Different Design steps

for Converters

Dr Pradyumn Chaturvedi, VNIT Nagpur

1. Specifications of Converter
2. Selection of Power semiconductor switches
3. Selection of Magnetic material
a. Selection of core
b. Design of inductor
c. Design of transformer
d. Selection of Litz wire
e. Selection of Capacitor
4. Design of Snubber circuit
5. Design of Heat sink
6. Design of Gate Drive circuit
7. Design of filter
8. Closed loop control
Device selection
 Power Electronics

is the technology
associated with
efficient conversion, control and conditioning
of electrical power
by
Static means
 The goal of Power Electronics

is to control the flow of energy from

electrical source to electrical load

• Thermal with the help of • Industry

• Hydro • Domestic
• Nuclear • Public Utility
• Wind • Transport
• Solar
• Tidal
CONVERTERS • Others

• Others
 Converters

Rectifier
AC DC

Chopper
AC DC
Inverter
 Converters - applications

Domestic : Fan,Television, Washing Machine, Micro-oven,

Computer, Air conditioner, Emergency Lamp, UPS, Door bell,
Eliminator, Clock, Lift, toys, Lighting, Air Cooler, sewing
machine, refrigerator

Transport : locomotive, Rope way, conveyor belt, hoist,

crane, electric car, hybrid car, aeroplane, two wheeler

Communication : Telephone, mobile charger, PA system,

satellite, antenna
 contd….

Drives : AC / DC motors, SRM, PMSM, BLDC, Stepper Motor

FACTS Devices : SVC, Statcom, SSSC, UPFC, UPQC,IPFC,

TCSC, PAR, Active filter, HVDC, HVDC Light, asynchronous
link, solar interface

and many more

Specifications of Converters

➢Power rating
➢Voltage rating
➢Current rating
➢Switching frequency
➢Applications
 Selection of Power
Semiconductor Switches
 MOSFETs: IRFP450 (1), IRFB13N50A (2),
IRFP448 (3), IRFB11N50A (4)
IGBTs: SGP23N60UFD (5), FB15R06KL4 (6),
FB10R06KL4 (7), IRG4BC15UD (8)
Device losses Vs frequency for different switching devices (Load current 5A)
 C
G

Photograph circuit symbol

(a) Output characteristics (b) Transfer Characteristics

• MOSFET – preferred • IGBT – preferred
upto 2.0 kW of power above 2.0 kW power
SPW35N60C3 IRGP50B60PD

VCES = 600V, IC = 33A

Vds=650V, Rds(on)=0.1 Ω, ID = 35A, TO-247 VCE(on) typ. = 2.5V
MOSFET – SPW35N60C3 IGBT – IRGP50B60PD

Typical Parameters Typical Parameters

• td(on) = 10 nsec. • td(on) = 33 nsec.
• tr = 5 nsec • tr = 26 nsec
• td(off) = 70 nsec • td(off) = 140 nsec
• tf = 10 nsec • tf = 50 nsec
Conduction loss (at 50% of Irated = 10 A) Conduction loss (at Irated= Ic= 10 A)

case - 1 case - 1
10amp 10amp

I2Rds = 102 x 0.1 = 10 w VceIc = 2.5 x 10 = 25 w

case - 2 case - 2

5amp 5amp (1) I2Rds = 52 x 0.1 = 2.5 w 5amp 5amp (1) VceIc = 2.5 x 5 = 12.5 w
(2) I2Rds = 52 x 0.1 = 2.5 w (2) VceIc = 2.5 x 5 = 12.5 w
Total loss = 1 + 2 Total loss = 1 + 2

M1 M2
= 2.5 + 2.5 I1 I2
= 12.5 + 12.5
=5w = 25 w
Application Considerations for Silicon
Carbide MOSFETs
Cree Inc
 Application Considerations for Silicon Carbide (SiC)
MOSFETs
The SiC MOSFET comparisons will be made with other similar silicon
devices to illustrate the differences in operating characteristics, not to pick
the best device.

The comparison silicon devices are as follows:

 1.2kV, 0.08 Ω SiC MOSFET Cree CMF20120D

 900V, 0.12 Ω Si super junction MOSFET (SJMOSFET) Infineon

IPW90R120C3
 1.2 kV, 20 A trench/field stop (TFS) Si IGBT Fairchild FGA20N120FGD

 1.2 kV, 20 A non-punch though (NPT) Si IGBT International Rectifier

IRGP20B120U
 1.2 kV, 0.30 Ω Si MOSFET (Si MOS8) Microsemi APT34M120J
The devices selected for comparison are representative of commercially
available Si IGBTs and MOSFETs with voltage and current ratings similar to
the CMF20120D.

