ZKZ 64717
07-09ISSN: 1863-5598
Electronics in Motion and Conversion July 2009
Typ. solder layer Sinter layer
Unbreakable sinter joint:Melting temp. is 6 x higher than operating temp.
TypTypTypT sssoldoldere laya erere SiSiSintter lllayer
220oC 150oC
>900oC
6 x higherSolidus temperature
Operatingtemperature
Sintered chips - for high operation temperaturesSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSiiiiiiiiiiiiiiiiiiiiiinnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnntttttttttttttttttttttttttteeeeeeeeeeeeeeeeeeeeeeeerrrrrrrrrrrrrrrrrrrreeeeeeeeeeeeeeeeeeeddddddddddddddddddddddd ccccccccccccccccccccccccchhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhiiiiiiiiiiiiiiiiipppppppppppppppppppppppppsssssssssssssssssssssss fffffffffffffffffffffffffffffffoooooooooooooooooorrrrrrrrrrrrr hhhhhhhhhhhhhhhhhhhhhhiiiiiiiiiiiiiigggggggggggggggghhhhhhhhhhhhhhhh oooooooooooooooppppppppppppppppeeeeeeeeeeeeeeeeeerrrrrrrrrrrrrrraaaaaaaaaaaaaaaaattttttttttttttttttiiiiiiiiiiiiiiioooooooooooooooooonnnnnnnnnnnnnnnn tttttttttttttttteeeeeeeeeeeeeeeeeeemmmmmmmmmmmmmmmmmmmmmmmmmmmmpppppppppppppppppppeeeeeeeeeeeeeeeerrrrrrrrrrrrraaaaaaaaaaaaaaaaaattttttttttttttttuuuuuuuuuuuuuuuurrrrrrrrrrrrrrrrrrreeeeeeeeeeeeeeeeeeeeeeeesssssssssssssssssSKiiP®
4th generation
Intelligent Power Module: IPM
3 in 1: Driver, semiconductor, cooling
400 kW – 1,8 MW
33% more power, same volume
5 x higher thermal cycling capability
Sintered chips
Australia +61 3-85 61 56 00 Belgium +32 23 00 07 93 Brasil +55 11-41 86 95 00 Cesko +420 37 80 51 400 China +852 34 26 33 66 Danmark +45 58 10 35 56 Deutschland +49 911-65 59-0 España +34 9 36 33 58 90 France +33 1-30 86 80 00 India +91 222 76 28 600 Italia +39 06-9 11 42 41 Japan +81 68 95 13 96 Korea +82 32-3 46 28 30 Mexico +52 55-53 00 11 51 Nederland +31 55-5 29 52 95 Österreich +43 1-58 63 65 80 Polska +48 22-6 15 79 84 Russia +7 38 33 55 58 69 Schweiz +41 44-9 14 13 33 Slovensko +421 3 37 97 03 05 Suid-Afrika +27 12-3 45 60 60 Suomi +358 9-7 74 38 80 Sverige +46 8-59 4768 50 Türkiye +90 21 6-688 32 88 United Kingdom +44 19 92-58 46 77 USA +1 603-8 83 81 02 [email protected] www.semikron.com
www.bodospower.com July 2009
C O N T E N T S
Viewpoint
It is Summer, Beach Time! . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
News . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-7
Green Product of the Month
Energy-Efficiency Calculator Navigates the Maze of External Power Supply Standards
Power Integrations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Blue Product of the Month
Power Pack Offers 33% Increase in Power Density
Semikron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Market
Electronics Industry Digest
By Aubrey Dunford, Europartners . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Guest Editorial
Have No Fear: Digital Power is Here!
By Chris Young, Sr. Manager, Digital Power Technology, Intersil . . . . . . . . . . . . . . . . . . . . . . . 12
Market
Incentives Fuel PV Inverter Opportunities
By Richard Ruiz Jr. , Research Analyst, Darnell Group . . . . . . . . . . . . . . . . . . . . . . . . . . . 14-15
Cover Story
The Bi-mode Insulated Gate Transistor (BIGT)
By Munaf Rahimo, Arnost Kopta and Ulrich Schlapbach, ABB Switzerland Ltd, Semiconductors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-19
Power Management
Integrated Digital Isolation Delivers Enterprise Energy Efficiency Benefits
By Amit Gattani, Akros Silicon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-22
Transformers
Planar Transformers are Essential for Truly Efficient Electric/Hybrid Vehicles
By Dean Curran, Managing Director at Himag Solutions Ltd., United Kingdom . . . . . . . . . 24-25
Power Supply
Smarter Rectification
By Helen Ding, International Rectifier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26-27
Power Supply
Get <1ppm Performance from a 10ppm Precision Reference
By Bob Frostholm, VP Marketing – Microbridge Technologies Corp . . . . . . . . . . . . . . . . . . 28-29
Power Modules
Power Module Testing with Combination Testers
By Günther Dörgeloh, MRS Electronic GmbH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30-32
Sensors
Low Consumption Flux-Gate Transducer
By Manuel Román, Guillermo Velasco, Alfonso Conesa and Felipe JerézPremo and TU of Catalonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .34-37
Power Supply
Choosing The Right Topology
By Andrew Skinner, Advanced Development Manager, TDK-Lambda . . . . . . . . . . . . . . . . 38-39
Motion Control
Comparison of Sensorless Algorithms for PMSM Rotor Position Detection
By Stello Matteo Bille, Dino Costanzo, Antonio Cucuccio,STMicroelectronics andAlfio Consoli, Mario Cacciato, Giuseppe Scarcella, Giacomo Scelba, University of Catania . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40-43
New Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44-48
Future precision.Future performance.Now available.
CAS-CASR-CKSRThe transducers of tomorrow. LEMcreates them today. Unbeatable in size, they are also adaptable and adjustable. Not to mention extremely precise. After all, they have been created to achieve great perform-ance not only today – but as far into the future as you can imagine.
Several current rangesfrom 6 to 50 ARMS
PCB mounted Up to 30% smaller size (height)Up to 8.2 mm Clearance / Creepage distances+CTI 600 for high insulation
+5 V Single Supply Low offset and gain driftHigh Accuracy @ +85°C Access to Voltage Reference Analog Voltage output
www.lem.com
At the heart of power electronics.
Biricha Digital Power offering
Digital Power Workshop based on
TI´s F28x family.
For more information and your free drill
hole stencil please visit
www.biricha.com
2 Bodo´s Power Systems® July 2009 www.bodospower.com
TThhee GGaalllleerryy
Naturalmatch!
Features+15V/-10V gate voltage
3W output power
20A gate current
80ns delay time
Direct and half-bridge mode
Parallel operation
Integrated DC/DC converter
Electrical isolation for 1700V IGBTs
Power supply monitoring
Short-circuit protection
Fast failure feedback
Superior EMC
2SP0320 is the ultimate driver platform for PrimePACKTM IGBT
modules. As a member of the CONCEPT Plug-and-play driver
family, it satisfies the requirements for optimized electrical
performance and noise immunity. Shortest design cycles are
achieved without compromising overall system efficiency in
any way. Specifically adapted drivers are available for all
module types. A direct paralleling option allows integrated
inverter design covering all power ratings. Finally, the highly
integrated SCALE-2 chipset reduces the component count
by 80% compared to conventional solutions, thus signifi-
cantly increasing reliability and reducing cost. The drivers are
available with electrical and fiberoptic interfaces.
PrimePACKTM is a trademark of Infineon Technologies AG, Munich
2SP0320
SAMPLES AVAILABLE!
CT-Concept Technologie AG, Renferstrasse 15, CH-2504 Biel, Switzerland, Phone +41-32-344 47 47 www.IGBT-Driver.com
Bodo´s Power Systems® July 2009 www.bodospower.com
Enjoy the summer and recharge your enthu-
siasm for a return to work.
Three lucky winners at the PCIM Europe
walked away with a large toy excavator truck
for creative summertime engineering in the
sand. Games were sponsored by Microchip
and a few more goodies were raffled off to
kids there - they are our future.
Here’s an engineer of the future.
PCIM showed technology on the move.
From talking with a number of exhibitors, it is
clear that all forms of renewable energy are
key. Reducing energy consumption in any
application through improved efficiency and
using renewable energy sources will keep
the world in better shape – our children will
benefit.
Each percent of increased efficiency will
count. The new semiconductor materials, Sil-
icon Carbide (SiC) and Gallium Nitride
(GaN), presented at my PCIM Podium dis-
cussion, will be key to progress in achieving
higher efficiency. Many inverter applications
in wind and solar renewable energy will ben-
efit from the new technology. Inverters in
solar sources already realize the benefits of
SiC rectifiers and active switches will soon
be in use. All applications in renewable ener-
gy have a great opportunity to minimize loss-
es and together will have a huge impact on
the grid and its transition towards a “Smart
Grid”.
PCIM China and PCIM Europe both indicat-
ed a great deal of interest in efficient design
for the future. Many of the shows’ visitors
were decision-makers in high-level positions
- a good indicator that plans are already
underway for a better future and recovery in
the in the second half of 2009. Now’s the
time to begin preparing for a number of good
shows in fall.
Bodo’s Power Systems magazine has estab-
lished more than two dozen agreements with
conferences and trade shows around the
world. I am proud to announce that I have
published about 350 pages so far this year -
a result of reliable delivery each month since
June 2006. In times of an uncertain econo-
my, I am happy to present you with the
leading voices in power electronics and to
help all my contributors, whether authors or
advertisers. There is no better way to com-
municate. We all share one world. As a pub-
lisher I serve the world: one magazine, on
time, every time.
My Green Power Tip for a summer month:
Treat yourself to a computer-free week and
turn off your cell phone. Life will continue
without all of these communication toys.
Afterwards you might reconsider the impor-
tance of such devices. Enjoy the break,
save energy and reflect on what life was like
just a quarter of a century ago.
Hope to see you at the beach, it is summer
time.
Best regards
It is Summer – Beach Time !
Events
SEMICON West
San Francisco
July 14-16
http://www.semicon.com
EPE Barcelona
Spain
September 8-10
http://www.epe2009.com
Digital Power Forum
Costa Mesa CA
Sept. 21-23
http://www.darnell.com
Digital Power Seminar
Freising Germany
Sept. 15-18
http://www.biricha.com
V I E W P O I N T
4
A MediaKatzbek 17a
D-24235 Laboe, Germany
Phone: +49 4343 42 17 90
Fax: +49 4343 42 17 89
www.bodospower.com
Publishing EditorBodo Arlt, [email protected]
Creative Direction & ProductionRepro Studio Peschke
Free Subscription to qualified readers
Bodo´s Power Systems
is available for the following
subscription charges:
Annual charge (12 issues) is 150 €
world wide
Single issue is 18 €
circulation
printrun
25000
Printing by:
Central-Druck Trost GmbH & Co
Heusenstamm, Germany
A Media and Bodos Power Systems
assume and hereby disclaim any
liability to any person for any loss or
damage by errors or omissions in the
material contained herein regardless of
whether such errors result from
negligence accident or any other cause
whatsoever.
Winner of one of the toy excavator truck
6 Bodo´s Power Systems® July 2009 www.bodospower.com
N E W S
Texas Instruments announced that it will
expand its microcontroller (MCU) portfolio
with the acquisition of Luminary Micro, the
market-leading supplier of ARM Cortex-M3-
based 32-bit MCUs. The addition of Lumi-
nary Micro’s Stellaris® family of Cortex-M3
processors will accelerate TI’s ability to pro-
vide the industry’s most complete MCU port-
folio. This acquisition means that customers
can now enjoy the innovative capabilities of
Stellaris MCUs along with the proven experi-
ence and technical strength TI brings as a
global semiconductor provider.
Stellaris devices will allow TI to address
mainstream 32-bit MCU markets, giving cus-
tomers access to the general-purpose pro-
cessing power of the industry-standard ARM
Cortex-M3 core and the Stellaris family’s
advanced communication capabilities,
including 10/100 Ethernet MAC+PHY, CAN,
USB On-The-Go, USB Host/Device,
SSI/SPI, UARTs, I2S, and I2C. The transac-
tion closed on May 14, 2009.
(Seewww.ti.com/stellarispr.)
“Combining Luminary Micro’s design experi-
ence in Cortex-M3 processors with TI's
expertise in ultra-low power MSP430 MCUs
and high-performance C2000™ real-time
controllers now gives TI customers one MCU
source for almost any application – all com-
plemented by the industry’s most expansive
embedded processing and analog portfo-
lios,” said Brian Crutcher, vice president of
TI’s Advanced Embedded Control (AEC)
business.
www.luminarymicro.com
www.ti.com/mcu
TI acquires Luminary Micro
American Superconductor announced that it
has received an order worth more than $10
million from ACCIONA Energy, a division of
ACCIONA SA (MC: ANA) and a world leader
in renewable power, for its new Dynamic
VAR Ride Through (D-VAR RT) solution.
Building on AMSC’s highly successful D-VAR
platform, which provides critical dynamic
reactive compensation required to connect
many wind farms around the world to the
power grid, the company’s D-VAR RT prod-
uct enables individual wind turbines to con-
tinue operating smoothly by “riding through”
voltage disturbances on power grids that
might otherwise interrupt their operation.
The D-VAR RT product meets stringent grid
interconnection requirements, including
Spain’s new Procedimiento de Operación
12.3 requirement for both existing and new
wind turbines.
According to the Global Wind Energy Coun-
cil, Spain was the world’s third largest wind
power market at the end of 2008 with an
installed base of more than 16,000
megawatts (MW). Disturbances such as
momentary voltage dips can disconnect
many wind turbines and cause instability on
the transmission grid. Developed by Spain’s
transmission system operator Red Electrica
de España (REE), P.O. 12.3 requires that
wind turbines remain connected to the grid
through such events.
www.amsc.com
AMSC Signs Contract with Acciona Energy
ANSYS, Inc., a global innovator of simulation
software and technologies designed to opti-
mize product development processes,
announced a first milestone in coupling
ANSYS® and Ansoft™ products, successful-
ly performing multiphysics simulations that
involve electromagnetic applications. As
electronics become more embedded into
automotive, aerospace, industrial and con-
sumer products, engineers must consider
factors such as circuitry’s ability to withstand
vibration shocks, heat generation and elec-
tromagnetic interference. The combined
depth and breadth of solutions from ANSYS
is key to solving problems that involve these
complex systems. In performing several
case studies, ANSYS engineers deployed
the electromagnetic effects determined by
Ansoft software directly in ANSYS thermal
and structural simulation. Work is ongoing to
fully integrate Ansoft software directly into
the ANSYS® Workbench™ platform for
future bidirectional and seamless operation.
For example, a high-power electronic con-
nector used in a military radar application to
connect a transmitter to an antenna must be
engineered from electromagnetic, thermal
and structural perspectives to ensure suc-
cess. The simulation was performed by cou-
pling Ansoft’s HFSS™ software with the
ANSYS Workbench environment, using
advanced thermal and structural capabilities.
Engineers used HFSS to ensure that the
device was transmitting in the proper path,
by calculating the high-frequency electro-
magnetic fields, power loss density distribu-
tion and S-parameters. In such high-power
applications, it is critical to determine the
temperature distribution to ensure the device
stays below temperatures that cause materi-
al failure, such as melting. The power loss
density results from the HFSS simulation
were used as the source for the thermal sim-
ulation performed within ANSYS® Mechani-
cal™ software, which simulated the tempera-
ture distribution of the device.
www.ansys.com
Physics Brought Together In Complex Electronic Design Application
IPC — Association Connecting Electronics
Industries® released the spring 2009 edition
of its quarterly business report, Supply Chain
Tracker, this week and although it showed
continuing economic contraction, it also indi-
cated the first signs of recovery in the elec-
tronics industry.
IPC's global statistical programs for several
key industry segments all show worsening
year-on-year growth rates in first quarter
2009, after growth rates turned negative in
late 2008. IPC's North American Electronics
Industry Performance Index fell 29 percent.
This is the third straight quarter the index
has declined. This index, a new addition to
Supply Chain Tracker, monitors the perform-
ance of the North American electronics sup-
ply chain.
Some leading indicators, however, are
beginning to show improvement. The April
2009 book-to-bill ratio for the North American
printed circuit board (PCB) industry climbed
for the third straight month from 0.89 to 0.97.
This ratio still indicates lagging demand, but
it is trending toward 1.0, the point of parity
between bookings and shipments. The North
American EMS book-to-bill ratio inched up to
0.95 at the end of first quarter. Semiconduc-
tor sales, while still in negative territory,
improved in first quarter 2009.
www.IPC.org
First Signs of Recovery
7www.bodospower.com July 2009 Bodo´s Power Systems®
N E W S
Rogers Corporation has made two important
announcements at the PCIM Europe 2009
Exhibition. The first time that the UL746C
registration has been granted to a supplier of
laminated busbar components. The RO-
LINX laminated busbars are used in industri-
al drives, mass transit, and alternative-ener-
gy power-distribution systems.
The laminated busbars are a critical compo-
nent in medium-to-high-power variable-
speed drives (VSDs) in these systems.
Rogers' components are designed to
achieve higher efficiency by limiting switch-
ing losses. In addition, the new UL746C rat-
ing provides the assurance of safety and
reliability backed by Underwriters Laborato-
ries. Previously, busbars and other power-
conversion components had been limited to
the UL94 safety rating for flammability.
