The magnetics typically end up
being the largest components on the board and for
this reason the tradeoff between compactness and reliability is acute.

Figure J
HYSTERESIS LOOP of MAGNETIC CORE
in Flyback Circuit

As operating frequency increases, lower flux levels are required to hold losses constant (Figure K).

The reliability of the power supply is directly related to temperature rise. Reliability will also effect the
safety of the power supply, particularly in the case of
the transformer.
Temperature rise for magnetic components comes
from two sources: core losses and copper losses. Typically the magnetics designer will estimate core losses
and copper losses. Then these loss values will be used
along with a thermal resistivity coefficient which is specific to the transformer geometry to predict a temperature
rise for the part. Temperature rise due to operation may
be as low as 10 degrees C and as high as 60 degrees
WEST COAST MAGNETICS
5

continued

Application Notes
ALNT 1440, Revision 1 FLYBACK CONVERTER DESIGN

Flux Density B-gauss

2000
1750
1500
1250

Magnetic
Flux Lines

1000
750
500
250
0

20

30 40

60 80 100

Eddy Currents
Concentrate
Current Flow at
Surface of Wire

200 300 400 600 1000

Frequency kHz
Figure K
OPERATING FREQUENCY vs. FLUX DENSITY
Core Losses Constant - Typical Ferrite
Power Material

Wire Cross Section

Figure M

EDDY CURRENTS
Contributing to AC Resistance

Core losses also vary with operating temperature.

This relationship is expressed in (Figure L), for a typical
ferrite power material. In general, for a given material,
core losses will decrease with increasing temperature
up to a transitional point, and then they will start increasing again. Every ferrite power material will have
a unique set of loss characteristics. It should not be assumed that one material can be substituted for another
merely because it has a similar permeability. It is imperative to estimate core losses using the manufacturers
loss curves for the material used in the design.

AC losses will dictate the use of finer gauge wires

or copper foil windings as the operating frequency increases. This tends to result in less copper area per unit
of bobbin window area.
Higher temperatures will also give rise to higher
copper resistivity. (Figure N) shows a normalized copper resistance factor as a function of temperature.
100
90
80

340

70

Temperature T(C)

380

Core Loss

300

mW
cm3

260
220
180

60
50
40
30
20

140

10
100

0
60
-60

-30

30

60

90

120

Temperature (C)
Figure L
CORE LOSS vs.TEMPERATURE:
Typical Ferrite Power Material

Copper Losses: Magnetic components used in

flyback converters will also have losses associated with
the copper wire or copper strip used to form the winding.

Copper Resistance Factor

Normalized to Unity at 20C
Figure N

Final design calculations must consider the effects

of AC and DC resistance and the temperature rise of the
wire when determining copper losses for the transformer.

West Coast Magnetics typically utilizes Class B

(130 degrees Centigrade) materials in the construction
of its parts. A good rule of thumb for thermal considerThese losses include DC losses and AC losses.
ations is to allow for a 40 degrees Centigrade operating
AC losses are created from eddy currents which circutemperature rise at maximum load. In most cases this
late on the surface of a conductor. See (Figure M) and
will allow adequate margin given ambient conditions to
from proximity effects generated from other windings.
remain safely below the 130 degrees Centigrade temperature rating on the materials used in construction.
WEST COAST MAGNETICS
6