 The TFS IGBT is representative of a low on-voltage device

 The NPT IGBT is representative of a low turn-off loss device

 The Si MOS8 is representative of a commercially available 1.2kV Si

MOSFET
 Output characteristics of a typical Cree CMF20120D and the Si NPT IGBT

Fig 1: Output characteristics comparison (TJ = 150 °C)

 The CMF20120D needs to be driven with a higher gate voltage swing than what is
customary with SJMOSFETS or IGBTs. Presently, a +20V and -2V to -5V negative bias
gate drive is recommended for the CMF20120D.
 The rate of rise of gate voltage will have a greater effect on the rate of rise of the drain
current due to the lower transconductance. Therefore, the gate drive needs to supply a
fast rise and fall time gate pulse to maximize switching speed.
 For the CMF20120D , the output impedance is lower and the device does not go into a
clean constant current region during this type of over-current fault, especially under
moderate over-currents. Therefore, the drain to source voltage will not increase as much.
 Fig 2: Forward conduction characteristics comparison (VGS = 20V, VGE = 15V)

➢ At 25 °C, the Si SJMOSFET and CMF20120D were somewhat similar. At

150 °C, the RDS(on) of the CMF20120D increases by only about 20%
from 25 °C to 150 °C, whereas both the Si SJMOSFET and Si MOS8
devices increase by 250%.
 Fig 3: High temperature leakage current comparison
➢ One of the key advantages to SiC is the high temperature capability afforded by the
wide bandgap. This is clearly reflected in the leakage current comparison at elevated
temperature shown in Figure 3. The CMF20120D has about 20x lower leakage
current at 150 °C. At 200 °C, the Si comparison parts leakage current increases
dramatically, to the point where the device fails due to excess power dissipation. The
CMF20120D leakage current is still acceptable at this temperature and is over 100x
lower than the Si devices.
Fig 4: Gate charge and energy comparison
Fig 5: Qg*RDS(on) Figure of Merit Comparison
Fig 6: Switching loss vs. temperature comparison (VDD = VCC = 800V, ID = IC = 20A,
RG = 10Ω)
Silicon Carbide Half-Bridge Module
(CAS120M12BM2)

Part Number Package

CAS120M12BM2 Half-Bridge Module

VDS 1.2 kV
Esw, Total @ 120A, 150 ˚C 2.1 mJ
RDS(on) 13 mΩ
Features
• Ultra Low Loss

• High-Frequency Operation

• Zero Reverse Recovery Current from Diode

• Zero Turn-off Tail Current from MOSFET

• Normally-off, Fail-safe Device Operation

• Ease of Paralleling

• Copper Baseplate and Aluminum Nitride Insulator

System Benefits Applications

• Enables Compact and Lightweight • Induction Heating
Systems • Solar and Wind Inverters
• High Efficiency Operation • DC/DC Converters
• Mitigates Over-voltage Protection • Line Regen Drives
• Reduced Thermal Requirements • Battery Chargers
• Reduced System Cost
 Electrical Characteristics (TC = 25˚C)

Symbol00 Parameter Min. Typ. Max. Unit Test Conditions

V(BR)DSS Drain – Source Breakdown Voltage 1.2 kV VGS,=0 V, ID=300 μA

VGS(th) Gate Threshold Voltage 1.8 2.6 V VDS= 10 V, ID= 6 mA

13 16 VGS =20 V, IDS= 120 A

RDS(on) On State Resistance mΩ

VGS = 20V, IDS = 120 A, TJ =
23 30
150˚C

td(on) Turn-on delay time 38 ns

VDD=600V, VGS=-5/+20V,
tr Rise Time 34 ns ID=120A,RG(ext)=2.5Ω,
Timing relative to VDS
td(off) Turn-off delay time 70 ns Note: IEC 60747-8-4, pg 83
Inductive load
tf Fall Time 22 ns
500
Conditions:
450 TJ = 25 °C
V DD = 600 V
400 I DS = 120 A
V GS = -5/+20 V
350

300
td(off)