The UL746C rating encompasses UL94, but
also covers hot-wire resistance to ignition
(HWI), high-current arc resistance to ignition
(HAI), comparative tracking index (CTI), and
relative thermal index (RTI) safety areas.
With the UL746C rating, Rogers' customers
are assured of reducing internal engineering
validation time, with a corresponding shorter
time to market for their own products.
Rogers ' second major announcement at
PCIM Europe 2009 comes from a joint effort
between their Thermal Management Solu-
tions Division (www.rogerscorp.com/tms)
and materials innovator Element Six. The
two companies will introduce HEATWAVE™
AlSiC D3 Technology as an advanced solu-
tion to their existing high-performance ther-
mal management products. The AlSiC D3
thermal materials are based on Element
Six's unique silicon cemented diamond
(ScD) technology combined with aluminum
encapsulation.
The thermal-management materials, which
are well suited to three-dimensional designs,
feature high thermal conductivity and low
coefficient of thermal expansion. Element Six
(www.e6.com), with more than 50 years in
materials development, has refined
advanced processes such as chemical vapor
deposition (CVD) to develop diamond-based
supermaterials with superior optical, thermal,
electrical, and chemical properties for a wide
range of industries.
www.rogerscorp.com
Breakthrough UL Rating & Advanced Thermal Solution
Isabellenhütte, Semikron and Siemens Drive
Technologies have cooperated to develop a
3 phase shunt module for the Siemens
SINAMICS G120 frequency inverter series.
This series is intended for measuring phase
currents of up to 20 A through shunts on the
circuit board. Apart from this, currents of up
to 400 A can be measured with outstanding
precision and long-term stability using the
jointly developed shunt modules in Semi-
kron’s Semitrans housing.
The PM240 high-performance parts for
SINAMICS G120 inverters make it possible
to extend the current range of Isabellen-
hütte’s Type BVR electron-beam welded
composite material precision resistors to 800
A and power output to 132 KW. For this pur-
pose, the resistance of shunts used in this
area was reduced from 3 to 1 mOhm, and
the module’s layout was optimized. The
housing’s interior was also adapted to the
terminal link design. The isolated shunt
installation, integration of 3 phases in one
case and optimized power loss heat dissipa-
tion by a DCB substrate, as well as the mod-
ule’s size, were not changed. Generally, the
module’s resilience and reliability were
improved. www.isabellenhuette.de
Shunts Replace Transformers in Large Inverters
The Korean company LS Industrial Systems
and Infineon Technologies AG announced
the establishment of the joint venture LS
Power Semitech Co., Ltd. which will focus on
the development, production and marketing
of molded power modules for white good
applications. The establishment of the joint
venture paves the way for Infineon and LS
Industrial Systems to more rapidly access
the promising market for energy efficient
household appliances, such as washing
machines, refrigerators and air conditioners,
and also for other low-power consumer and
standard industrial applications. The use of
variable-speed motors to reduce the energy
consumed by household appliances is grow-
ing in response to regulatory requirements
and consumer demand. Concurrently, smart
design of drive control electronics to make
best use of these motors presents manufac-
turers with further opportunities for efficiency
and savings.
www.infineon.com
Molded Power Module Business for White Goods
According to the latest update of the SEMI
World Fab Forecast database, spending on
fab construction projects has seen a consis-
tent quarterly decline since 2008, and on a
year-over-year basis is expected to fall by 56
percent in 2009. On a global scale, construc-
tion spending is at its lowest level in 10
years. However, the latest data from the
report suggest an increase in investments
for both fab construction projects and fab
equipping in the second half of 2009, with
the trend continuing into 2010. In 2010,
investments in fab construction projects are
expected to almost double and spending on
equipping fabs may increase by as much as
90 percent year-over-year from the signifi-
cant declines expected in 2009, according to
the report.
Investments are actually increasing in the
Americas, with a total quarterly spending
increasing to almost US$1 billion, mainly due
to major investments announced by Intel, as
the company moves forward on a planned
upgrade to 32nm.
According to the report, 19 fab facilities
closed in 2008, and about 35 facilities will
close in 2009, though the number of clo-
sures should decline in 2010 as only 14
facilities are expected to close. Nine fabs
are expected to launch operations in 2009.
Overall the trend of new facilities commenc-
ing operations has slowed since 1995, due
to the fact that most new fabs are 300mm
Megafabs for memory production, meaning
fewer but larger fabs are needed.
www.semi.org
SEMI World Fab Forecast Reveals Signs of Increased Investment
Power Integrations, the leader in high-volt-
age integrated circuits for energy-efficient
power conversion, introduced a new online
tool that enables designers of external
chargers and adapters to instantly determine
whether their product complies with world-
wide energy-efficiency regulations. The new
External Power Supply Efficiency Compli-
ance Calculator quickly and easily compares
power supply performance measurements
against the maze of specifications that now
apply to external chargers and adapters, sig-
nificantly simplifying the design engineer’s
task of verifying compliance. The calculator
currently checks compliance to the following
standards:
ENERGY STAR EPS (version 2.0): Spon-
sored by the U.S. Department of Energy and
the Environmental Protection Agency,
ENERGY STAR is one of the most visible
efficiency certifications worldwide.
EISA 2007: The first mandatory U.S. federal
EPS efficiency standard, the EPS limits in
EISA 2007 were adopted from the California
Energy Commission’s Appliance Efficiency
Regulations.
European Commission Code of Conduct
(version 4): The European Commission
Code of Conduct (CoC) issued version 4 of
its EPS specification in April 2009.
EC Eco-design Directive: The European
Commission’s Eco-design Directive for exter-
nal power supplies, scheduled to take effect
in April 2010, will align with the EISA 2007
standard for Tier 1 and ENERGY STAR (ver-
sion 2) for Tier 2.
China USB Charger Specification (YD/T
1591-2006): China’s Communication Indus-
trial Standard mandates a USB connector
and power output with a no-load consump-
tion of ≤300 mW for mobile telecommunica-
tion terminal equipment power supplies.
EC Integrated Product Policy (IPP): In
2008, a group of leading mobile-phone man-
ufacturers developed a “Five-Star” rating
system for mobile phone adapters/chargers,
specifying no-load power consumption down
to ≤30 mW -- well below any current or pro-
posed government standards.
Comments Rich Fassler, manager of energy-
efficiency programs at Power Integrations:
“Energy-efficiency specifications and stan-
dards have become increasingly complicat-
ed, and the landscape is constantly chang-
ing. For example, in April 2009 alone, two
important updates occurred -- the EC Code
of Conduct issued version 4 of its specifica-
tion, and the upcoming EC Eco-design stan-
dard was approved by European Parliament.
More than ever, power supply designers and
those sourcing external power supplies for
use with their end products need an easily
accessible, up-to-date database of world-
wide current and proposed regulations.”
Continues Fassler: “Power Integrations’ new
External Power Supply Efficiency Compli-
ance Calculator means that designers no
longer have to consult multiple sources
when checking the efficiency compliance of
their EPS designs – they can simply enter
their data and achieve an immediate, com-
prehensive, and accurate analysis.”
For more information about energy-efficiency
standards and standby energy waste, please
visit Power Integrations’ Green Room web-
site at www.powerint.com/greenroom.
References:
PI Energy-Efficiency Compliance Calculator:
www.powerint.com/sites/default/files/images/f
inal_final.swf
ENERGY STAR:
http://www.powerint.com/en/green-room/reg-
ulations-agency/energy-star-us
EISA 2007: http://www.powerint.com/green-
room/agencies/u-s-federal-government
EC CoC: http://www.powerint.com/en/green-
room/regulations-agency/eu-code-conduct
EC EuP Ecodesign Directive:
http://www.powerint.com/en/green-
room/agencies/ec-eup-eco-directive
China USB Charger Spec: http://www.pow-
erint.com/sites/default/files//greenroom/docs/
china_usb_spec_050409.pdf
EC IPP: http://www.powerint.com/en/green-
room/regulations-agency/ec-ipp-mobile-
device-charger-rating
Power Integrations is the leading supplier of
high-voltage analog integrated circuits used
in energy-efficient power conversion. The
company’s innovative technology enables
compact, energy-efficient power supplies in
a wide range of electronic products, in AC-
DC, DC-DC and LED lighting applications.
Since its introduction in 1998, Power Integra-
tions’ EcoSmart® energy-efficiency technolo-
gy has saved an estimated $3.4 billion of
standby energy waste and prevented mil-
lions of tons of CO2 emissions. The compa-
ny’s Green Room web site
(www.powerint.com/greenroom) provides a
wealth of information about “energy vam-
pires” and the issue of standby energy
waste, along with a comprehensive guide to
energy-efficiency standards around the
world. Reflecting the environmental benefits
of EcoSmart technology, Power Integrations
is included in clean-technology stock indices
sponsored by the Cleantech Group (Amex:
CTIUS) and Clean Edge (Nasdaq: CELS).
www.powerint.com
G R E E N P R O D U C T O F T H E M O N T H
8 Bodo´s Power Systems® July 2009 www.bodospower.com
Energy-Efficiency CalculatorNavigates the Maze of External
Power Supply StandardsOnline Resource for Design Engineers Quickly Determines
Compliance with Worldwide EPS Energy-Use Rules
10 Bodo´s Power Systems® July 2009 www.bodospower.com
SKiiP4, the new generation of intelligent
IGBT power modules, boasts a longer serv-
ice life than non-sintered modules and can
be used in higher temperature applications.
The SKiiP power pack is the most powerful
intelligent power module on the market and
is 33% more powerful than its predecessor
SKiiP 3. The IPM is used predominantly in
wind and solar power applications, traction
applications, elevator systems and industrial
drives with high outputs of between 400 kW
and 1.8 MW.
For comparable conditions and module
sizes, the SKiiP4 provides 33% more power
than the current version of this module fami-
ly, SKiiP3. On the one hand, this allows for
the development of more powerful or more
compact frequency converters, thus reducing
costs. This increase in power is down to the
use of an innovative pressure contact sys-
tem, an improved heat sink and IGBT4 and
CAL4 chip technology. In addition, six paral-
lel half bridges have been used for the first
time at the upper power end instead of four,
as was the case up till now.
In SKiiP4 modules, the semiconductor chips
are not soldered to the ceramic substrate but
are joined using sinter technology, meaning
that higher operating temperatures are pos-
sible with no compromise to – or in some
cases even increased - reliability. The sinter
bond is a thin silver layer which has a lower
thermal resistance than a bond with solders.
Thanks to the high melting point of silver,
premature material fatigue can be prevent-
ed.
Like its predecessors, SKiiP 4 is based on
well matched components: heat sink, power
module, driver and protective sensors/func-
tions. Here, the mounting and connecting
technology, which is based on a pressure
system, plays a crucial role. Customers can
also opt to have unique burn-in tests where
the IPMs are run under real operating condi-
tions performed. These tests enable prema-
ture silicon failure to be identified and the
defective chips removed. In the tests the
modules are exposed to the maximum possi-
ble junction temperature.
The solder-free pressure contact system and
the integrated laminated power rails ensure
homogenous current distribution. Every
IGBT and diode chip is connected to the
main terminal separately, keeping the mod-
ule resistance very low. The chips are not
soldered to the ceramic substrate but are
joined in a sinter process. As these modules
have no base plate, the solder-free connec-
tion between DCB and heat sink is quasi-
flexible, which is why the thermal cycling
capability has no upper limits. SKiiP4 will be
available with blocking voltages of 1200V
and 1700V in dual-pack topologies with
three, four or six parallel half bridges per
IPM.
Digital signal transmission for the switching
signals is the key to ensuring both a very
high degree of reliability and interference
immunity for switching signals. In addition to
various technical advantages, this ensures
high signal integrity and, consequently, inter-
ference immunity. Transmission depends on
component parameters, is highly robust and
not sensitive to temperature fluctuations or
ageing effects. The switching and sensor
signal transmission channels feature galvan-
ic isolation, meaning the user does not have
to provide additional isolation. To round off,
the new SKiiP4 IPM also features a multi-
stage output stage which ensures the reduc-
tion of overvoltage and includes various
other protective functions. Finally, a diagno-
sis channel is available for optimum evalua-
tion on the customer side.
.
www.semikron.com
B L U E P R O D U C T O F T H E M O N T H
Power Pack Offers 33%Increase in Power Density
Up to 33% more power in the new SKiiP 4 IPM.
11www.bodospower.com July 2009 Bodo´s Power Systems®
M A R K E T
ELECTRONICS INDUSTRY DIGESTBy Aubrey Dunford, Europartners
GENERAL
After an 8 percent decline
in 2008, the German elec-
tronic components market
is expected to decline 16
percent in 2009 to € 13.2
billion, so the ZVEI. Indus-
try observers however are
optimistic that the market will recover in
2010, growing at most 5 percent.
During the current year, the world market will
decline by 16 percent to $ 333 billion (in
euro, a decline of 9 percent to € 246 billion,
before increasing by 5.5 percent to $ 351 bil-
lion in 2010.
SEMICONDUCTORS
Utilization of worldwide semiconductor
capacity is expected to rise to 60 percent in
the second quarter of 2009, up from 49 per-
cent in the first quarter, so iSuppli. This will
mark the first quarterly sequential increase
since the second quarter of 2008.
More than 60 companies and research insti-
tutions in Germany develop technologies on
a large scale that will significantly reduce
energy consumption of microchips and infor-
mation technology. The group, called the
'Cool-Silicon-Cluster', which is located in the
Dresden area, officially starts its project
financed by both the Federal and the State
governments. Altogether the federal govern-
ment as the initiator of the cluster research
will make € 40 M in funding available. With
the additional funds from the Saxony's Min-
istry of Science and the Arts, the cluster will
be able to apply for funds totaling more than
€ 100 M in the coming years. Combined with
contributions from Cool Silicon partners a
total volume of more than € 150 M will be
available.
The management of the Swindon foundry in
Cheney Manor had acquired the semicon-
ductor foundry business from MHS Electron-
ics UK. MHS had acquired this business in
February 2008 and the company was subse-
quently placed into administration in Febru-
ary 2009.
Worldwide silicon wafer area shipments
were 940 million square inches during the
first quarter 2009, a 34 percent decrease
from the previous quarter. The new quarterly
total area shipments are 57 percent below
Q1 2008 shipments and at the lowest levels
since 2001, so SEMI.
North America-based manufacturers of semi-
conductor equipment posted $ 253 M in
orders in April 2009 and a book-to-bill ratio
of 0.65, so SEMI. The bookings figure is
three percent greater than the March 2009
level and about 77 percent less than in April
2008.
European Investment Bank (EIB) lends €
400 M to Wacker Chemie for its new polysili-
con production plant construction which will
locate in Saxony region of Germany. Wacker
now secured € 800 M for the construction of
the plant. The new plant will have 10,000
tons in annual production capacity which will
begin from 2011. The plant is expected to
create 450 jobs.
PASSIVE COMPONENTS
World connector sales of $ 7.363 billion in
1Q09 are comparable to sales levels in
2003, so Bishop & Associates. The first
quarter of 2009 is over, ending with orders
down 41.7 percent, and sales down 35.7
percent. The industry consensus for 2009
sales appears to be a decline of 25 percent
at $ 32.983 billion.
Revenues for the PCB industry in Germany
continue declining in February; compared to
last year, the market was down 40 percent
for the month, so the ZVEI/VdL. Both incom-
ing orders for February, as well as the cumu-
lative orders for the first two months of the
year, were less than half compared to the
same periods last year. The book-to-bill-ratio
reached 0.78.
Littelfuse announced restructuring plans
which include the transfer of manufacturing
from the company’s Duensen, Germany site
to Mexico. Research and development,
sales and other functions will remain in Ger-
many.
Epcos has acquired the development activi-
ties of Mems microphones from Technitrol.
The acquisition sum is in the mid single-digit
million euro range. This acquisition opens up
the growth market for miniaturized MEMS
microphones to Epcos. They are used espe-
cially in mobile phones and Bluetooth head-
sets. The market volume of these applica-
tions is in the triple-digit million euro range.
Planned investments for financial 2009/10 at
AT&S amount to € 30 M, half of which
relates to projects begun in 2008/09. AT&S
sales compared with the last financial year
fell by 7.4 percent to € 449.9 M. Final con-
solidated net income came out at -€ 5.8 M
for the biggest European PCB provider.
OTHER COMPONENTS
Seagate Technology has initiated a restruc-
turing plan that includes a reduction of
approximately 1,100 employees or 2.5 per-
cent of the company's global workforce.
K2 Energy Solutions, a manufacturer of
rechargeable battery systems, has
announced the construction of a lithium-ion
battery factory in Varkaus, Finland. The $ 44
M facility is expected to be completed in fall
of this year, with full battery production
beginning in early 2010.
GE Transportation announced plans to build
a $ 100 M plant for advanced storage batter-
ies. GE has invested more than $ 150 M to
develop advanced battery technologies,
including a high energy-density sodium-
based chemistry battery.
DISTRIBUTION
European distribution billings in 1Q09
declined by 3 percent, when compared to
the previous quarter and declined by 21.4
percent compared to the same period last
year, so the IDEA.
TTI, a global distributor of passive, connec-
tor, electromechanical and discrete compo-
nents, has signed a Europe and North Amer-
ica franchise agreement with Epcos, a Ger-
man supplier in power capacitors, SAW com-
ponents, surge arresters, thermistors, varis-
tors and piezo actuators. In Europe, Epcos
is the top manufacturer of aluminium elec-
trolytic capacitors, EMC components and fil-
ters, ferrites, film capacitors, and transform-
ers and chokes. Omron Electronic Compo-
nents also announced TTI, as their 2008
Distributor Partner of the Year. Awards were
presented to both companies that comprise
the TTI group --Mouser Electronics and TTI.
The MSC-Gleichmann group and Vectron
International signed a pan-European distri-
bution agreement for Bulgaria, Germany,
Greece, Austria, Poland, Rumania, Slovakia,
Ukraine and the Czech Republic.Vectron is a
supplier of high-precision frequency control,
sensor and hybrid product solutions.
This is the comprehensive power related
extract from the « Electronics Industry Digest
», the successor of The Lennox Report. For
a full subscription of the report contact:
[email protected] or by
fax 44/1494 563503.
www.europartners.eu.com
12 Bodo´s Power Systems® July 2009 www.bodospower.com
Change is scary. But,
change is inevitable.
In fact, you should be
happy that change is
inevitable because it is change that facili-
tates your gainful employment! If it were not
for change, your career would have been
over after the first design.
Not only is change inevitable but, especially
related to power design requirements, the
rate of change (the change of change) is
increasing. Thus, it is the happiest of times
and the scariest of times.
Technology evolution, regulatory adjust-
ments, and economic factors, drive change
in power conversion requirements. Yet it is
not this change that we, as power conver-
sion engineers, fear. In fact, this is the
change that we embrace; it brings challenge
and excitement to our jobs; it is, in many
ways, what makes us happy. We are more
than happy to use what we know to solve
ever evolving challenges.
What is so scary? Changing our toolset is
scary. Using what we know is comfortable;
using what we don’t know is unsettling.
Changing to a new controller is a scary thing
to do. Let’s face it, there are no perfect con-
trollers. But, you have figured out how to
use the controller that you are using right
now. You have found ways to work around
all of the little “issues”…the little “surpris-
es”…that popped up when you first tried to
use it. We’ve all had our disasters along the
way, though, where those little surprises
were no so little and came at exactly the
wrong time. We live with the scars from
these encounters. We are wary, very wary
of changing controllers.
This fear of changing controller technology is
both real and justified. Let’s examine what
could go wrong:
# The features do not behave the way you
understood them to behave. For example,
you thought that the response to a current
fault was foldback or hiccup but it turns
out that it shuts down and does not
restart.
# You need more features than the part can
provide so you add circuitry that takes
more time and money than you had
planned. For example, it does not handle
prebias adequately so you need to add
additional circuitry to take care of prebias
startup.
# The part has features that interfere with
what you need to do so you add circuitry
to negate the feature and this takes more
time and money than you had planned.
# When you first selected the controller, it
met or marginally met your needs but now
that you are well into the design, your
needs have changed and the part is going
to take a lot of work to fix.
# There is a bug in the part. 3-6 months for
a metal mask spin. 6-12 months for an all
layer change.
# The part is more noise sensitive than your
lab testing initially revealed. You find this
out in production…too late to add filters.
# The manufacturer had a process change
between you designing with the samples
and the production parts were made avail-
able.
# Your competitor used a different part and
now offers features that your controller
cannot provide.
In each of these examples, the common ele-
ment is the inflexibility of the controller to
adapt to your needs. You get what you get
and for good reason. Imagine the task of
your controller manufacturer when deciding
on what part to build. The manufacturer will
talk to a dozen or more customers and each
will have a different set of requirements.
The result is a compromise in terms of the
lowest common denominator of controller
requirements. Although intended to have
something that pleases everyone, the result
falls short of everyone’s expectations.
Digital controllers break this paradigm.
Today’s advanced digital controllers combine
the best of analog (for high speed events)
and digital (for flexibility) technologies.
These controllers typically have a dedicated
function PWM engine that is configurable by
way of an onboard microcontroller with non-
volatile memory and a communications link.
The controller is a hybrid of a sort in that it is
fixed function (power converter) but config-
urable through firmware. Not only are the
operational parameters (voltage, current,
switching frequency, etc.) configurable, but
also the response (faults, transients,
enable/disable) and features. Additional
capability includes telemetry, flexible pin defi-
nition, and adaptive behavior to optimize the
performance in real time.
So, when changing over to digital controller
technology, fear no more:
# Digital controllers can provide multiple
behaviors for any given feature. Simply
open up the provided graphical user inter-
face (GUI) and point and click on the
behavior that you prefer. Not exact what
you are looking for? Then most likely the
behavior is defined in firmware and you
can get a fast fix in a few days with a
qualified part in a few weeks.
# The nice thing about digital is that is
comes with, most likely, more features
than you need. Turning them on or off
takes a simple point and click on the GUI.
Don’t see the feature that you need. You
guessed it…it’s probably a firmware addi-
tion that can be realized in a few days.
# Is noise a problem? Digital controllers
come with digital noise filters…again,
point and click to turn on and configure.
# The competitor has come up with some-
thing new or you have an idea that can
increase your sales…work with your digi-
tal controller supplier…there is a good
chance you can see your ideas realized
with firmware.
# Do you need to make an adjustment to
the design at the last minute? Put away
the soldering iron…open up the
GUI…take a few seconds to make the
change and you are done. No more lifting
traces, no more searching for the right
replacement component, no more late
nights!!
# In fact, stop guessing about what is going
on when you first turn it on. With the
onboard telemetry you can analyze the
operation and faults in real time.
Changing power controller technology is no
longer scary. In fact, controller change is
exciting and enabling. Have no fear…Digital
Power is here!
www.intersil.com
G U E S T E D I T O R I A L
Have No Fear: Digital Power is Here!By Chris Young, Sr. Manager,
Digital Power Technology, Intersil
ElektrischeAutomatisierungSysteme und Komponenten
Fachmesse & Kongress24.–26. Nov. 2009Nürnberg
SPS/IPC/DRIVES/
Mes
ago
Mes
sem
anag
emen
t G
mb
H, P
ost
fach
10
32 6
1, 7
0028
Stu
ttg
art,
sp
s@m
esag
o.c
om
, Tel
. +49
711
619
46-8
28
Ihre kostenlose Eintrittskarte
www.mesago.de/sps
14 Bodo´s Power Systems® July 2009 www.bodospower.com
Led by the growing photovoltaic (PV) market, the outlook for inverters
used in alternative energy resource technology is expected to remain
strong. Industry growth in this application will be driven by a combi-
nation of government incentives and declining PV module prices.
Projected to make up over 95% of the market, the inverters used in
PV installations, both small (1-5kW) and large (>6MW), will far out-
pace those used in either wind or fuel cell applications. In fact,
according to the Electric Power Research Institute (EPRI), grid-con-
nected PV systems are expected to account for more than 90% of
PV capacity additions in 2011, up from about 50% in 2000, and from
less than 10% a decade earlier. The accelerating worldwide growth in
grid-tied PV will be driven by continuing technology performance and
cost improvements, strong deployment incentives, and growing con-
sumer interests, as well as renewable portfolio standards, concerns
about climate change, and other policy mandates.
In Europe, the primary driving force in the PV market are feed-in tar-
iffs, which have been successfully used in 16 EU countries, most
notably Germany, Italy and Spain. In fact, in Europe renewables com-
prise the fastest- growing segment of the energy market. The primary
factor driving the PV industry in Germany is the Renewable Energy
Sources Act (EEG in German) which mandates feed-in tariffs. A key
feature of the German Renewable Energy Act is that it guarantees
feed-in tariffs for 20 years, and as a result, no price fluctuations need
to be taken into consideration during the planning process of a
renewable energy program. This law provides an incentive for Ger-
man property owners to own PV installations, providing a ready-
made domestic market for PV products made in Germany. In Ger-
many, installations are 80% residential, creating local installation jobs
and making PV visible as a source of electricity generation, with grid
parity expected between 2012 and 2015.
The other large European feed-in tariff program is in Spain. Together,
in 2008, they made up over 84% of the region’s installed PV capacity.
The PV boom in the Spanish market was significant for a number of
reasons, most notably for the generous feed-in tariffs which led to a
global market explosion of more than 100% compared to 2007. How-
ever, the feed-in tariffs implemented in the Spanish PV market result-
ed in a number of unintended consequences, including the overpro-
duction of solar modules, which many felt were already headed
towards an oversupply situation. Due to a number of issues, includ-
ing an overwhelming backlog of solar PV applications, difficulty
financing projects, falling oil prices and other challenges, the new
government in Spain halted the costly stimulation of solar PV. The
result was a 40% decrease, or more than 2500 MW of global market
volume in 2009. So although Spain’s PV market grew by >2.66 GW
of new installed power in 2008, new laws will cap that growth at
500MW in 2009.
The photovoltaic systems resulting from these programs range from
small off-grid installations providing hundreds of watts to large central
utility power plants. However, the fastest-growing sectors of the PV
industry are large commercial and central power generation systems.
These systems can range in size from tens of kilowatts for commer-
cial rooftop systems to over 100 megawatts for utility-scale central
power generation plants occupying many acres of land. The inverters
for these systems are central inverters, although a parallel combina-
tion of several string inverters may also be used. Some of the com-
panies manufacturing these inverters include SMA Technologie AG,
Xantrex, Satcon Technology Corp, Studer Innotec, Fronius, Advanced
Energy Industries, Magnetek, PV Powered, Solectria Renewables
and others. Each of these companies should see significant opportu-
nities as the size of PV installations continues to increase.
Typical of these installations is a 300 kilowatt (kW) solar tracker plant
in Toledo, Spain. The objective of the project is to combine water, ani-
mal life, plants and buildings into an integrated facility. The solar plant
will be composed of six 100 kW installations, with the first phase
including three Xantrex GT100E central inverters. The GT100E com-
mercial solar inverter is a 100 kW three-phase advanced power elec-
tronics system for grid-connected solar arrays. An even larger facility
can be seen in Germany at the Waldpolenz Energy Park, which will
eventually comprise 40 megawatts (MW) of production capacity. SMA
will provide 70 central inverters, the Sunny Central 500HE, each with
a capacity of 500 kWh. In each case, two inverters are located in one
station for a single module field. A total of 35 of these stations will
make up the facility.
As PV facilities continue to increase in size, inverter technology will
need to meet the challenges presented with the trend towards larger
installations. An example of this can be seen in SMA America Inc.,
the U.S. subsidiary of SMA Solar Technology AG, which introduced
the Sunny Central 250U, a 250kW inverter designed to be the heart
of a large-scale commercial PV system. Generally acknowledged to
be the largest manufacturer of inverters in the world with an estimat-
ed global market share of 35%, SMA’s entrance into the higher-
wattage commercial market is a significant development for the entire
industry. Discussions with SMA indicate that the company sees a
growing industrial/commercial market and wants to expand outside
the residential sector. Other companies featuring inverters in the
250kW to 500kW range are Xantrex, Advanced Energy and PV Pow-
ered, followed by Satcon and Magnetek, which each introduced a
1MW inverter system. The new generation of photovoltaic inverters
fueled by this growth present system integrators with new solar array
design and installation challenges.
In addition, the growing trend towards large-scale developments is
evident in the number of partnerships and cooperative efforts being
formed in the PV industry. In 2009, Xantrex Technology Inc., a sub-
sidiary of Schneider Electric, announced the launch of the Solar
M A R K E T
Incentives Fuel PVInverter Opportunities
By Richard Ruiz Jr., Research Analyst, Darnell Group
www.bodospower.com
Power Conversion Substation (SPCS) for the North American mar-
ket. Sustainable Energy Technologies Ltd. also announced that it
had entered into an agreement with FOTOVOLTAICA 10 CM SA
(“Grupo Diez”) to collaborate on an inverter system solution for a 33
kW solar concentrator tracker developed for Grupo Diez. Grupo
Diez is a large solar project developer based in Toledo, Spain.
Other significant business combinations and cooperative efforts
include Sunfilm AG and Sontor GmbH, which merged to become
what they claim will be one of the world’s largest providers of tan-
dem junction silicon-based thin-film modules, which is one of the
most significant growth areas within the photovoltaic industry.
Given this trend, a large number of existing manufacturers and
new entrants are seeing the market for PV inverters becoming more
competitive. As a result, prices are expected to fall, and products
with innovative features and greater efficiency are expected to flood
the marketplace. This anticipated activity is being reflected in the
number of new investments in production capacities that have been
announced over the past year. As an example, Sputnik Engineering
AG is increasing its annual production capacity for string and cen-
tral inverters with its new facility, which went into initial operation in
March in Biel, Switzerland. According to the company, its goal is to
see its capacity reach 400 MW annually by the end of 2009.
The new Sputnik plant will assemble central inverters with outputs
from 50 kW upwards. The company identified Germany, Spain and
Italy, along with France and Greece as areas showing significant
opportunities for market growth. In addition, in 2008, SMA Technolo-
gie AG, began construction on what the company says is the world’s
largest solar inverter factory in Kassel, Germany. SMA expects con-
tinuous growth in the coming years, and to meet this increasing
demand the company is developing a new 15,000 square meter
production facility that is designed to be completely CO2-neutral
and has a virtually independent power supply with a MW-scale
Building Integrated PV (BIPV) system. A BIPV system involves inte-
grating photovoltaic modules in to the building envelope material
and power generators. Common in residential buildings in many
countries including Japan, Germany, the US, Switzerland and oth-
ers, the further expansion of BIPV in commercial projects will help
expand the PV industry and provide additional opportunities for
manufacturers of inverters for years to come.
http://www.darnell.com/store/product_info.php?products_id=92
www.powerpulse.net
MAKING MODERN LIVING POSSIBLE
DANFOSS SILICON POWER
SILICONPOWER.DANFOSS.COM
Wind power the cool wayCustomized power modules made for high performance
It cannot be stressed enough: Efficient cooling is the mostimportant feature in power modules. Danfoss Silicon Power’s cutting- edge ShowerPower® solution is designed to secure an even cooling across base plates. In addition, our modules can be customized to meet your wind power requirements in detail, offering: High quality, high performance, extended life and very low life cycle costs. In short, when you choose Danfoss Silicon Power as your supplier you choose a thoroughly tested solution withunsurpassed power density.
Please go to siliconpower.danfoss.com for more information.
ShowerPower®
Digital Power Workshop
based on
TI´s F28x family.
For more information and
your free drill hole stencil
please visit
www.biricha.com
Biricha Digital Power offering
The past two decades has seen major improvements in silicon power
semiconductors in terms of reduced losses and wider Safe-Operat-
ing-Area (SOA) performance. The main drive behind these advances
is the continuous need for higher power densities in new power sys-
tem designs. Therefore, devices such as the IGBT and the free-
wheeling diode continue to provide Megawatt applications such as
traction, industrial drives and transmission and distribution with opti-
mum components which have enabled with each improved genera-
tion a clear leap in power levels.
The IGBT in particular with its inherent advantages including a con-
trolled low power driving requirement and short circuit self limiting
capability has experienced many performance breakthroughs, which
has resulted in its wide employment in many high power applications.
The most recent low loss and high SOA improvements were mainly
due to the introduction of the Soft-Punch-Through (SPT) or Field-
Stop (FS) thinner silicon concepts combined with advanced planar or
trench emitter structures. Similar advances were also achieved for
the anti-parallel freewheeling diode to match the continuously improv-
ing IGBT. Today, high power IGBT modules have voltage/current rat-
ings ranging from 1700V/3600A to 6500V/750A.
High Power Semiconductor Devices and the Future Options
Currently, it could appear that the development of high voltage silicon
power devices has reached a limit with regards to further reductions
in the device total losses. The state-of-the-art SPT/FS structures are
close to the so-called “Silicon Design Limits” and the emitter plasma
enhancement will only provide smaller steps with fine optimization of
the IGBT cell designs. In addition, the extremely robust modern IGBT
designs already provide the necessary SOA performance with ade-
quate margins. Hence, one can conclude that the possibility for
achieving major leaps in increased power densities for silicon is
becoming more restricted. Although, increasing the device total area
monolithically or equally through device paralleling can provide the
higher power solution for some applications, this approach will
remain selective due to its negative impact on the cost, size and
complexity of the overall system. Therefore, the quest for more
advance device concepts will remain as the trend continues for next
generation Megawatt systems with increased efficiencies.
The assumed technological barrier or silicon limit has led to the
recent trend towards increased operating temperatures compared to
the traditional 125°C maximum junction temperatures operation limit.
The low losses and high SOA of modern IGBTs and diodes has
enabled this step to be taken while also focusing on reduced leakage
currents and improved package reliability as the major limiting fac-
tors. An increase of around 10-15% in total output current capability
can be predicted with this approach.
Furthermore, there are continuous developments in Wide Band-Gap
(WBG) materials such as SiC and GaN for power semiconductors
due to its ten fold thinner base region structures having substantial
loss reduction potentials and the high operating temperature capabili-
ty when compared to silicon. With the clear progress achieved for
ultra fast unipolar power diodes rated up to 1700V and many switch
concepts demonstrated recently, WBG power devices are now being
regarded as the next major performance leap. Nevertheless, the cur-
rent cost of such devices and some technological and performance
aspects yet to be fully resolved especially for higher voltage/current
devices will continue to seriously delay the introduction of WBG com-
ponents in Megawatt applications. This while also taking into account
that silicon power devices could still provide another breakthrough in
performance.