250
td(on)
200
tr
150

100
tf
50

0
0 5 10 15 20 25 30 35 40

External Gate Resistor,

G(ext) ( R Ohms
)
SiC MOSFET based 50kW DC/DC
Boost Converter in PV Application
Typical 3-ph String Inverter Topology with Si Devices

MPPT BOOST+3Level Inverter:

 1200V IGBTs for Booster+600V IGBTs for three-level inverter
 Low frequency with10kHZ-20kHZ because of Si IGBT limitation
 Heavy weight and large size, thus low power density
 High passive magnetic cost because of low frequency
 ZVT Three-level Boost Converter Topology

ZVT Three-level Boost Benefits: ZVT Three-level Boost Drawbacks:

• 600V Si MOSFET devices • Additional switcher for aux circuit

• Higher frequency operation • Complicated control method

• Output DC link voltage unbalance

• Zero voltage soft switching
issue
• Higher efficiency
 Key Advantages of Using SiC MOSFETs

SiC power semiconductors are superior to

silicon in 3 critical properties:
• Wider bandgap: SiC supports 10 times

higher electric fields than Si

• Higher thermal conductivity: SiC

supports 3 times the power density of Si

• Reliability: 10X better than silicon

Off State Off State

SiC MOS Key Benefits Vs Si:

• Low conduction losses

• Low switching losses

On State • Enabling high frequency

Turn-on Turn-off • No current tailing for IGBT

 Key Advantages of Using SiC Schottky diodes

Almost zero reverse

recovery energy
independent of device
temperature.
 50kW, 4 phase Interleaved Boost Converter Features
 2x devices hard paralleled
per phase
 4 phase interleaved Boost
with full SiC devices
 Input voltage: 400V-
600Vdc
 Output voltage: 800Vdc

 Output power: 50KW

(12.5KW per channel)
 Controller preset

 2x independent MPPT
channels
 Electrical Specifications

Parameter Unit Value

DC output voltage VDC 800

Max. output power kW 50

DC input voltage VDC 400 – 600

Efficiency % 97.8 – 99.14

Switching Frequency / phase kHz 75

Operating temp* ºC -25 to +35

Storage temperature range ºC -35 to +85

Isolation voltage kV tbd

PCB Assembly Of The 50kW Evaluation Unit
 Measured Versus Calculated Efficiency Over
Varying Load

Note: Gate to source turn on resistor is 15Ohm and turn off resistor is 5Ohm
Ambient temperature is 25°C with fan cooling
 Waveforms With 600VDC Input And 800VDC Output
(D=25%)

Il(10A/div)

Vds1 (500V/div)

Hard paralleled
MOSFETs
Vds2 (500V/div) on each phase

5 uS/div

Turn- Off detail Turn- On detail

200 nS/div
200 nS/div
 Waveforms With 400VDC Input And 800VDC Output
(D=50%)

Il(10A/div)

Vds1 (200V/div)

Vgs (10 V/div)

5 uS/div

Turn- Off detail Turn- On detail

200 nS/div 200 nS/div

 Cree PV Inverter Tops
1kW/kg with All-SiC Design
 SiC Schottky Diodes:
• 600V, 650V, 1200V, 1700V

• 1A to 50A

• Packaged and bare die

Gate Drivers:
C2M SiC MOSFETs:
• Single and dual
• 1200V, 1700V
• Cree designed
Package Styles: • 1Ω to 25mΩ (4A to 60A)
• Partnered designs
• TO-220 • Packaged and bare die
• TO-247

• Full-PAK

• DPAK (TO-252)
SiC Power Modules:
• D2PAK (TO-263) • 1200V, 1700V
• QFN • 100A to 300A Half-bridges
• Isolated TO-220 • 50A Six-pack
Bare die for power
modules
 Commercial three-phase roof top
system

 Need for:
➢ Small size

➢ Light weight

➢ High power density

➢ Lower installed system cost

 SiC devices enable the best

solution to achieve this need

Power density
needed
> 1 kW/Kg
 What Is Used Today - Example
Kaco Blueplanet 50.0 TL3

Full Power MPPT Voltage Range 480 – 850 VDC 50 kW, TL3 series
Operating MPPT Voltage Range 200 – 850 VDC

No. Independent MPPT Input 1

Nominal output power 50 kW

CEC Efficiency 97.5%

Peak Efficiency 98.3%

Power Factor > 0.99

Output Voltage 480 Vac

Operating Temperature Range -30 °C to 60 °C (de-rated > 45 °C)