The Limits of Silicon Power Devices
To further explore the potential of silicon based power devices in gen-
eral, and IGBTs/diodes in particular, the performance/design limits
must be clearly defined in order to determine a future development
trend. Today, by carrying out a standard performance study of state-
of-the-art high power IGBT modules, one can clearly see that the
main limiting factor in terms of maximum output current capability (i.e.
power density) is the diode in both inverter and rectifier mode opera-
tion. In addition, the diode also presents another major restriction
with regard to its surge current capability. Both limits are clearly a
result of the limited diode area available in a given package footprint
design which has a typical IGBT to diode area ratio of around 2:1.
This limit in diode current capability was fundamentally established
after the introduction of modern low-loss IGBT designs. As mentioned
earlier, the simple solution of increasing the diode area is not a pre-
ferred one and in any case remains restricted by the package stan-
16 Bodo´s Power Systems® July 2009 www.bodospower.com
C O V E R S T O R Y
The Bi-mode Insulated GateTransistor (BIGT)
A High Voltage Switch for Next Generation Megawatt Applications
An advanced high voltage reverse conducting IGBT concept referred to as the Bi-modeInsulated Gate Transistor (BIGT) is currently being developed. The BIGT can operate atthe same current densities in both IGBT and diode modes by utilizing the same available
silicon volume as it targets to fully replace the traditional IGBT/Diode two-chip approachwith a BIGT single chip.
By Munaf Rahimo, Arnost Kopta and Ulrich Schlapbach, ABB Switzerland Ltd, Semiconductors
17www.bodospower.com July 2009 Bodo´s Power Systems®
dard footprint design. This leads to the conclusion that the develop-
ment effort must target an improved diode performance to match at
least the current IGBT designs. In other words, there is currently no
need for improved switch generations unless the diode experiences a
major revolution in terms of reduced losses and thus, higher power
capabilities. As discussed above, while ignoring cost and material
issues, WBG based diodes could provide the required performance
due to the low switching losses, but the high conduction losses of
high voltage WBG unipolar and bipolar diodes compared to current
silicon diodes will restrict their application to relatively high frequen-
cies for Megawatt applications. Other aspects to consider are the soft
reverse recovery performance of WBG diodes in general under high
currents and high voltages which have not yet been thoroughly evalu-
ated.
The above analysis leads to the following conclusion; in order to
increase the power density of high voltage IGBT modules while also
solving the real limiting factors due to the diode performance, an
IGBT and diode integration solution is needed, or what has been nor-
mally referred to as a Reverse Conducting RC-IGBT. The practical
realization of a single-chip technology will provide an ideal solution
for next generation high voltage applications demanding compact
systems with higher power levels, which is proving to be beyond the
capability of the standard two-chip approach.
The Bi-Mode Insulated Gate Transistor (BIGT)
Similar to power MOSFETs, the traditional goal for a reverse con-
ducting device having an integral diode is to obtain higher power for
a given footprint package area by eliminating the need for a separate
anti-parallel diode. This approach has been demonstrated experi-
mentally in recent years for medium voltage IGBTs (600V-1200V)
mainly operating at low currents and/or soft switching conditions in
special applications. On the other hand, we at ABB have been inves-
tigating the RC-IGBT concept by demonstrating its feasibility for high
voltage chips under heavy paralleling. The realization of the RC-IGBT
concept has always faced a difficult challenge due to a number of
technological design and process barriers such as (a) the conflicting
requirement for plasma enhancement for the IGBT and diode (b)
matching the silicon and buffer design parameters for both the IGBT
and diode with regard to device softness (c) the inherited on-state
snap-back phenomenon associated with RC-IGBTs and finally (d) the
layout and alignment design of the shorted collector for minimizing
non-uniform charge distributions during device operation.
Further development work to resolve the above issues and improve
the device characteristics has resulted in a clear breakthrough in per-
formance by adopting an advanced collector backside P/N layout
design, fine doping profiles and controlled lifetime reduction for
enabling optimal operation in both IGBT mode and diode mode. The
new device concept is referred to as the Bi-mode Insulated Gate
Transistor (BIGT). The results obtained show that the BIGT exhibits
C O V E R S T O R Y
Figure 1: The BIGT basic structure
Power Your Recognition InstantlyBased in Munich, Germany, ITPR Information-Travels Public Relations is a full-service consultancy
with over a decade of experience in the electronics sector.
As a small exclusive agency, we offer extremely high ROI,
no-nonsense flexibility and highest priority to only a handful of companies.
Strategical SupportCorporate/Product Positioning, Market/Competitive Analysis, PR Programs, Roadmaps,
Media Training, Business Development, Partnerships, Channel Marketing, Online Marketing
Tactical PRWriting: Press Releases, Feature Articles, Commentaries, Case Studies, White Papers
Organizing: Media Briefings, Road Shows, Product Placements in Reviews and Market Overviews,
Exhibitions, Press Conferences
Monitoring and Research: Speaking Opportunities, Editorial Calendars, Feature Placement,
Media Coverage, Competitive Analysis
Translations: Releases, By-Lined Articles, Websites, etc.
Call or contact us today for a free consultation on how PR
can dramatically affect your company’s bottom line.
ITPR Information-Travels Public Relations
Stefanusstrasse 6a, 82166 Gräfelfing-Munich, Germany
Tel ++49 (89) 898687-20, Fax ++49 (89) 898687-21,
w w w . i n f o r m a t i o n - t r a v e l s . c o m
low losses in both modes of operation with no typical snap-back
behavior in the transistor on-state mode when compared to a stan-
dard RC-IGBT, while also maintaining high levels of SOA perform-
ance. The BIGT offers in addition a number of device performance
advantages as described below. The BIGT technology consists of a
hybrid structure integrating an IGBT and an RC-IGBT into a single
chip as illustrated in figure (1). A circuit symbol is also proposed for
the BIGT.
The main target of this combination is to eliminate snap-back behav-
ior at low temperatures in the BIGT transistor on-state mode by
ensuring that hole injection occurs at low voltages and currents from
the P+ collector region in the IGBT section of the BIGT. The BIGT
provides an optimum solution especially for thin devices with SPT
type buffer designs where the snap-back phenomenon is pronounced
in RC-IGBTs. The backside layout design and dimensioning provides
smooth transition into full chip IGBT conduction while maximizing the
RC-IGBT area for diode conduction. Therefore, the BIGT concept has
resulted in a better trade-off between the above mentioned parame-
ters compared to the standard RC-IGBT design.
To optimize the BIGT for low dynamic and switching losses, the other
main challenge was to enable low diode mode recovery losses while
not having a considerable effect on the transistor mode on-state loss-
es. A three step approach is utilized to achieve this target. The first
step is the fine control of the doping profiles of the emitter p-well cells
and collector P+/N+ regions. As shown in figure (1:top), the
Enhanced Planar (EP) cell design does not include any highly doped
P+ well regions and provides the BIGT with a fine pattern p-well pro-
file for obtaining low injection efficiency for better diode performance
while maintaining the typical low IGBT losses associated with EP
designs. The second optimization step employs a Local p-well Life-
time (LpL) control technique utilizing a well-defined particle implanta-
tion which further reduces the diode recovery losses without degrad-
ing the transistor losses and blocking characteristics. Further reduc-
tion in the reverse recovery losses is achieved with a uniform local
lifetime control employing proton irradiation. A further optional 10%
reduction of in recovery losses can be obtained with a MOS Control
Diode function as demonstrated in the module results presented in
the following sections.
The BIGT technology has mainly been developed for high voltage
devices and the work presented here was carried out on a
3300V/62.5A BIGT (active area = 1cm2). The on-state characteristics
of the BIGT in transistor and diode modes are shown in figure (2) at
25°C and 125°C. For safe paralleling of chips, the curves show a
strong positive temperature coefficient even at very low currents in
both modes of operation due to the optimum emitter injection efficien-
cy and lifetime control employed in the BIGT structure.
A remarkable performance feature of the BIGT is that it provides soft
turn-off behavior under all operating conditions in both transistor and
diode modes. The optimized collector P+ doping profiles will ensure
that during the turn-off tail in both modes, the passing electrons
towards the N+ regions will induce a large potential across the collec-
tor PN junction forcing a controlled level of hole injection into the
base region. The main advantage of this method lies in the fact that
normally the diode silicon specification does not match the IGBT sili-
con for obtaining soft recovery performance. Thus, such conflicting
requirements could result in diode mode snappy behavior in an inte-
grated structure. However, in a BIGT, soft behavior is granted under
all conditions for both the transistor and diode modes even under
extreme conditions as shown in figure (3) and (4) respectively.
3300V BIGT Module Results
High current 3.3kV BIGT (140 x 130)mm modules were fabricated
and tested under conditions similar to those applied to state-of-the-art
IGBT modules. The BIGT module contained 24 BIGT chips for the
estimated current rating of 1500A. The nominal transistor mode
switching characteristics of the BIGT modules are shown in figure (5)
along with the associated switching losses at 125°C. The BIGT diode
mode reverse recovery performance is mirrored in the turn-on wave-
forms. The freewheeling reverse recovery losses were approximately
2.3J. The SOA performance of the BIGT in transistor and diode
mode at a high DC link voltage is shown in figure (6). Both modes
show rugged characteristics similar to the current IGBT and diode
modules. Furthermore, the BIGT also provides improved short circuit
and softness performance compared to state-of-the-art IGBTs.
The BIGT advantage is clearly demonstrated here since this module
can practically replace a similarly rated larger (140 x 190)mm module
which normally contains 24 IGBTs and 12 diodes. The larger stan-
dard IGBT module has the further disadvantage of employing much
C O V E R S T O R Y
18 Bodo´s Power Systems® July 2009 www.bodospower.com
Figure 3: 3.3kV/62.5A BIGT in transistor mode turn-off softness test
Figure 4: 3.3kV/62.5A BIGT in diode mode reverse recovery softnesstest
Figure 2: 3.3kV/62.5A BIGT on-state in transistor and diode mode
19www.bodospower.com July 2009 Bodo´s Power Systems®
less diode area which is normally a limiting factor in rectifier mode of
operation and surge current capability. On the other hand, when the
larger (140x190)mm module employs only BIGT chips i.e. 36
devices, its rating can potentially reach up to 2250A. Figure (7)
shows the targeted scaled output current performance for the BIGT
compared to today`s EP-IGBT module at 125°C in both Inverter and
Rectifier modes. The curves show that the diode performance is a
limiting factor for the standard module approach while for the BIGT
module the transistor mode defines the limit. The curves show
approximately a 30% increase in output current capability up to 2 kHz
with the BIGT technology.
Finally, the expected thermal cycling load pattern in a BIGT module is
shown in figure (8). Due to the fact that no inactive periods are pres-
ent per IGBT/Diode compared to the standard approach, a lower
temperature difference and more efficient thermal utilization of the
module will result in better thermal cycling capability and eventually
improved reliability performance.
www.abb.com/semiconductors
C O V E R S T O R Y
Figure 5: 3.3kV/1500A BIGT (140x130)mm module nominal turn-on(left) and turn-off (right) and associated losses. (I: 500A/div, V:500V/div, Vge: 10V/div, Time(x): 1usec/div)
Figure 6: 3.3kV/1500A BIGT (140x130)mm module SOA transistormode turn-off (left) and diode mode recovery (right) and associatedpeak-power. (I: 500A/div, V: 500V/div, Time(x): 1usec/div)
Figure 7: 3.3kV (140x190)mm BIGT module performance chart.
Figure 8: Predicted BIGT thermal load cycling in PWM application
www.we-online.com
E M C C O M P O N E N T S
I N D U C T O R S
T R A N S F O R M E R S
R F C O M P O N E N T S
P O W E R E L E M E N T S
C O N N E C T O R S
C I R C U I T P R O T E C T I O N
A S S E M B LY T E C H N I Q U E
Eco TransformerTransformer for energy saving electronic devices
Universal input voltage: 85-265 VAC
Samples free of charge
4kV isolation voltage
Guaranteed in stock
High efficiency
Reference design of all major IC manufacturers
To reduce energy consumption in the enter-
prise, it’s critical to focus on all aspects of
energy: usage, allocation and management.
With advent of the new PoE Plus (IEEE
802.3at) compatible PSEs, there is the
added control functionality that allows for
dynamic allocation of per-port Power Device
(PD) power by the PSE. Using this
approach, power can now be allocated and
managed much as data bandwidth is today.
A new PoE System-on-a-Chip (SoC) that
integrates an isolation barrier has been
specifically developed to affect significant
energy savings in the enterprise. The
device’s integrated isolation increases the
power-conversion efficiency of individual
PDs in a simple and cost effective way, while
also enabling network software to reduce
enterprise energy consumption with better
allocation and management of power. The
overall energy saved by using this device
can offer considerable benefits in an enter-
prise setting where thousands of IP phones,
WAPs, thin-client PCs and security cameras
all draw power from more efficiently-loaded
power supplies at the network center.
PD Efficiency Improvements
The cumulative effect
of efficiency improve-
ments at each of the
PD nodes can be
quite substantial
across the enterprise.
For example, if on
average a PoE Class 2 IP phone draws an
average power of 5.2W from the PSE,
improving the PD power conversion efficien-
cy by 5% (example from 86% to 91%) will
provide over 500mW of savings at the PD,
and approximately 700mW of savings
referred back to the PSE input. Similarly a
20W IP camera can save 1.3W at the PD
and over 1.6W referred to PSE input. Table
1 below shows savings for different types of
appliance, assuming average power con-
sumption based on the device PoE Class.
In a simple model of 1,000-employee enter-
prise that consists of 1100 IP phones, 60
wireless access point nodes, and 40 IP cam-
eras, power savings at each PD node due to
efficiency improvements will lead to over 860
Watts of power saving, and over 16.6kWH of
energy saving per year when reflected to the
PSE input, including savings from less heat
dissipation in the data-center.
Intelligent Power Allocation
Allocation of PSE power in a traditional PoE
environment consists of setting aside the
maximum power that each port will be
expected to deliver to a PD at any phase of
its operation. In addition, maximum cable
power losses must also be set aside. Gener-
ally there is a significant margin between the
maximum advertised power and the actual
power consumed. These maximums force
the significant over-design of PSE power
supplies. In addition, power supplies are
much more efficient when used near their
peak capacity, rather than under light load-
ing.
For example, a traditional PSE will budget
7.7W of power for a Class 2 IP phone
(6.95W for PD, 0.71W for 100m cable loss),
and 30W of power for a Class 4 IP camera
(25.5W for PD, 4.5W for cable loss). If the
PSE could determine a PD’s dynamic power
consumption and the cable distance
between the PSE and the PD, the PSE
could substantially optimize its power alloca-
tion and hence reduce total required power
at the PSE. Assuming average power drawn
by the PDs, as in Table 1, and average Eth-
ernet cable length of 50m between the PSE
and the PD, intelligent allocation will reduce
the power budget at the PSE by over 35%
(7.3kW vs. 11.2kW for static allocation) in
our 1,000-employee enterprise model.
P O W E R M A N A G E M E N T
20 Bodo´s Power Systems® July 2009 www.bodospower.com
Integrated Digital Isolation Delivers Enterprise Energy
Efficiency BenefitsThe cumulative effect of efficiency improvements can be quite substantial
Power over Ethernet (PoE) standards have made it possible to deliver power to networkdevices along the same Ethernet cable used for data, eliminating the need to provide a‘wall wart’ power cube for each device. Network power is thereby consolidated at thePower Sourcing Equipment (PSE) in the data-center, which has tremendous benefits of
lowering cost, simplifying infrastructure and providing backup power management. Overall power management has been crudely addressed in previous-generation PoE
clients via power classification, but these have largely been limited and static solutions.
By Sajol Ghoshal, Chief Technical Officer and Chief Architect, Akros Silicon
Table 1 – Average Power Savings per PD Node
PD Efficiency Improvement (91% vs. 86%) Used Power (W) Savings at PD (W) Savings at PSE A/C Main (W)
IP Phone (average usage) 5.2 0.54 0.69
WAP (13W) 10.7 0.68 0.88
IP Camera / IP PTZ Camera (13W / 25W) 10.7 / 20.0 0.68 / 1.28 0.88 / 1.64
This not only reduces the power supply size
needed at the PSE, delivering direct cost
savings, but it also allows the PSE power
supplies to operate near full load conditions
for higher efficiency, reducing power
wastage and conversion to heat in the data-
center. Energy savings for the 1,000-employ-
ee enterprise example can be over
11.8kWH, including the savings from
reduced cooling requirements.
Enterprise Energy Efficiency Benefits
Using the 1,000-employee model, , if each
PD saves an average of 0.5-1.3Watts, the
total savings for the enterprise would be
nearly 16.6kWh or €3,500 annually,
assuming average European com-
mercial energy rates. With typical-
ly-rising energy costs, this factors
into a 5-year savings of over
€20,000. The total savings
approaches €35,000 over five
years (Table 2 below) after adding
in savings effect of intelligent
power allocation between the PSE
and PDs. Considering a five-year
average life span of a PD, this
translates to close to €29 in recur-
ring cost savings per PD node, a
substantial direct monetary benefit.
It also helps enterprises achieve
their energy efficiency objectives.
PoE Device with Integrated Digi-
tal
Isolation
A new device by Akros Silicon
called the AS1854 integrates Akros
GreenEdge™ 2kV digital isolation
technology with next-generation
PoE PD and power conversion
technology. This combination deliv-
ers groundbreaking power integra-
tion, and enables a new range of
Digital Power PoE PD capabilities
and solutions. A Type 1 (IEEE®
802.3af) and Type 2 (IEEE® pre-
802.3at) compliant PD is integrated
into the device with high-voltage
isolation and quad-output digital
power DC-DC converters – result-
ing in a complete PoE and power
management solution in a single
device with minimal external com-
ponents.
The number of required voltage
sources in a PoE-PD platform is
similar to many embedded plat-
forms that include a processor,
DRAM/SRAM, Flash memory and
I/O, each with different voltage and
power requirements. For a VoIP
21www.bodospower.com July 2009 Bodo´s Power Systems®
P O W E R M A N A G E M E N T
Table 2 – 1000-Employee Enterprise Model Energy Saving Benefits
Energy savings from PD node efficiency,
including cooling
16,626 kWh
Yearly savings from PD node efficiency €3,500
Energy savings from Live Monitoring 11,836 kWh
Yearly savings from Live Monitoring €2,500
Total yearly savings
(PD efficiency and Live Monitoring)
€6,000
Total Lifetime Savings (5 Yr) €34,600 (€28.84 per node)
deadline for submission, July 17, 2009, go to web for details:
www.apec-conf.orgwww.apec-conf.org
CALL FOR PAPERS!CALL FOR PAPERS!
SPONSORED BY
2010February 21–25, 2010
Palm Springs Convention Center,Palm Springs, CA
THE PREMIER
GLOBAL EVENT
IN POWER
ELECTRONICSTM
THE PREMIER
GLOBAL EVENT
IN POWER
ELECTRONICSTM
22 Bodo´s Power Systems® July 2009 www.bodospower.com
phone, there is often a display that may
demand a unique voltage, whereas a pan-tilt-
zoom (PTZ) camera will have multiple
motors as well as low-voltage electronic
power needs. By integrating four flexible dc-
dc outputs in one power SoC, including digi-
tal isolation, the AS1854 simplifies PoE PD
system design and delivers the benefits of
power-system efficiency and intelligent man-
agement (Figure 1)
Integrated digital isolation offers a break-
through for improving power system efficien-
cy and dramatic cost savings. Using a built-
in isolated synchronous driver for secondary
side rectification, the AS1854 not only elimi-
nates the need for external pulse transform-
ers to create the sync drive, it allows the
device to intelligently manage timing of the pri-
mary and secondary drives. This optimizes “Sync delay” and “Over-
lap delay” parameters, practically eliminating PWM timing related
energy losses (Figure 2). Figure 3 shows the improvements in PoE
power system efficiency delivered by the AS18x4 series products –
which improves both light-load and full-load efficiency compared to
traditional designs. The GreenEdge technology delivers efficiency
improvements of over 8% in PoE Class 2 range, the one most com-
monly used for VoIP phones, over typical traditional designs.
Additionally, the device’s integrated isolation technology enables
direct digital management of both isolated primary power and sec-
ondary system power. Primary-side isolated ADC in the AS1854
device allows for direct power usage monitoring by the PD’s μCon-
troller across the isolation barrier. This “Live Monitoring” feature col-
lects real-time power usage to pass to the PSE’s Network Manage-
ment software via standards based LLDP messaging. This enables
the PSEs to deduce cable power loss and true PSE power allocation
needed per PD port, eliminating blind allocation of maximum power
to each PD.
This approach for the first time enables PSEs to serve the same
number of PDs with a much lower power and efficiently operating
power supply. In addition, savings can be augmented upstream using
a live monitoring capability, which allows for better power allocation
and higher-efficiency PSE power supply utilization. Although isolated
communication channels can be accomplished using discrete compo-
nents such as opto-couplers, this approach tends to be slow, bulky
and expensive, making the design more costly, much larger and
sometimes even impractical.
Summary
PD efficiency improvements and intelligent allocation capabilities can
deliver significant monetary benefits to the enterprise with lower ener-
gy usage and lower cost of power supplies in the data-center. A new
device using Akros’ GreenEdge™ integrated isolation facilitates ener-
gy-saving design improvements on the PD, reducing device power
consumption with efficient power conversion across a wide loading
range. This new approach to integrated high-voltage power manage-
ment provides a unique opportunity for cost-effective, end-to-end
“green power” applications. By employing the AS18x4 product family,
designers now can create value-added and differentiated feature sets
in a cost-effective manner, passing on recurring energy-saving bene-
fits to their end customers.
www.AkrosSilicon.com
Figure 2: Primary and Secondary PWM timing management acrossintegrated isolation
Figure 3 – GreenEdge™ Efficiency Improvement
P O W E R M A N A G E M E N T
Figure 1: Akros Silicon’s AS1854 in PoE Powered Device System
Efficiency goes hand-in-hand with the electric/hybrid vehicle revolu-
tion and designers now need to squeeze every last ounce of efficien-
cy from each part of the vehicle. It has often been the case that the
power-supply design engineer was the last person to be brought onto
a project: a topology already dictated; available space limited. How-
ever, with efficiency now at the top of the agenda, automotive drive-
train design engineers are embracing energy saving technology, and
planar transformers are at the forefront of the revolution.
Planar transformers often offer an electrical efficiency of up to 99.5%
and provide higher power density than their conventional counter-
parts. In addition to this, not only are planar transformers consider-
ably smaller than their conventional counterparts, they also offer a
significant weight advantage: the lighter the moving object, the less
energy required to move it. Hence, when you consider the added
efficiency that the reduced weight of a planar transformer provides to
an electric/hybrid vehicle, together with the improved electrical effi-
ciency, it is simple to understand why the demand for planar trans-
formers is growing rapidly.
Planar transformers are ideal for all switch-mode power supply appli-
cations and the choice of magnetic components for your design is as
important as selecting the best topologies and switching frequencies.
Electric/hybrid vehicle battery charging designs often require a power
rating in excess of 2kW, at relatively high current levels. These appli-
cations ideally lend themselves to a copper ‘lead-frame’ planar trans-
former construction, comprising thin, flat, pre-formed copper layers.
Himag Solutions is a world leader in the high-frequency 2kW+ power
range, and has now successfully manufactured single planar trans-
former components providing up to 30kW of power throughput.
Having established the weight advantage provided by a planar trans-
former, what is it that makes them so much more electrically efficient
than their conventional ‘wire-wound’ equivalents? There are a few
elements, one being the ‘construction’ method of a planar trans-
former: the pre-formed copper lead-frames (or PCB layers) are easily
interleaved between one another, primary, secondary, primary, etc...
up to five or six times is not uncommon, and are accurately posi-
tioned in-line with one another in the vertical dimension. This inter-
leaving reduces the losses: with particularly improved coupling
between windings, a significant reduction in leakage inductance and
reduced skin effect. Furthermore, by managing the interleaving with-
in the design, the leakage can be very accurately controlled to meet
the customer’s design specification.
T R A N S F O R M E R S
24 Bodo´s Power Systems® July 2009 www.bodospower.com
Planar Transformers are Essential for Truly Efficient
Electric/Hybrid VehiclesWeight advantage makes them more efficient than
conventional ‘wire-wound’ equivalentsWhen the first mass-produced motor car, the ‘Model T’, rolled off the production line in
1908, Gerald Ford famously told the world that “you can have it in any colour, as long asit’s black”. Today, a century later, the motor car has become the backbone of our
developing world, and it now appears that we can still only have one colour, although‘GREEN’ is the new black. “Necessity is the mother of invention” said Plato, and after
years of waiting it has taken a global economic downturn for the definition of ‘necessity’ tobecome: ‘replace the petrol engine with something more efficient and less polluting’. Final-ly some political and financial backing, from the likes of Obama and Brown, has turned the
automotive world’s attention away from ‘gas-guzzling’ to focus more on ‘energy-saving’
By Dean Curran, Managing Director at Himag Solutions Ltd., United Kingdom
Figure 1: E64 1kW-9kW Planar Transformer
Invariably planar transformers are custom designed (or ‘bespoke’ as
we say in the UK) for the specific end-user application. This means
that instead of the power supply design engineer having to design his
system around limited standard ‘off-the-shelf’ transformers, he can
design in the knowledge that the planar transformer will do exactly
what he needs it to do. Needless to say, the very nature of the cus-
tom designed planar transformer means that yet more efficiency is
achieved: it is obvious that if you have to ‘make do’ with a standard
part, your system will not run as efficiently as a custom designed part
specifically matching your needs.
The advantages offered by the planar winding construction are also
opening the market for ‘retro-fit’ planar transformers. Himag Solu-
tions is now frequently asked to design a planar transformer to
replace a conventional transformer that runs too hot: the planar
equivalent not only runs cooler but also improves the inverter drive
waveforms, thus reducing switching losses in the power converter as
well. We have successfully applied planar transformer design tech-
niques to conventional ferrite cores to meet the ‘retro-fit’ space con-
straints, whilst achieving an overall more efficient, cooler, component.
Beyond the electrical efficiency of planar transformer design, advan-
tages are also found in heat dissipation. Planar transformers use
‘flat’ ferrite cores that are very low profile but offer a higher
surface:volume ratio than their conventional counterparts, providing
significantly improved cooling. Furthermore the particular open con-
struction method used by Himag Solutions is very flexible and pro-
vides an excellent thermal path from the planar structure to the main
heatsink via the ferrite cores. In addition to this, it is also very easy
to incorporate ‘spacers’
between the windings to allow
airflow through the centre of
the transformer, improving
thermal performance yet fur-
ther, especially when forced air
cooling is available.
The final thing to mention is
the dreaded ‘C’ word: COST!
For some time now planar
transformers have been disre-
garded as being too expen-
sive. Contrary to popular opin-
ion, the costs have significantly
reduced in recent years
through cheaper material
costs, as the market grows,
and improved manufacturing
techniques. It is common for a
5kW planar transformer to cost
in the region of $23 nowadays,
and in volumes of 100k+
prices below $10 per piece are
easily achievable. Moreover,
in the 2kW-30kW power range,
the use of pre-formed copper
‘lead-frames’ in single and
multi-turn forms can achieve
virtually any desired turns ratio
at low cost, allowing rapid
design iteration and ultimately
improved time to market.
It is extremely exciting to be leading a planar transformer company
through the efficiency revolution that we are embarking upon. With
hybrid/electric vehicles at the forefront, we are also experiencing
huge demand from the defence/military, solar/wind power, medical
and induction heating/charging sectors. The planar transformer mar-
ket is set for significant growth over the next few years and one thing
is certain, if you’re not using or considering using a PLANAR
TRANSFORMER, then your competitors surely are.
www.himag.co.uk
25www.bodospower.com July 2009 Bodo´s Power Systems®
Figure 2: E102 7kW-30kW Planar Tramsformer
Figure 3: E102 7kW-30kW Planar Transformer Dimension Drawing
Advantages of Planar Transformers
• High Efficiency
• Low Weight
• Low Profile
• Low Leakage Inductance
• Highly Efficient Cooling
• Multiple Topologies
• Multiple Winding Options
• Customised Terminations
• Dimensional Accuracy
• Tightly Controlled Parasitics
BICKER ELEKTRONIK GmbHPhone +49 (0)906 / 70595-0
www.bicker.de
Point Of SaleBEA-740 | 450 Watt• Additional +24 VDC output
for printers, displays, scanners,cash terminals etc.
26 Bodo´s Power Systems® July 2009 www.bodospower.com
The demands for high operating efficiency and small size in con-
sumer electronic products such as home theatre systems, game con-
soles and LCD televisions is moving power supply design towards
resonant topologies capable of operating at high switching frequen-
cies. At the same time as allowing high-frequency operation permit-
ting the use of smaller magnetic components, the resonant convert-
er’s soft commutation enables the SMPS to operate efficiently and
with low EMI.
Refining the Resonant Converter
Among the various resonant converter topologies, the LLC converter
has become the switching scheme of choice. Although as easy to
build as a basic series-resonant LC converter, it overcomes draw-
backs such as difficulty in maintaining regulation at light load. It also
effectively improves efficiency when the input voltage is high such
that switching loss is more dominant than conduction loss. The LLC
resonant topology utilises an additional shunt inductor across the pri-
mary winding of the transformer, as shown in Figure 1. This is usually
realised using the magnetising inductance of the transformer, which
is controlled by adjusting the transformer air gap. This topology pro-
duces a complex resonant tank with buck-boost transfer characteris-
tics in the soft-switching region.
Referring to figure 1, in normal operation the primary-side MOSFETs
operate at 50% duty cycle and the output voltage is regulated by
varying the switching frequency of the converter. The converter has
two resonant frequencies; a lower resonant frequency (given by Lm,
Lr, Cr and the load), and a fixed higher series resonant frequency Fr1
(given by Lr and Cr only). The secondary side half bridge can be
soft-switched for the entire load range by operating the converter
either above or below Fr1.
The secondary side half-bridge is conventionally implemented using
a pair of diodes. However, this arrangement is relatively inefficient, as
diode losses contribute significantly to the overall power loss of the
SMPS. With increasing current draw in future generations of feature-
rich consumer products, these losses will continue to increase, since
the diode rectifier conduction loss is proportional to the product of its
forward conduction current as well as the forward voltage drop.
Increasing dissipation also demands the use of larger diodes, leading
to progressively more bulky power supplies.
Hence there are two powerful factors forcing designers to demand a
more satisfactory secondary side topology for LLC resonant convert-
ers. For power supplies in the region of 50W and higher, the demand
for higher power density to minimise case dimensions is the domi-
nant concern. In the region of 200-400W, boosting efficiency to satis-
fy initiatives such as Energy Star and CEC 80+ is a major reason for
power supply designers to seek ways to eliminate the losses incurred
in the secondary side diodes.
Synchronous Secondary Side Rectification
Using synchronous rectification in the secondary side holds out the
promise of reducing the large losses incurred in the half-bridge
diodes. Since MOSFET conduction losses depend on I2 x RDS(ON),
splitting current between two synchronous MOSFETs reduces the
dissipation in each device by four, thereby halving the total dissipa-
tion.
However, the most familiar control techniques for synchronous recti-
fiers are not workable in LLC resonant converters. For example in a
primary-controlled synchronous rectifier, where the MOSFET control
signals are derived from primary side signals, the LLC resonant con-
verter has a phase lag between the input voltage of the resonant tank
and the rectified secondary side current. This prevents the primary
side gates from being used to drive the secondary side rectifiers. The
alternative self-controlled rectifier technique, where the control sig-
nals are derived from the secondary voltage across the power trans-
former, suffers from timing mismatches between the rectified second-
ary side current and the voltage across the main power transformer,
which is a 50% duty cycle square wave. These prevent satisfactory
operation below the resonant tank frequency of the converter.
Using a current transformer, on the other hand, is a workable control
technique for resonant converters. The drawbacks of this technique
include a high component count, which increases footprint and
impairs reliability, and the need for a relatively expensive fast com-
parator.
Placing a controlling IC in the secondary side to manage the switch-
ing of the MOSFETs potentially offers a simpler and more cost-effec-
P O W E R S U P P LY
Smarter Rectification Driving up SMPS Efficiency and Power Density
Performance, size and efficiency advantages have made the LLC series-resonant converter the preferred power supply topology in applications such as high-end consumerproducts. Going forward, improvements to the secondary side architecture are necessary
to further enhance efficiency and space savings.
By Helen Ding, International Rectifier
Figure 1: Half-Bridge LLC series-resonant converter
27www.bodospower.com July 2009 Bodo´s Power Systems®www.bodospower.com July 2009 Bodo´s Power Systems®www.bodospower.com July 2009 Bodo´s Power Systems®
tive alternative. By eliminating the current
transformer and fast comparator, an IC-
based solution saves size and component
count. However, to implement all the
required functions in a single component
requires certain competencies, such as the
ability to co-integrate the control functions
with high-voltage sensing capabilities in the
same device, as well as managing high
switching frequency and high current-driving
capabilities. International Rectifier’s IR1168,
for example, uses high-voltage IC (HVIC)
technology and patented techniques to deliv-
er a secondary side rectifier driver IC
designed to drive two N-Channel power
MOSFETs used as synchronous rectifiers in
resonant converter applications. In addition
to providing two gate drivers, the device also
provides adaptive shoot-through protection
to prevent both channels from simultaneous-
ly turning on. It is also capable of operating
in normal and burst-mode conditions. In
addition, clamped gate-driver operation sig-
nificantly reduces power dissipation.
Single-Chip Control
Figure 2 illustrates a typical application
schematic for the IR1168. In normal operat-
ing mode, the IC senses the voltage drop
across each MOSFET at pins VS1/VD1 and
VS2/VD2, and turns each MOSFET on and
off via pins GATE1 and GATE2.
At the core of this device are two high-volt-
age (200V) high-speed comparators. These
differentially sense the drain to source volt-
age of the MOSFET, using the RDS(ON) of the
device as a shunt resistance, and hence
determine the polarity and level of the device
currents. Dedicated internal logic then man-
ages the turning on and off of each device in
close proximity to the zero-current transition.