Power density = 0.29 kW/Kg

Cooling Forced convection

Need: Increase this value > 3×

Weight 173 kg

Isolation Transformer No Cree SiC Technology can

achieve this…
Volume 840 × 355 × 1360 mm
Cree SiC Technology Enables ½ Volume and <1/3 Weight
Cree SiC based 3-ph, PV String Inverter comparison

Full Power MPPT Voltage 480 – 850 VDC 450 – 800 VDC
Range

Operating Voltage Range 200 – 850 VDC 400 – 950 VDC

No. Indep MPPT Input 1 2

Nominal output power 50 kW 50 kW

CEC Efficiency 97.5% 97.8%

Peak Efficiency 98.3% 98.7%

Power Factor > 0.99 > 0.99

Output Voltage 480 Vac 480 Vac

Operating Temperature Range -30 °C to 60 °C -30 °C to 60 °C (no

(de-rated > 45 °C) de-rating)

Cooling Forced air Forced air SiC based 3-ph PV string

Weight 173 kg 50 kg Inverter
Isolation Transformer No No

0.41 0.21 Power Density = 1

Volume (m3) 840 x 355 x 1360 mm 700 x 300 x 1000 mm
kW/Kg
 Cree All-SiC PV Inverter Reference Design Overview

Aux. Power Supply

Aux. Power Supply

450 to 800 VDC Driver Driver

2 x MPPT channels

825 VDC

Channel A

Channel B

Scalable to any numbers

of MPPT channels
~
MPPT Boost, 3- ph Inverter,
75 kHz 48 kHz

Control Plane
3LT-Type Inverter Chosen for High Frequency Efficiency
2L
2 L Si 3L- NPC Si 3L- T Si 2 L SiC 3L- T SiC
100.0%

99.5%

99.0%
3L- NPC
98.5%

98.0%

97.5%

97.0%
3L-T
96.5%

96.0%
0 5 10 15 20 25 30 35 40 45 50

Swithcing Frequency (kHz)

Inverter Schematic: All SiC Active Power Semiconductors

Input 4- ph interleaved DC 3- ph T- Type Output

Filter Boost Converter Link Inverter Filter

2 x C2M0080120D, SiC MOSFET in 10 x 700uF 2 x C2M0025120D, SiC MOSFET in parallel3 x 4uF, 470Vac film
600 VDC film per phase leg upper and lower switch. 3 x 230uH, metglascore
parallel per phase leg.

2 x C4D10120D, SiC MPS diodes in 1 x C2M0025120D,

SiC MOSFET per-Tleg.
parallel per phase leg.
1 x C5D50065D,
SiC MPS Diode per- Tleg.
 50kW 4-Channel Boost Front End
Boost Chokes*
2 pcs C2M0080120D
EMI Filter Choke per each phase

MPPT
Ch A Controller

MPPT
Ch B

Phase Gate
driver
3-Level, T-Type Inverter Topology

3500 µF
270 µH

 Main lower & upper MOSFET position: 2x Cree 1.2kV SiC Mosfet, 50A, 25 mOhms

 Uses Cree SiC MOSFET body diode in main anti-parallel with main power switches

 T-Branch MOSFET position: 1x Cree 1.2kV SiC Mosfet, 50A, 25 mOhms

 T-branch Diode position: 2x Cree 650V, 50A, C3D50065D

SiC System Efficiency v/s Si based system
Better Overall efficiency
@ 1/3rd the weight

CEC Efficiency:
▪ Kaco – 97.5%
▪ Cree – 97.8%
 Operating Temperature Data
Vin = 600 / 700VDC
Tamb= 22ºC
Po = 48 kW LC output filter
Tcore = 66 (with potting
)
103 ºC (no potting)

Tcore = 38/35ºC
Tc = 54/53ºC Tc = 70//68ºC

Tc = 39/38ºC

Tc = 61/54ºC Tc = 46/42ºC
 60W SiC Based Auxiliary Power Supply
Cree 60W Single-end Flyback reference similar to that used in 50kW:
 Input voltage: Wide input range 200Vdc to 1000Vdc

 Output voltage: 12Vdc/4.5A; 5V/0.5A; -12Vdc/0.25A – Operating frequency: 75KHZ

PQ2625
transformer
Cree 1700V
SiC MOSFET:
C2M1000170D

Cree 60W Aux Power Demo board withSiC

1700V
MOSFET

Cree 1700V, 1Ω MOSFET:

 Higher blocking than available silicon

 Higher efficiency for cooler operation

 Enables higher reliability than comparable silicon designs

 All SiC 50kW PV String Inverter Test Assembly
Boost converter
DC Link Film Capacitors
2 x 3.5mF, 1.1kV

Controller

Output Power = 47.8kW

Output Voltage = 492
Vacl- l rms
Output Current = 56A
rms
DC link Voltage = 850 VDC
3- phase, balanced resistive load
Output filter
Inverter Power stage
 Summary
 Using SiC MOSFETs available today from Cree enables up to 3x the power
to weight ratio when compared to current Si based systems.

 Higher Fsw reduces cost of the magnetics in the boost and output filter while
delivering improved efficiency compared to Si solutions.

 Installing smaller, lighter inverters will lower installation costs.

 Less parts and the use of film capacitors in the DC link can improve
reliability of the system.

 Cree is developing a complete range of SiC discrete and module devices to

support the power conversion market.

 SiC makes economic sense today!

Magnetic Materials for Power
Electronics

60
 Introduction
Desired characteristic of magnetic materials are:

✓High relative permeability μr

✓High saturation magnetic flux density Bsat
✓Low coercivity Hc
✓High electrical resistivity ρc
✓High Curie temperature (Tc), low hysteresis and eddy-current
losses per unit volume

61
62
➢ Ferromagnetic materials are thus selected for inductors and
transformers in power electronic circuits
63
64
Ferrite Cores

65
 Ferrite

➢ Widely used material for inductors and transformers in power electronics.

➢ They are made up of iron oxide (Fe2O3), combine with other metals such
as cobalt (Co), copper (Cu), magnesium (Mg), manganese (Mn), nickel
(Ni), silicon (Si) and zinc (Zn).

➢ Typical saturation flux density is in the range 0.25–0.45 T for the different
material grades.

➢ High resistivity so lower eddy current loss.

➢ Low curie temperature. This must be taken into account while designing.

66
➢ Most commonly used types are Mn-Zn and Ni-Zn ferrites.

67
Different shapes of ferrite core

Pot core
PQ Core EE Core

ETD Core U/C Core Toroid Core

68
➢ Saturation flux density of ferrite core decreases with temperature

69
Powder Core

70
 Powder Core
➢ Iron or iron alloy powder is compressed with insulation materials

➢ Insulating material provide distributed air gap, so low value of

effective relative permeability.

➢ The distributed gap helps to tolerate high DC current before the

iron saturates

➢ Have high resistivity, low hysteresis and eddy current losses, and
excellent inductance stability under both DC and AC conditions

71
➢ Effective inductance value is stable over a wide temperature
range

➢ Have high saturation flux density

➢ The lack of localized air gap eliminates a fringing effect

➢ Curie temperature is about Tc= 450◦C

➢ Effective permeability is in the range of 15-550

➢ Basically available in toroidal shape. Sendust and MPP are

available in both toroidal and E cores
72
73
 Comparison of Bsat , Max.operating
frequency and Curi temperature of ferrite
and powder core

74
DC bias performance comparison

If Inductor current have large DC component, then

performance of Ferrite core is poor

75
Core Loss Comparison

76
Relative Cost comparison of Powder
core and Ferrite

77
Temperature Stability Comparison

78
 Performance comparison of Ferrite
and Powder cores

Material Core DC Relative Saturation Curie Operating Tempe-

Loss Bias Cost Flux Tempe- frequency rature
Density rature stability
(Tesla)

AmoFlux Low Better Medium 1.5 400° C 2 MHz Better

High Flux Moderate Best Medium 1.5 500° C 1 MHz Better
Sendust Low Good Low 1.0 500° C 2 MHz Good
Very Low Better High 0.75 460° C 2 MHz Best
MPP
XFlux High Best Low 1.6 700° C 500 kHz Good
Iron Highest Good Lowest 1.2 - 1.5 770° C 200 kHz Poor
Powder
Ferrite Lowest Poor Lowest 0.45 100 - 250° 2MHz Poor
C

79
Carbonyl Iron powder core

80
 Carbonyl Iron powder core
➢ It is another type of powdered iron core. Carbonyl iron cores are
obtained by heating carbonyl iron Fe(CO)5

➢ Has a relatively low permeability, in the range (1–50).

➢ Have low hysteresis losses up to very high frequencies (10MHz)

and can be used in high frequency applications.