The device uses the SmartRectifier™ control
technique, which compares the sensed volt-
age across the MOSFET with two negative
thresholds to determine the turn-on and turn-
off transitions for the device. The most nega-
tive of these two thresholds, VTH2, detects
current through the body diode and hence,
controls the turn on transition for the power
device. Similarly, a second negative thresh-
old, VTH1, determines the level of the current
at which the device turns off. A third thresh-
old voltage, VTH3, acts as a reset threshold
governing the resetting of an internal one-
shot when the cycle is completed and the
VDS voltage turns positive and starts to
increase again. This way the system is ready
for next conduction cycle.
By governing the drive level of the second-
ary side MOSFETs according to these three
thresholds, the IR1168 ensures accurate
performance without the need of a PLL or
external timing sources. In fact, the high
accuracy of turn-off transitions is a key bene-
fit of this technique, since it prevents reverse
current across the MOSFETs and also min-
imises the body diode conduction time. Addi-
tionally, internal blanking logic is used to pre-
vent spurious gate transitions and guarantee
operation in fixed- and variable-frequency
operation modes.
The waveforms of Fig-
ure 3 show the IR1168
in normal operating
mode. When the con-
duction phase of the
MOSFET is initiated the
current begins to flow
through the MOSFET
body diode, generating a
negative VDS voltage.
Since the body diode
has a much higher volt-
age drop than that
caused by the MOSFET
RDS(on), this negative VDS triggers the turn-
on threshold, VTH2. At this point, the IR1168
will turn on the gate of the MOSFET. This, in
turn, causes the VDS voltage to drop down to
a value defined by the MOSFET drain cur-
rent (ID) multiplied by RDS(on).
Since this fall in voltage is usually accompa-
nied by some amount of ringing that could
trigger the input comparator to turn-off, a
fixed Minimum On Time (MOT) blanking
period is used that will maintain the power
MOSFET on for a minimum time duration.
Once the MOSFET has been turned on, it
remains on until the rectifier current decays
to the level where the voltage will cross the
fixed turn-off threshold VTH1. Once the
threshold is crossed, the current will start
flowing again through the body diode, caus-
ing the VDS voltage to jump negative again.
Depending on the amount of residual cur-
rent, VDS may once again trigger the turn-on
threshold. Hence, VTH2 is blanked for a time
duration called TBLANK after VTH1 is trig-
gered. The period TBLANK is shown in the
diagram, and is terminated when the device
VDS crosses the positive reset threshold
VTH3. The IC is then ready for the next con-
duction cycle.
Conclusion: Performance Advantages in
Next-Generation Products
Figure 4 compares a 240W multi-rail power
supply for an LCD TV application built using
the IR1168 synchronous-rectifier controller
against a conventional resonant converter
design with secondary side Schottky diodes.
The IC is housed in a low-cost SO-8 pack-
age and delivers a single-chip solution capa-
ble of generating the control signals for both
MOSFETs. The four Schottky diodes were
replaced by IR1168 with four SO-8 MOS-
FETs daughter card. The two large heatsinks
cooling the Schottky diodes for the 12V and
24V rails are also eliminated.
In practice, this smaller SMPS featuring sec-
ondary side synchronous rectification has
shown a 1.5% increase in efficiency. A 25°C
reduction in operating temperature for the
SMPS has also been recorded, leading to a
significant increase in overall system reliabil-
ity.
www.irf.com
www.bodospower.com July 2009 Bodo´s Power Systems®
P O W E R S U P P LY
Figure 2: Typical secondary-side applicationcircuit for IR1168
Figure 3: Operation of SmartRectifier™ syn-chronous-rectification control with IR1168
Figure 4: IR1168 retrofit in a 240W LCD TV
Selecting precision voltage references can be an arduous task.
Today, there are hundreds of products offered by dozens of chip
companies, offering various combinations of initial accuracy, tempera-
ture coefficients and cost. Selecting the right part for the right applica-
tion requires considerable research into the details of the datasheets
as well as sourcing volume pricing quotations.
While several parameters are important in the selection of voltage
references, the two that stand out as potentially contributing the
greatest errors are initial accuracy and temperature coefficient (TC).
Noise, thermal hysteresis, line and load regulation and long term sta-
bility should not be neglected when making a selection, but their con-
tributions to error are shadowed by initial accuracy and TC. It is not
surprising therefore, that chip manufactures concentrate their market-
ing efforts to proclaim best in class performance achievements in
these two domains.
Initial Error
Initial error is the deviation of the actual output voltage from the
desired specification. Many voltage references have trim pins that
allow the user to set or pre adjust the output value in an attempt to
offset the inherent initial error of the reference chip. (See Fig. 1) The
output is adjusted by setting the ratio of resistance above and below
the “wiper” shown in the figure.
Of course the ability to precisely select the desired output voltage
then becomes an issue of finding the correct resistance values, tak-
ing into consideration their initial accuracies (or perhaps more cor-
rectly, their inaccuracies) as well as their contribution to the TC of the
output voltage. The specific ratios attainable are constrained by the
limited selection of fixed resistance values.
In this application we have set about to show how to improve the per-
formance of an average precision voltage reference, the Analog
Devices ADR 425A, by more than an order of magnitude.
Analysis of the circuit requirements showed that by replacing the
470KÙ ballast resistor with a 120KÙ, resistor, and using the
MBT143E 1:9 ratio Rejustor, the Vout could be adjusted quite pre-
cisely. See Fig 2
To assure the best possible performance, the trimming is accom-
plished at the assembled board level. During this calibration process,
software algorithms provided by Microbridge automatically adjust the
two resistances of the MBT143E down from their as manufactured
values based on the real time feedback from Vout, taking about 1-2
seconds, until the desired output accuracy is achieved. The rejustor
values will be trimmed to slightly different resistances for each
ADR425A to compensate for both the minor manufacturing variances
of the precision reference as well as the differing non idealities asso-
ciated with each board layout. Thus the use of precision fixed resis-
P O W E R S U P P LY
28 Bodo´s Power Systems® July 2009 www.bodospower.com
Get <1ppm Performance from a 10ppm Precision Reference
Rejustor Technology turns a good voltage reference into an ‘industry best’
Precision References require the user to make major cost/performance tradeoffs. Untilnow, ultra-high performance came at a high price. By using a low cost, tunable resistor
divider (Rejustor) from Microbridge, precision references can be made more accurate andtheir temperature coefficients can be offset to offer unmatched performance over their
operating temperature range.
By Bob Frostholm, VP Marketing – Microbridge Technologies Corp
Figure 1: Typical external resistor configurations to trim a precisionvoltage reference
Figure 2 Fully calibrated and temperature compensated precision ref-erence design using MBT143E 9:1 Rejustor divider from Microbridge
tors would become an almost impossible task without massive reiter-
ative measurement and hand selection.
Temperature Coefficient
Temperature coefficient is the measure of the stability of the output of
the precision voltage reference with temperature changes. Precision
means nothing if it cannot be maintained over the useful operating
temperature range of the system into which the precision voltage ref-
erence is designed. TC is referred to as the second most important
specification to consider when selecting a reference. First-order tem-
perature correction (or linear temperature correction) is the largest
contributor to errors associated with temperature variations.
Datasheet examination of the ADR 425A reveals the device is speci-
fied to have an initial accuracy of +/- 3mV (+/- 0.15%) and a relative-
ly linear TC of 10ppm/°C, making it an average performer among
several competitive devices. See Figure 4.
Microbridge Rejustors provide designers with a new tool with which
to craft and adjust an application circuit. Rejustors are passive preci-
sion resistor dividers whose ohmic resistance and TC can be
trimmed independently. Once trimmed these values remain constant
and require no power to maintain their values. The TC-Offset vs. Off-
set characteristics of MBT143E divider are shown in Figure 3. The
Offset is the deviation of the divider output voltage Vin*(R1/(R1+R2)),
measured in mV per volt of divider input voltage Vin, away from
Vin*(R1o/(R1o+R2o)), where R1o and R2o are the nominal unadjust-
ed divider resistance values.
The TC-Offset is the temperature coefficient of that divider output
voltage, measured in uV per degree-C (K) per volt of divider input
voltage. Microbridge’s electrical TC adjustment software allows one
to pick target values for Offset and TC-Offset as a point within the
roughly-parallelogram-shaped region shown in Figure 3. For exam-
ple, if initially the divider input voltage were low by 5% (50mV/V) from
its designed value, and, additionally, it has an undesired +75uV/VK
temperature variation, and if it is desired that the drive level be tem-
perature-stable at the nominal Vin*(R1o/(R1o+R2o), then one pro-
grams the divider to the point (+50mV, -75uV/KV), as shown in the
figure.
The adjustment software allows you to pick a target spot within this
roughly parallelogram-shaped region. One specific example point is
shown, at Offset = +50mV/V and TC-Offset = -75uV/VK, (+50mV/V, -
75uV/VK).
For the purposes of this example the ADR425A was characterized
over the temperature range of +0°C to +85°C. Characterization and
analysis of the temperature stability of the ADR425A reveals an out-
put voltage variation of ~5000uV over the temperature, or
~12ppm/°C , close to the specification for the A grade product
(10ppm/°C) shown as the curve in Figure 4 identified as “ADR425A
as it is”.
Based on these characterization results, Rejustor calibration targets
were set and the MBT143E was adjusted “in circuit” using Rejust-it
software from Microbridge at the same time that the configuration
was being tested.
By adjusting the ohmic value of the two rejustor resistances to set the
“perfect” divider ratio, the initial output of the ADR425A was improved
from 4.99612V to 5.000125V or 0.0025% initial error. By adjusting
the TCR of the two resistances in the divider, the majority of the posi-
tive temperature coefficient was eliminated, improving the TC from
12ppm/°C to 0.8ppm/°C
Conclusion
You don’t have to pay a lot more for performance. For less than
$0.50 Rejustor technology enhances the performance of a low cost
precision voltage reference has been demonstrated to outperform
devices costing several dollars more.
www.mbridgetech.com
29www.bodospower.com July 2009 Bodo´s Power Systems®
P O W E R S U P P LY
Figure 3: A typical plot of the achievable sets of values for voltagedivider made from a specific example of Rejustors
Figure 4: Typical performance of ADR425A precision voltage refer-ence before and after calibration and compensation with MBT143ERejustor divider
Bodo´s Power Systems® July 2009 www.bodospower.com
Parameter testing
The static parameters of power semiconductors are the most impor-
tant parameters that must undergo testing after production. The IDSS
(Darin off-state current) and IGSSf/r (forward and reverse off-state
current of the gate) leakage currents, in particular, provide informa-
tion on any possible mechanical damage to the chips. Gate threshold
voltage and breakdown voltage are important indicators of doping.
But is that still enough today?
Static tests are often inadequate to meet the high demands for quali-
ty in the manufacture of special-purpose machinery and in the auto-
mobile industry. Future requirements in the automotive industry are
clearly moving in the direction of robust design and validation. But a
robust design of components that are released according to the most
modern methods does not help if process problems occur in produc-
tion and nullify such efforts. To keep the required failure rate <10
ppm, advanced tests must guarantee the quality of production.
The Dynamic Test
Dynamic tests examine the switching behavior of the component
under a load. The component will ultimately be used as an electronic
switch in converters and similar units. For IGBT, the switch at high
environmental temperatures is especially critical RBSOA (reverse
bias save operation). Components with avalanche characteristics
must pass the switching unclamped inductive load test. This test
destroys weak components.
The Thermal Impedance Test
The production of heat in any noteworthy quantity is a very undesir-
able, yet unavoidable, property of every piece of power electronics.
In this context, it is very important that the semiconductors have the
ability to dissipate as much of the heat as possible from the place of
origin (depletion layer) to the environment (heat sink) via the enclosure.
The bubbles and voids that appear during soldering of a chip impede
the required heat dissipation. The component will work for a limited
time, but sooner or later a thermal failure will occur. Measuring the
thermal impedance, Zth, provides information on the thermal connec-
tion of the chip. It is unnecessary to examine all of a characteristic:
measurements at one or two informative measurement points are
enough.
With Only One Contact
The required quality demands that these tests (switching unclamped
inductive load test, RBSOA and Zth) be performed as a 100% test of
the components. The simplest solution would be to use specialized
testers with independent contacts, possibly even for additional hot
and cold tests. But it’s readily apparent that this approach does not
make economic sense. But it is nonetheless possible to combine the
tests mentioned above in one tester, despite the different ancillary
conditions of each test. This approach saves the effort of multiple
contacts and handling time.
And the following is also possible. First, the static parameters can be
determined to deduce the processes for chip manufacture and
assembly. Second, stress tests can be performed with an increasing
load. The static parameters would then be measured again to ensure
that the stress tests did not do any damage. And all that can be done
with just one contact.
The Test System
The testers from MRS Electronic are based on 19” components and
are assembled in a type of modular system. This approach enables
optimal tailoring of the vertices (maximum test voltage, maximum cur-
rent, and the required multiplex productions) to the requirements.
P O W E R M O D U L E S
30 Bodo´s Power Systems® July 2009 www.bodospower.com
Power Module Testing withCombination Testers
Quality requires 100% testing of semiconductors
Since the Die cannot be tested thoroughly before the assembly, all parameters have to betested after the module was assembled. A complete test consists not only of parametertesting but also the thermal connection of the chip to the heat sink and the dynamic
switching behavior. Thus a test sequence consist of a pretest measuring the cold parame-ters, then the stress test dynamic and Zth and retest to make sure the part is still functional.
To conduct all these test in one station, combination testers are available.
By Günther Dörgeloh, MRS Electronic GmbH
Figure 1: Schematic Structure of an N-Channel IGBT Cell
31www.bodospower.com July 2009 Bodo´s Power Systems®www.bodospower.com July 2009 Bodo´s Power Systems®
The tests run in real time on a high-performance digital signal
processor (DSP). A regular commercial PC is used to control the
tester.
Measuring the Static Parameters
For the static parameters, a source instrument adjusts the test volt-
age or the test current and measures the variables. The measure-
ment module works continuously.
Except for the indispensable anti-alias filter, the hardware does not
contain any additional filters. The DSP performs each processing of a
signal.
From a large assortment, the optimal FIR filter can be selected for
each test. Short filter times are used for time-critical measurements.
Sensitive measurements, such as leakage currents in the lower nA
range, require longer filters. Digital signal processing enables very
exact, reproducible measurements of leakage current even in
extremely electromagnetic industrial environments.
Measuring Dynamic Behavior
The two dynamic tests used most widely are the switching
unclamped inductive load test and the double impulse test. Both
involved stress tests that destroy problematic components.
The first, the switching unclamped inductive load test, links the test to
an inductive load without an override (unclamped). The transistor is
switched off once the specified cut-off current is reached. Because
there is no override path, the current continues to flow through the
unit being tested and drives it into the avalanche breakthrough or into
linear mode if an active clamp is required over the gate.
Within a very short time (usually <500 is), a very high rating in the
chip is converted into heat.
In cases of inhomogeneous, defective doping or defective source or
emitter gate metallization, hot spots on the silicon develop that can
lead to fusion of the crystal (see Figure 2). IGBTs, in particular have
an unavoidable parasitic thyristor structure that ignites when the fail-
ures noted above occur and that can lead to a latch-up with the
resulting destruction of the component. These kinds of components
may not leave production in any circumstances. FETs have only a
part of these parasitic structures, so that there is no danger of a
www.bodospower.com July 2009 Bodo´s Power Systems®
P O W E R M O D U L E S
Figure 2: Failure of a Transistor in a Switching Unclamped InductiveLoad Test
XPT-IGBT the newest generation of short-circuit rated IGBTsDrive with the XPT-IGBT
Features:• Easy paralleling due to the positive
temperature coefficient
• Rugged XPT design results in:
– Short Circuit rated for 10 �sec.– Very low gate charge– Square RBSOA @ 3x Ic– Low EMI
• Advanced wafer technology results in low Vce(sat)
• 10-50 A in 1200 V
Used in:• AC motor drives• Solar inverter• Medical equipment• Uninterruptible power supply
For more information pleaseemail [email protected] or call Petra Gerson: +49 6206 503249
TYPE Configuration PackageMIXA20WB1200TED CBI E2-Pack
MIXA60WB1200TEH CBI E3-Pack
IXA37IF1200HJ CoPack ISOPLUS 247
IXA20I1200PB Single TO 220
www. ixys .com
For more part numbers, go to www.ixys.com
latch-up. Nevertheless, these structures are
enough to cause comparable failures (a con-
trolled increase of the parasitic bipolar tran-
sistor).
The Double Impulse Test
The setup for the double impulse test is
comparable to the setup for the unclamped
inductive load test, although an override path
exists in this case. If complete converter
modules or phase legs are being tested, the
existing diode conveniently located on the
opposite side of the transistor can be used
as an override diode. This approach corre-
sponds to later usage. A complete switching
cycle is now run: commutation of the current
into the override diode and reserve recovery
of the override diode when switching back
on. Test tester monitors the ancillary condi-
tions of the test. A rapid digital storage oscil-
loscope can also be used to perform addi-
tional detailed analysis of current and volt-
age curves online.
The dynamic tests of IGBTs are especially
interesting at high environmental tempera-
tures. The current gain of both parasitic tran-
sistors increases as the temperature rises,
as does the extrinsic base resistance. Both
increase the danger of a latch-up. An appro-
priate choice of test parameters enables
testing compliance with the safe operating
area of each component. The transformed
heat is mostly limited to the chips.
Measuring the Thermal Impedance
Measuring the thermal impedance, Zth, can
determine if the component can dissipate, in
long-term operation, the heat that is created.