➢ The quality factor Q, (Q = jwL/R) is very high even for

frequencies above 150 MHz.

➢ Main applications are in EMI/RFI devices and in high linearity RF

current sensing
81
Amorphous Alloys

82
 Amorphous Alloys
➢ Made up of ferromagnetic and metallic elements like Si, B ,
Carbon etc.

➢ They have glass like structure so also known as metallic glass.

➢ The relative permeability μr is very high, more than 150 000.

➢ saturation flux density is in the range from 0.7 to 1.8 T

➢ As temperature goes from 25oC to 250oC, the saturation flux

density reduced by 30% or more.

83
➢ POWERLITE®, Metglas®, VITROVAC® etc are some of
the commercially available amorphous high frequency core.

➢ Maximum operating frequency is about 250KHz.

➢ Usually available in U and Toridal shape.

➢ Used in pulse transformer, current transformer, SMPS

power factor correction choke etc.

84
Nano crystalline Cores

85
 Nano crystalline Cores
➢ Made up crystals with a typical size of 7–20 nm that are iron (Fe)
based . In addition there are traces of Si, B, Cu, molybdenum
(Mo) and niobium (Nb)

➢ They combine the high saturation magnetic flux density of silicon

steels with the low loss of ferrites at high frequencies.

➢ saturation flux density is about 1.2T- 1.5 T.

➢ The nanocrystalline cores are used up to 150 kHz.

➢ High relative permeability μr.
➢ applications in current transformers, pulse transformers and
common-mode EMI filters
86
 Refernces
1. Data sheet of Magnetics®, http://www.mag-inc.com/

2. Data sheets from Metglas®, http://www.metglas.com/

3. Valchev, Vencislav Cekov, and Alex Van den Bossche. Inductors and
transformers for power electronics. CRC press, 2011.

4. Kazimierczuk, Marian K. High-frequency magnetic components. John

Wiley & Sons, 2013.

5. McLyman, Colonel Wm T. Transformer and inductor design handbook.

Vol. 121. CRC press, 2004.

6. Hurley, William G., and Werner H. Wölfle. Transformers and Inductors

for Power Electronics: Theory, Design and Applications. John Wiley &
Sons, 2013. 87
Selection of Magnetic Materials
 Conduction Process

The conditioning of power flow in PES is

done through the use of electromagnetic
I A m2

and reactive elements –

ε V/m
• inductors,
I
• transformers
J=σε
• capacitors and
In this section the basics of
electromagnetics is reviewed.
 Electromagnetics

The voltage across and current through a conducting element is related through
Ohm's law. This law may be stated as follows –
When an electric field (of intensity ε V/m) is set up across a conducting material
(of conductivity σ1 / Ωm), there is an average flow of electrical charge across
the conducting material (of current density J A/m2).

J=σε

When expressed in terms of element voltage and current, this reduces to the
familiar statement of Ohm's law –

I = V/R

R = l/ σA
 Magnetization Process

Φ A m2
B = μH
H A/m
Φ=F/R
I
R = l / A μ0 μr (μ0=4π10-7 H/m)
B=μH

Magnetic Equivalent Circuit

Design of Inductor
 Design of Inductor

The inductor consists of a magnetic circuit and an electrical circuit. The

design requires –
• The size of wire to be used for the electric circuit, to carry the rated
current safely.
• The size and shape of magnetic core to be used such that –
• The peak flux is carried safely by the core without saturation.
• The required size of the conductors are safely accommodated
in the core.
• The number of turns of the electric circuit to obtain the desired
inductance.
 Material constraints

•Any given wire (conducting material) can only carry a certain maximum

current per unit cross section of the wire size. When this limit is exceeded,

the wire will overheat from the heat generated (I2R) and melt or

deteriorate. The safe current density for the conducting material is

denoted by J A/m2.

• Any magnetic material can only carry a certain maximum flux density.

When this limit is exceeded, the material saturates and the relative

permeability drops substantially. This maximum allowable flux density for

the magnetic material is denoted by Bm T.

 Core material characteristics:
the relation between B and H

B B

µ
0

H H
µ0

Free Space Magnetic Core Material

B = µH and µ = µ0µr
1. B-H characteristics : as narrow as possible
e.g. Mg-zinc : narrow B-H curve,
have low volume resistivity

2. Volume resistivity must be reasonably high

e.g. Ni-zinc : higher volume resistivity,
wider B-H characteristics
Bm → 10% - 15%
- increases number of turns
- increases core size
- less utilization of core material

3F3 – from ferrocube

H7F – from TDK
K or R – from magnetics
EPCOS (Siemens)
 Design Relationships
In order to design an inductor of L Henry, capable of carrying an rms
current of Irms and peak current of Ip –

Let the wire size be aw m2. Irms

aw = (1)
J

Let the peak flux density in the core

of area (AC) be Bm on account of N*Ip = L*Φp = N*Ac*Bm (2)
the peak current Ip in the inductor.

The winding of the inductor is accommodated in the window of the core.

Let the window area (AW) be filled by conductors to a fraction of kw.

Irms
kwAw= Naw = N (3)
J
 Cross-multiplying Eq. (2) and Eq. (3), we get

Irms
LIpN = NAcBmkwaw (4)
J

LIpIrms = kwJBmAcAw (5)

• kw depends on how well the winding can be accommodated in

the window of the core. kw is usually 0.3 to 0.5.

• Bm is the maximum unsaturated flux is about 1 T for iron and

0.2 T for ferrites.

• J is the maximum allowable current density for the conductor.

For copper conductors J is between 2.0x106 A/m2 to 5.0x106 A/m2.

1. Compute
LIpIrms
AcAw=
kwBmJ

2. Select a core from core tables with required Ac*Aw.

3. For the selected core, find Ac and Aw.
4. Compute & select nearest whole number of N*.

LIp
N=
AcBm

5. Compute & select nearest whole number of wire guage

and aw* from wire table.

Irms
aw =
J
6. Compute the required air gap in the core
μ0N*Ip
lg=
Bm

7. Check the assumptions

• core reluctance<<air gap reluctance
l
Rc<<Rg ; << lg
μr
• no fringing
lg << Ac

8. Recalculate Irms
J* =
a*w

9. Recalculate N*aw
kw* =
Aw

10. Compute from geometry of the core, mean length per turn and the
length of winding. From wire tables, find the resistance of the winding.
 Different Types of Ferrite Cores

RM Cores

• RM cores are ideal for a extremely low-loss, highly stable filter coils and
other resonance-determining inductive components
• a low-distortion broadband transformers for small signals
• a RM cores without centre hole are also used in power electronics.
 Low ProfileRM Cores

RM cores with substantially reduced total height

These cores can be used for small-signal, interface and matching transformers
as well as for transformer and storage chokes in DC/DC converters with a high
switching frequency.
The low-profile models are es
pecially suitable for applications in which the winding is printed on the board
and the cores are inserted from both sides.
P – Cores

Magnetically closed pot cores, very low leakage

• These cores are suitable for a resonant-circuit coils
(filters) with high inductance stability and Q,
• low-distortion broadband small-signal transformers in
materials T38 and N30.
• P cores without centre hole of N87 material are suitable
for power applications. The larger, effective magnetic cross-
section means that they feature higher AL, better flux
density distribution and thus lower power loss.

P3.3 ….. P22

 EP – Cores

For compact transformer

• With their cubic geometry, EP cores are recommended for space-
saving onboard solutions.
• In conjunction with our materials, these models feature excellent
characteristics for broadband, small-signal transformation with low
leakage field.
• EP core are consequently ideal for xDSL applications.
 EPX / EPO – Cores

Optimized ferrite cores for xDSL applications

• Internet access with DSL is the fastest growing broad band technology.
• For the new requirements EPCOS has created ferrite cores and
materials for DSL line transformers. These new EPX and EPO cores
increase loop reach at a given data rate for DSL applications.
PM – Cores
PM cores for handling very high powers
• The benefits of this pot core have been
proven in numerous applications in tele-
communications and industrial electronics:
• wide flux area for high power at minimum
number of turns,
• low magnetic leakage and stray
capacitance, good shielding owing to the
closed form, precisely ground air gaps.
U / I – Cores

For power, pulse and

high-voltage transformers

E – Cores
 EQ / I Planner Cores
ELP / I – Cores
Toroid – Cores
Design of Transformers
 Design of Transformers

The transformer consists of more than one winding. Also, in order to keep the
magnetization current low, the transformer does not have air gap in its
magnetizing circuit. Consider a transformer with a single primary and single
secondary as shown in above fig. Let the specifications be
Primary : V1 v; I1 amp;
Secondary : V2 v; I2 amp;
VA Rating : V1*I1 = V2* I2;
Frequency : f Hz
For square wave of operation, the voltage of the transformer is
V1 = 4 f Bm AC N1 ; V2 = 4 f Bm AC N2 (1)
The window for the transformer accommodates both the primary
and the secondary. With the same notation as for inductors,
kw AW = N1*I1 + N2*I2 (2)