Measuring the thermal impedance occurs in
three phases:
Measurement of a temperature-dependent
parameter (diode flow voltage, saturation
voltage, and so on)
Application of well-defined (electric)
energy
Second measurement of the temperature-
dependent parameter directly (in as short a
time as possible) after the power impulse
Application of well-defined energy means
that constant power must be adjusted over a
specific time. The time is determined from
the thermal time constant of the structure to
be tested, usually between 10 ms and 500
ms. The power is selected to enable good
measurement of the temperature difference
without exceeding the maximum barrier layer
temperature. The amount of warming that
has occurred can be calculated from the
change of the temperature-dependent
parameter.
This measurement of Zth corresponds exact-
ly to reality and is clearly superior to other
approaches.
Figure 3 illustrates a detailed analysis of a
module that was rejected by the tester dur-
ing the product part approval process
(PAPP) phase. Curves ZTH_1 to ZTH_4
were determined during the evaluation
phase to specify the thermal behavior of
good parts. The routine test uses a power
impulse that lasts 500 ms. It is briefly inter-
rupted after 50 ms for an intermediate meas-
urement of the temperature-dependent
parameter. No abnormalities are seen after
50 ms, but extreme abnormalities are seen
after 500 ms. The reason for the difference
is that the chip is very well soldered with
direct copper bond (DCB) and can dissipate
the heat very well to the DCB.
The bubble is located between the DCB and
the copper base plate. That’s why the bubble
cannot be seen by an X-ray. The solid cop-
per base plate absorbs so much of the X-
rays that the bubble disappears in the noise.
Only an expensive ultrasound examination
would confirm the results of the tester. Only
an analysis of both Zth results that are deter-
mined in a sort of inspection impulse con-
tains valuable information for the process
engineers.
This component was perfect in terms of the
static parameters and passed both dynamic
tests without any problems. In the field, how-
ever, it failed in a short time because of
overheating.
MRS Electronic was founded in 1971 as a
manufacturer of interface electronics and
control equipment for data processing
devices. The first test systems of MLH sys-
tems were developed in 1975 to test power
diodes and NPN transistors. The company is
headquartered in Rottweil (near Stuttgart,
Germany) and has today about 50 employ-
ees. It works with test systems for semicon-
ductors along with measurement and regu-
lating technology for industrial applications,
automotive electronics, and electronics man-
ufacturing as a service.
Increasing quality demands require compre-
hensive, 100% testing of semiconductors.
Static tests are no longer enough. Only sup-
plemental dynamic testing procedures, such
as switching unclamped inductive load and a
double impulse test can meet the require-
ments. Inspection of thermal impedance, Zth,
is very important in the process. The sim-
plest approach uses a test system that can
perform all these tests with only one contact.
www.mrs-electronic.de
32 Bodo´s Power Systems® July 2009 www.bodospower.com
Figure 3: Rejection of a Chip with Poor Ther-mal Fastening
P O W E R M O D U L E S
Digital Power Workshop based on
TI´s F28x family.
For more information and
your free drill hole stencil
please visit www.biricha.com
Biricha Digital Power offering
13th European Conferenceon Power Electronics
and Applications
Receipt of synopses:Monday 3 November 2008
Receipt of full papers:Monday 11 May 2009
EPE 2009 Barcelona, Spain
www.epe2009.com
Throughout time different methods of meas-
urement of the electrical current have been
developed. The different methods are based
on purely electrical, magnetic or optical prin-
ciples, or making use of the behaviour that
some materials have in presence of a mag-
netic field [1].
The more suitable method in each case
depends on the characteristics of the current
to be measured: DC, AC or both simultane-
ously, frequency, peak value, accuracy, iso-
lation, etc. Among the different methods we
find: shunts, current transformers, Hall-effect
and Rogowski effect transducers, Flux-gate
effect transformers, and other alternative
methods of less common use.
A brief summary of the main methods is pre-
sented.
Shunt
The method is based on the measure of volt-
age that appears on a resistance (shunt)
due to electric current, according to Ohm’s
law.
The method is extremely simple and suitable
for measurement of DC and AC accurately,
but its major drawback is the absence of iso-
lation between the power and measure cir-
cuits and the high power consumption in
case of high current measurement.
Current Transformer
Based on electromagnetic principles, is an
AC transformer where the secondary current
is related to the primary according to the
transformer turns ratio.
It consists of a toroidal core, where the sec-
ondary winding is wrapped. The conductor
through which the current to be measured
circulates become the primary winding.
The transformer is carefully conceived in
order that the leakage current and the losses
in the core are very small, so that no signifi-
cant errors are introduced in the measure.
The main advantage of this method is its
simplicity and robustness, while the main
drawback is that it is only suitable for AC
measures.
Hall Effect Transformer
The Hall sensor measures the voltage that
appears in a semiconductor in the presence
of a magnetic field perpendicular to the
plane of the material when a current circu-
lates along this material (Hall effect).
The transformer consists of a toroidal mag-
netic core with a gap for the Hall probe
placement. The magnetic core is used to
direct the magnetic field originated by the
current to be measured. An additional elec-
tronic circuit processes the signal generated
by the Hall probe.
The principal advantage of this system is the
ability to measure DC and AC currents, up to
frequencies of 100 kHz, with an acceptable
precision and a galvanic isolation.
Rogowski Transformer
The Rogowski transformer has a toroidal
structure, but with a coil wrapped on a non-
magnetic core (named Rogowski coil) and its
structure can be either rigid or flexible.
The current to be measured crosses the
Rogowski coil and generates a voltage pro-
portional to the rate of change of this current
and the mutual inductance between the coil
and the conductor. The value of the current
to be measured is proportional to the integral
of this voltage.
The advantage of this transducer is the
impossibility of saturation of the magnetic
core (i.e. air or plastic), but is not appropriate
for DC measures and its accuracy and band-
width are conditioned by the integrator cir-
cuit.
Flux-gate Transformer
This transformer, with physical structure sim-
ilar to the Hall transformer, is based on the
detection of the saturation state of a magnet-
ic circuit. The magnetic core is built using a
high permeability material, which is
immersed in the magnetic field to be meas-
ured.
The magnetic material is excited by a signal
that, in absence of external magnetic field,
leads the magnetic material to the symmetri-
cal saturation. This symmetry is lost with the
existence of an external magnetic field.
S E N S O R S
34 Bodo´s Power Systems® July 2009 www.bodospower.com
Low Consumption Flux-GateTransducer
Both AC and DC High-Current Measurement is achieved
This article presents the design and implementation of a transducer system for the meas-urement of AC and DC high-current using the flux compensation technique in the meas-urement transformer, also called Flux-gate technology. The system can measure currents
greater than 700 A (peak value) with a 100 kHz bandwidth measured at -3 dB.
By Manuel Román*, Guillermo Velasco*, Alfonso Conesa * and Felipe Jeréz ***EUETIB – CEIB and **Grupo PREMO S.A.
*Technical University of Catalonia (UPC) - Department of Electronic Engineering (DEE)*Compte d’Urgell 187, 08036 – Barcelona SPAIN
Picture 1: fluxgate currents transducer range
The injection of current in an auxiliary wind-
ing creates a compensating magnetic field
that restores the symmetry of the hysteresis
cycle. The injected current compensates the
magnetic field created by the current to be
measured, and its value is proportional to
this current.
This system is suitable for the measurement
of DC and AC currents, with high accuracy,
high frequency and high current range.
Other Methods of Measurement
So far the most important methods used for
the measurement of the electrical current
have been reported.
All of them are based on the detection of the
magnetic field created by this current, with
the exception of the direct measurement by
means of a shunt.
Other methods for current measurement are
based on the properties of materials sensi-
tive to magnetic field, like those based on
the magneto-resistive and magneto-optical
principles or those based on the magneto-
diode, magneto-transistor and superconduc-
tors components, etc… [2 - 3].
Comparison Between the Presented
Methods
Table I presents a comparison between dif-
ferent methods of current measurement
described above using some parameters
that determine their main characteristics and
usual applications.
Flux-gate transducers
These transducers have similar physical
structure to Hall transformers. They are
based on the detection of the saturation
state of a magnetic circuit. The ferromagnet-
ic material used presents high permeability
and is immersed in the magnetic field to
measure. This system is appropriate in a
large range of DC and AC current measure-
ment with high accuracy, until maximum fre-
quencies around 100 kHz.
The “Flux-gate” term makes reference to the
principle of operation on which some trans-
ducers are based for the measurement of
current with isolation. In this type of trans-
ducers the magnetic field generated by the
current to measure is detected by means of
a sensor.
In a similar way to the transducers based on
the Hall-effect, the usually called standard
Flux-gate transducers use a toroidal magnet-
ic circuit which includes an air gap with the
field measuring element and a secondary
winding. The main difference between the
transducers Hall and standard Flux-gates
consists on the element to measure the field
that crosses the magnetic circuit, as shows
Figure 1. Hall cells and saturable inductors
are used respectively.
The current transducer proposed in this work
is based on non-standard Flux-gate trans-
ducer. The non-standard transducer devel-
oped uses its own toroidal core as field
detector, and does not include any gap in
the magnetic path. An auxiliary winding is
added to the core and that results in a core-
wounded set, which is used as a saturable
inductor flow detector.
In order to detect a null field in the magnetic
circuit, the secondary winding is excited with
the necessary current. In this circumstance,
the transducer works with zero field condi-
tion, as shown in Figure 2. This condition
verifies that the current imposed, by the sec-
ondary winding, is directly proportional to the
primary current to be measured (IP).
The relationship between the primary and
secondary current is (1), being NS the num-
ber of turns of the secondary winding.
(1)
The detection of the zero flux condition in
the magnetic path of the transducer is based
on the change of the inductance value of the
saturable inductor formed by the ferromag-
netic core and the auxiliary winding. In
absence of current to be measured (IP), the
net flux through the saturable inductor core
is zero.
Under these conditions, if a squared voltage
waveform is applied to the auxiliary winding
(NA), the current waveform in auxiliary wind-
ing will be as shown Figure: 3.
The effect of the current to be measured (IP)
on the auxiliary winding current (NA), when a
square voltage waveform is applied, leads to
an average current value different from zero,
as shows Figure 4.
The average value obtained and its sign will
depend on the specific value and direction of
the current IP.
Flux-gate transducer designed
The designed transducer operates in a
closed-loop, according to the general
scheme shown in Fig. 2. The null average
value condition in the current by the auxiliary
winding (NA) is used to determine the condi-
tion of zero flux in the core. The basic oper-
ating principle of the designed transducer is
in Figure 4.
SSP INI ⋅=
S E N S O R S
35www.bodospower.com July 2009 Bodo´s Power Systems®
Figure 1: Structure of a Flux-gate transduc-er: a) standard and b) without gap in themagnetic path
Figure: 2 Basic principle of behaviour ofFlux-gate transducers
Figure 3: Voltage waveforms of excitationand current by the auxiliary winding underzero flux conditions
Figure 4: Voltage waveforms of excitationand current by the auxiliary winding undernon zero flux conditions
Table1: Comparison of methods for current measurement
Parameter Shunt Current transformer
Hall transformer
Rogowskitransformer
Flux-gate transformer
AC/DC measure AC/DC AC AC/DC AC AC/DCBand Width Low Low Middle Middle HighIsolation No Yes Yes Yes YesLinearity High High Middle High Very High Precision Middle Middle Middle Middle Very HighOffset Yes No Yes No NoHigh current Bad Middle Middle Good Very Good Saturation effect No Yes Yes No NoTemperature dependence
Middle Low High Very Low Low
Power Consumption High Low Low Low MiddleSize Very Low Low Low Middle Middle
A disadvantage presented by this structure is
the possible injection of noise on the primary
current measure (IP). This noise comes from
the auxiliary current (IA), and can be cou-
pled in the primary current due to the trans-
former effect in the magnetic core of the
transducer. The solution usually adopted to
avoid this phenomenon is the use of a sec-
ond core with a new auxiliary winding. These
two cores with their auxiliary windings must
be identical under ideal conditions [4].
Now the secondary winding (NS) in which
the compensation current of flux in the trans-
ducer is applied will be common to both aux-
iliary cores.
The purpose of this second auxiliary core
(also a second saturable inductor) is to com-
pensate the injected noise in the primary
current by the first saturable inductor. If the
second auxiliary core (NA2) is excited with
an equal current but in reverse direction to
the current used for exciting the first auxiliary
core (NA1), the currents induced on the pri-
mary current conductor (IP) will be equal
and in opposite direction, cancelling its
effect.
The block diagram that represents the
designed measurement system is shown in
Figure 6.
Signal Generator for the Excitation of
Auxiliary Windings
It is based on a comparator circuit with hys-
teresis (or Schmitt trigger). This circuit will
change the value of its output voltage when
the circulating current on the main winding
excitation (IA) exceeds a threshold value.
The magnetic component of measure is
included in the oscillator circuit, and the
electrical characteristics of this component
will influence the frequency of oscillation of
the squared signal generator circuit.
For the transducer built, this frequency is
around 300Hz.
Symmetry Detector of the Auxiliary
Current (IA)
In absence of primary current (IP) the aver-
age value of the current of excitation (IA) is
zero. The effect produced by the existence
of a primary current is the appearance of an
average value different from zero and sign
dependent of the sense of of this current.
For the automatic adjustment of the value of
the secondary current winding (IS), the use
of a PI controller is proposed, in order to
ensure that the primary current excitation
winding has zero mean value.
This controller cannot guarantee the proper
functioning of the measurement system at
the start-up process if a primary current (IP)
even of moderate value is already circulat-
ing, because under these conditions the two
inductors will be saturated.
A primary current through the measurement
system in non zero flux conditions, produce
a high frequency current (some tens of kHz)
in the main excitation winding (IA) and non
zero average value, with independent sign of
the primary current circulating sense (IP).
To overcome this drawback, an additional
controller is included which ensures that the
zero flux condition is reached regardless the
value that the primary current (IP) could take
at the start-up time.
The operation of this controller is based on
the property mentioned previously, where the
frequency of the current of the main excita-
tion winding (IA) is high frequency when the
system is not balanced and low frequency
when the system is operating in the vicinity
of the point of zero flux. The presence of this
additional controller increases the robust-
ness against possible situations, as sporadic
transients, ensuring the accomplishment of
the equilibrium conditions.
As Figure 6 shows, this controller includes a
triangular low-frequency oscillator, a frequen-
cy detector for the excitation current (IA) and
an analog switch controlled by the frequency
detector.
While the measurement system does not
operate in zero flux conditions, the input of
the current compensation driver (IS) will be
connected to the triangular low-frequency
signal generator. This waveform guarantees
that, in some moment, a value of the winding
current compensation (IS) next to the neces-
sary condition of zero flux will be reached.
When this happens, the frequency of the
current of the main excitation winding (IA)
decreases. This situation is detected and the
originally proposed PI controller then is con-
nected to the input of the current compensa-
tion driver.
Valid Measure Indicator
The output of the low-frequency detector cir-
cuit is connected to the indicator of valid
measure. This indicator is only activated
when it detects that the current of the pri-
mary winding excitation is of low frequency,
effect that will occur when the system works
in zero flux conditions.
An LED indicator and a relay are the output
elements to indicate the condition of zero
flux valid measure.
Driver to Generate the Compensation Cur-
rent
This circuit is used to generate the current
that will flow through the secondary winding
compensation (NS).
A class D amplifier has been used for the
implementation. These amplifiers present the
advantage of high efficiency compared to lin-
ear amplifiers, but add harmonics of the
same switching frequency and higher.
Is based on a pulse width modulator (PWM),
which generates a squared voltage wave-
form with a duty cycle proportional to the PI
controller output signal.
The output PWM squared waveform of the
modulator is applied to the compensation
winding (NS) through a current driver imple-
mented with a half-bridge inverter.
The own inductance of the NS winding filters
the current that circulates along it. So, the
output voltage of the system measured in a
shunt resistance connected in series with
this winding is proportional to the primary
current to be measured.
Measure of High Frequency Currents
The system of measure proposed, based on
the Flux-gate principle, is only suitable for
the measurement of current in DC or in AC
at low frequencies. The maximum frequency
for the AC measurement is fixed by the
working frequency of the zero flux detection
system.
36 Bodo´s Power Systems® July 2009 www.bodospower.com
S E N S O R S
Figure 5: Principle of operation of the Flux-gate transducer designed
Figure 6: Block diagram of the designedFlux-gate transducer
For the measure of high frequency AC cur-
rents and to obtain a suitable dynamic
behaviour in case of fast variations of cur-
rent, a third core is included. This new core
is embraced only by the compensation wind-
ing (NS) and it works exclusively as a con-
ventional current transformer.
Power Supply
Voltage power supply of the transducer is
obtained from a flyback DC / DC converter.
In this way, two stable output voltages (+12V
positive and the other negative -12V) are
derived from a single input voltage which
can be between 10V and 30V.
Experimental results
Figure 7 shows the final look of the con-
structed prototype. It displays the trans-
former of measurement, the PCB with the
used electronics components and the
designed box to contain the measurement
system.
This prototype has been successfully tested
and here we present some of the results
obtained for four different types of measure-
ments.
In the first two experiments 35 times the
main conductor has been coiled on the
measurement transformer. Hence, the cur-
rent measured by the transducer will be 35
times greater than the real value of the IP.