From the above equations,

V1*I1 + V2*I2 = 4 kw*J*Bm*AC*AW (3)
V A = 2 f Bm*J*AC*AW (4)

VA
AC*AW = (5)
2 f Bm*J*kW

The above equation relates the area product (ACAW) required for a
transformer to handle a given VA rating.
 Design Steps
The design requires –
• Size of wire and number of turns to be used for primary and secondary windings.
• Core to be used.
• Resistance of the winding.
• Magnetizing inductance of the transformer.

1. Compute the Area product (ACAW) of the desired core.

VA
ACAW =
2fkwJBm
2. Select the smallest core from the core tables having an area product higher
than obtained in step (1).
3. Find the core area (AC) and window area (AW) of the selected core.
4. Compute the number of turns
V1 V2
N1 = N2 =
4fBmAC 4fBmAC

5. Select the nearest higher whole number to that obtained in step (4), for
the primary and secondary turns.
 6. Compute the wire size for secondary and primary.

aw1 = I1/J aw2 = I2/J

7. Select from the wire tables the desired wire size.
8. Compute the length of secondary and primary turns, from the mean
length per turn of the core tables.
9. Find from the wire tables, the primary and secondary resistance.
10. Compute from the core details, the reluctance of the core.

lc
R=
A C μ0μ r

11. Compute the magnetizing inductance.

N2
Lm=
R
Selection of wires
 Selection of wires

Expression for skin effect

δ = 0.066 / √ f meters
1. Use of Litz wire
• provide more surface area
• promotes current sharing between the wires
• ∑ individual cross-sectional area = area of solid conductor
2. Use of flat foil conductor
• low voltage, high current winding within transformer
• Thickness of foil is no more than two skin depth
• Reduce winding loss due to skin effect
3. Litz wire offers the best resistance to the induction of eddy current –
• dielectric isolation between windings
• The degree of magnetic coupling needed by each winding to the
core and to the other winding to minimize the leakage inductance
• which winding are better implemented using Litz wire of foils wires.
 High-current secondary
½ primary
secondaries
½ primary

Insulation tape

Foil winding
Solid wire
Litz wire

Method of minimizing winding eddy currents and

maintaining good coupling dielectric isolation
 Design of Litz wire

➢ Litz wires- conductors made up of multiple, individually

insulated strands, twisted or woven together.
➢ Frequency range from 10 kHz to 1 MHz.
➢ To minimize the skin effect in the winding.
➢ The number of strands in a Litz wire is increased; to decrease
the cross sectional area each strand.
➢ The number of strands increases the fraction of window area
filled with copper decreases and fraction filled with insulation
increases.
➢ Increased dc resistance.
 2
➢ The winding loss is represented by, Ploss = I ac . Fr . Rac

 2   2  02  N 2  d c2  k
Fr = 1 +
768   c2  bc2

➢ The diameter of copper in single strand,

1

F  F b  h
( )
1 1 2
(1− −

d =d   
)
  p lp b

n N 
 
c r

➢ The total resistance factor equal to ratio of ac resistance of Litz

wire to dc resistance of equivalent single stranded wire is given by,

Fr' = Fdc + Fr
➢ Fdc is ratio of dc resistance of litz wire to the dc resistance of a
single strand winding using a wire with the same diameter as
that of litz bundle.
(−
1
)  ( 1 − 1) (1 − ) 
2

F = Flp
r
' 
 n + n 

 
 3
6 6
(6 − )
− h 
  N     d
2 2 2 2
    Fp  Flp  bb    k
 

 N
s 0 r
=
768   c2  bc2

➢ The optimal number of strands as below,

1
 2   1   3 − 2 
nopt =  − 1    − 1   
     
 Total resistance factor and its components

No. of strands

Fr is total resistance factor, Fdc and Fr are dc & ac resistance factors

Fig. - Variation in resistance factors with strands of

Litz wire wound inductor.
Table - Design table for L1 for optimal and sub-optimal stranding

Relative
AWG Strands Fr
Cost

49 866 1.9469 16.43

44 350 1.9837 2.040

42 225 2.04 1.4047

40 135 2.18 1.088