The IP current is measured in a 3 mV/A
shunt, and the IS current, generated by the
designed transducer, is measured on a 1 Ωresistance. Since 1000 turns for NS have
been used, the output voltage of our trans-
ducer will be of 1 mV/A.
Figure 8 shows the result of a 525A DC cur-
rent measurement performed with the
described shunt (CH3) and with the
designed transducer (CH2).
Figure 9 shows the result of the measure-
ment of a 350 A of amplitude squared cur-
rent and 1 kHz of frequency performed with
the described shunt (CH3), and with the
designed transducer (CH2).
Figure 10: shows the result of the measure-
ment of a sinusoidal current of 400 A ampli-
tude and European network frequency (50
Hz), performed by the designed transducer.
Finally, Fig.11 shows the result of a 100 kHz
sinusoidal current of 133 A amplitude meas-
urement, performed by the described shunt
(CH3), the designed transducer (CH1) and
another industrial Flux-gate transducer
(CH2).
Conclusions
A transducer based on the Flux-gate tech-
nology for the measurement of DC and AC
current has been designed and tested.
The power supply of the transducer and the
driver for generating the compensation cur-
rent (IS) are based on high-efficiency DC/DC
power converters. This choice guarantees
the low consumption of the designed system
compared to the existing solutions in the
market.
The main characteristics of the designed
system are the following ones:
• Maximum peak current: 1000 A
• Primary rated current (IN): 700 A
• Small signal bandwidth (5% of IN):
DC to 100 kHz
• Conversion Ratio: 1:1000
• Supply voltage: from 10 to 30 VDC
In order to obtain a more detailed characteri-
zation of the described transducer, the refer-
ence DCT-700A can be consulted at the
products general datasheet of Group
PREMO S.A. [5]. This datasheet is accessi-
ble online.
Acknowledgment
This work has been partially funded by the
PROFIT program of the Spanish Ministry of
Industry and Energy under the project refer-
ence FIT-330100-2006-20.
References
James E. Lenz. ‘A Review of Magnetic Sen-
sors’. Proceedings of the IEEE. Vol.: 78, Iss:
6, pp. 973 - 989. June 1990.
Erik R. Olson and Robert D. Lorenz. ‘Inte-
grating Giant Magneto-resistive Current and
Thermal Sensor in Power Electronic Mod-
ules’. Eighteenth Annual IEEE Applied
Power Electronics Conference and Exposi-
tion, 2003 (APEC’03). Vol.: 2, pp. 9 - 13.
February 2003.
Emerging Technologies Working Group and
Fiber Optic Sensors Working Group. ‘Optical
Current Transducers for Power Systems: A
Review’. IEEE Transactions on Power Deliv-
ery. Vol.: 9, Iss: 4, pp. 1778 - 1788. October
1994.
T. Sonoda, R. Ueda and K. Koga. ‘An ac and
dc Current Sensor of High Accuracy’. IEEE
Transactions on Industry Applications. Vol.:
28, Iss: 5, pp. 1087 - 1094. September-Octo-
ber 1992.
Group PREMO S.A. ‘II - General Datasheet’.
2007. Online Access: www.grupopremo.com
www.grupopremo.com
37www.bodospower.com July 2009 Bodo´s Power Systems®
S E N S O R S
Figure 7: The designed Flux-gate transducer
Figure 9: Measurement of a 1kHz squaredwaveform current
Figure 10: Measurement of a 50Hz sinu-soidal current
Figure 11: Measurement of a 100kHz sinu-soidal current
Figure 8: Measurement of a DC current
There are many well known circuits that fit certain applications very
well, but if we want to find the best solution for a given application
then we need to understand how the choice of topology may affect
the key parameters of the power supply, which are cost, overall size,
EMC performance and efficiency.
The importance each of these parameters varies between application
and trade-offs made between each of them have a significant effect
on the choice of topology. For example, a solution targeting the low-
est cost may be quite different from one targeting the highest efficien-
cy; counter-intuitively a solution targeting the highest power density
(smallest size) won't necessarily achieve the lowest losses. With the
development of the new TDK-Lambda EFE series, a high density AC-
DC supply aimed at embedded front-end applications, an iterative
design method was used to determine the best possible topology.
The key design criteria set for the development of the EFE300 for
instance were the overall dimensions of 5x3-inch and 1U compatible,
300W continuous power rating with 2m/s airflow and 400W peak
power rating for 10 seconds, all achieved over the whole supply volt-
age range 90-264Vac, and curve B conducted EMC (EN55011,
EN55022). Another important design objective was to design the sup-
ply following the tried and tested TDK-Lambda design rules for com-
ponent deratings, spacings, PCB design rules and so on.
These design rules have been proven successful by our engineers
when designing earlier products and following them rigidly reminds
us not to make compromises simply to achieve a small size. In addi-
tion, they are important in creating a product that can be easily man-
ufactured and is reliable. It is always possible to squeeze more com-
ponents into a small space to achieve a “high density” power supply,
but components that are mechanically stressed or over-heated due to
insufficient space being allowed for them will lead to reductions in
overall reliability.
In order to choose the most appropriate topology we need to consid-
er the attributes of possible solutions, shown in Table 1, by adopting
a simple scoring process. Each topology is evaluated in two stages;
firstly for switching and efficiency characteristics, shown in Table 2,
and then for circuit complexity, EMC performance and other relevant
factors of the target application, shown in Table 3.
Each attribute is ranked, from “1” meaning bad through to “5” good,
and the results of both stages are weighted together.
For the topologies considered, Table 2 shows the scoring for the
switching- and efficiency-related characteristics. In general, a topolo-
gy with soft switching and low circulating current will be preferred.
The key issues for the attributes that are scored are:
Circulating VA: The amount of energy circulating within the topology
that is not contributing directly to powering the load. For higher output
voltages, this criterion is less of a concern due to the lower second-
ary-side currents involved.
P O W E R S U P P LY
38 Bodo´s Power Systems® July 2009 www.bodospower.com
Choosing The Right TopologyDesign rules have been proven successful
Whatever the final application, power supply users want reasonably priced products thatare small and efficient, yet have excellent EMC performance. The article explains how the
choice of power supply topology can greatly affect these key parameters.
By Andrew Skinner, Advanced Development Manager, TDK-Lambda
Table 1: Possible topologies
“NV” topologyZero-voltage switching variable frequency modified LLC with self-driven synchronous rectifiers. TDK-Lambda patented.
Half-bridge LLC resonant converterZero voltage switching variable frequency converter with activelydriven synchronous rectifiers (self drive is not possible)
Full-bridge Current Doubler Zero voltage switching pulse-width modulation controlled converter with current-doubler rectification.
Half-bridge Current DoublerHard switched half-bridge pulse-width modulation controlledconverter with current-doubler rectification.
ZVZC switching DC-DC transformer plus pre-regulator.Zero voltage and zero current switching resonant DC:DC “transformer” with buck pre-regulator (zero current or boundary mode controlled)
ZCS Resonant Converter Zero current switching , variable frequency resonant converter
Quasi-resonant converter Flyback converter operated with valley switching and activelydriven synchronous rectifier
Table 2: Switching and efficiency characteristics
Topology
Primary switching Secondary switching Circulating VA Efficiency
Turn-on Turn-off Turn-on Turn-off Low outputvoltage
High output voltage Light load Full load
NV topologyself-driven synchronous rectifier
ZVS ZVS ZVS High di/dt 4 3 2 4
Half-bridge LLC synchronous rectifier ZVS ZVS ZVS ZCS 3 5 3 5
Full-bridgecurrent doublersynchronous rectifier
ZVS ZVS/ hard ZVS High di/dt 5 3 2 5
Half-bridge current doublersynchronous rectifier
ZCS/hard ZVS ZVS High di/dt 4 3 3 5
ZVZCSResonant dc transformer plus pre-regulator
ZVS ZVS/ ZCS ZVS ZCS 2 4 2 4
ZCSResonant converter -fixed on-time variable frequency
ZCS ZVS ZVS ZCS 3 4 4 3
Quasi-resonantflyback synchronous rectifier with diode emulation
ZCS ZVS ZVS ZCS 1 3 2 4
Efficiency: Topologies with continuous output current and actively
controlled synchronous rectifiers will be preferred, although generally
this will lead to a lower circuit complexity score. With the appropriate
selection of components and switching frequency, most circuits can
be made highly efficient; this score is more a measure of the design
effort required to achieve high efficiency.
The key attributes relating to the circuit complexity, EMC perform-
ance and other application relevant factors of the topologies is shown
in Table 3. Let’s take a look at the key issues for each.
Multiple outputs: A current-driven circuit without an output inductor is
preferable since this simplifies the mechanical design.
Semi-regulated outputs: Where several outputs (that track one anoth-
er with reasonable accuracy) are required a current driven design
with a low secondary-side di/dt is preferred.
Size: Generally this is dependent on parts count within the power cir-
cuit for low power applications, and efficiency and even loss distribu-
tion, within components, at higher power.
Complexity: A topology that naturally limits current, does not place
stringent timing requirements on the controller and has self-driven
synchronous rectifiers will result in a better score for the power cir-
cuit. A topology that has only low-side drive requirements and
requires only a simple controller will score better for control circuit
complexity.
Complexity is also a measure of cost, generally a topology that is
complex and has a higher parts count will have a higher labour cost
associated with its manufacture. The component costs will also be
expected to be higher for a couple of reasons. Firstly, complex
topologies tend to be less popular and hence controllers for them sell
in lower volume, use a larger die size or require a larger package
with more pins. The second reason, especially at lower power rat-
ings, is that the topology generally requires more devices and will
tend to use relatively small die thereby increasing the relative cost of
the device packaging.
EMC: Topologies with zero-voltage switching (ZVS) on the primary
side will have a better score since this often results in lower conduct-
ed EMC.
Wide adjustment range: Topologies that have zero-voltage switching
often utilise the transformer magnetising current to assist in this. If
the output voltage needs to vary over a wide range, then the mag-
netising current will also vary resulting in variable EMC, efficiency
and component stress. Most soft-switching topologies therefore oper-
ate with a reduced output voltage range compared to hard-switching
pulse-width modulated topologies.
The scores are combined with weighting where the most important
factors for a high density product (size and efficiency) are more heav-
ily weighted than complexity (digital control makes the use of circuits
with complex control algorithms, for example, an acceptable design
choice). RFI is also given high weighting as curve B performance is a
design objective. After applying these weightings, we arrive at the
scores in Table 4. There are four topologies that have similar scores.
The low ripple current in the output filter of the NV topology make it
the preferred choice at the power level and size of interest. In a 5x3-
inch footprint, the number of output filter capacitors is limited due to
PCB real estate constraints and can lead to electrolytic capacitors
operating with high ripple current, which will result in more self-heat-
ing and reduced life. As the achievable operating frequency with the
NV topology is relatively high, a few ceramic capacitors are used
instead, resulting in a compact output filter design with enhanced life.
The industrial markets addressed by TDK-Lambda often require
power supplies that have characteristics modified to suit the end
application, such as non-standard voltages, modified loop-response,
modified signals, etc. By utilising the power of digital control, these
requirements are met more easily and, at the same time, the con-
cerns associated with the more complex control algorithm of a reso-
nant converter are alleviated since with digital control any control law
can be implemented. The EFE series was therefore designed with an
8-bit microcontroller to control the resonant converter and to manage
housekeeping functions.
Although there are similarly rated products available using LLC and
DC-DC transformer with pre-regulator, the high reliability design rules
used at TDK-Lambda don’t allow the small component spacings
needed to fit these topologies into a 5x3-inch format.
Choosing the right topology is vital to the success of the product and
requires detailed knowledge of all the available choices. The correct
choice results in a product with optimum cost, performance and relia-
bility characteristics. Especially for high density power supplies, such
as TDK-Lambda’s EFE series, squeezing an unsuitable topology into
the available space can seriously damage your wealth!
www.emea.tdk-lambda.com
P O W E R S U P P LY
39www.bodospower.com July 2009 Bodo´s Power Systems®
Table 3: Key attributes relating to the circuit complexity, EMC performance and other application relevant factors of the topologies
Topology
Multipleoutputs
Semiregulated outputs
Size ComplexityEMC
Wide adjustrange
Low power
High power
Powercircuit Control
NV topology self-driven synchronous rectifier
5 4 5 3 4 1 5 2
Half-bridge LLC synchronous rectifier
5 5 4 5 4 1 5 2
Full-bridge current doubler synchronous rectifier
2 3 1 5 3 2 5 2
half-bridge current doubler synchronous rectifier
2 3 2 5 2 4 3 5
ZVZCSResonant dc transformer + pre-regulator
5 4 4 4 4 4 5 3
ZCS Resonant converter -fixed on-time variable frequency
5 4 4 2 5 3 4 2
Quasi-resonant flyback synchronous rectifier with diode emulation
5 5 5 1 5 5 4 3
Table 4: Weighted scores for TDK-Lambda EFE300
Topology Weighted score Comments
NV topology self-driven synchronousrectifier 0.89 Continuous output current
Half-bridge LLC synchronous rectifier 0.86 High output filter ripple current and requires active synchronous rectifier gate-drive
Full-bridge current doubler synchronous rectifier 0.64
half-bridge current doubler synchronous rectifier 0.63
ZVZCS Resonant dc transformer +pre-regulator 0.85 High output filter ripple current
ZCS Resonant converter -fixed on-time variable frequency 0.75
Quasi-resonant flyback synchronous rectifier with diode emulation 0.9 Very high output filter ripple current
40 Bodo´s Power Systems® July 2009 www.bodospower.com
This article presents an investigation in the algorithms for sensorless
detection of rotor angular position of Permanent Magnet Synchro-
nous Motors (PMSM) and the actual implementation STMicroelec-
tronics has made in its Field Oriented Control (FOC) firmware library
developed for its 32-bit Cortex-M3 microcontroller family, the STM32.
Three phase AC motors (induction and synchronous motors) histori-
cally owe their greatest success in adjustable speed application to
the advent of vector control drives. In particular, Field Oriented Con-
trol (FOC) has guaranteed the highest reachable performances, but
its scope of employment has always been in costly drives mainly
used in the automation and robotics field.
In fact, it requires three essential components, in addition to the con-
verter/inverter electronic stage: intense computational power, accu-
rate sensors for current, and rotor speed/position measurement.
For these reasons, in the last decade, low cost applications (white
goods, HVAC, industrial, pumps) have appointed induction motors
and V/F controls as best compromise for cost containment and target
performance.
This scenario has rapidly changed.
The always stringent requirement of energy efficiency has been com-
plemented, and complicated, by an almost worldwide commitment to
environmental sustainability.
PMSM sinusoidal motors have consolidated – because of advance-
ments in magnet chemistry and for their reduced copper content - as
viable and robust as induction motors.
Powerful 32-bit microcontrollers are now offered as a suitable choice
in the same cost area of 8-bits, throwing open the doors to implement
more advanced control methods, such as FOC.
Some of these microcontroller families, such as the STM32, feature
fast and precise ADCs and advanced PWM timers designed specifi-
cally for motor control, to the extent that the last two cost barriers –
current and speed sensors – have now faded. Current feedback can
now be achieved using inexpensive shunt resistors, while
speed/position sensors have been replaced by mathematical algo-
rithms that bring additional benefits such as improved reliability and
reduced circuit complexity and size.
Because of this, low-cost motor control drives can finally take advan-
tage of the intrinsic properties of FOC, i.e. closed loop current control
(and hence stringent over-current protection, which in turn means a
more focused choice of power devices), best efficiency and dynamics
(even during transients), and quiet operations.
This article presents an investigation in the algorithms for sensorless
detection of rotor angular position of Permanent Magnet Synchro-
nous Motors (PMSM) and the actual implementation STMicroelec-
tronics has made in its Field Oriented Control (FOC) firmware library.
Two methods will be shown in details, as representatives of the two
categories to which they belong: a flux estimator, and a closed-loop
Luenberger’s state observer. They both use the mathematical model
of the machine, while the second one also exploits current feedback.
The aim is to compare them, on the basis of computational effort,
performances, and robustness to parameter variations.
PMSM equations
With reference to the figure, the motor voltage and flux-linkage equa-
tions of a PMSM (isotropic magnetic structure) are generally
expressed as:
M O T I O N C O N T R O L
Comparison of Sensorless Algorithms for PMSM Rotor
Position DetectionPermanent Magnet Synchronous Motors (PMSM)
and the actual implementation
It’s no news that richer performance vector control drives for AC motors could be beneficial, in terms of dynamics and efficiency as well as for every aspect an ideal control
should have. The good news is that the time has come to make wide use of these drives,running low-cost applications with first grade quality.
Stello Matteo Billé, Dino Costanzo, Antonio Cucuccio, STMicroelectronicsAlfio Consoli, Mario Cacciato, Giuseppe Scarcella, Giacomo Scelba, University of Catania
Figure 1: The motor voltage and flux-linkage
41www.bodospower.com July 2009 Bodo´s Power Systems®www.bodospower.com July 2009 Bodo´s Power Systems®
where :
rs is the stator phase winding resistance
Lls s the stator phase winding leakage inductance
Lms is the stator phase winding magnetizing inductance
θr is the rotor electrical angle
Φm is the flux linkage due to permanent magnets
The complexity of these equations is apparent, as the three stator
flux linkages are mutually coupled, and what is more, as they are
dependent on the rotor position, which is time-varying and a function
of the electromagnetic and load torques.
The reference frame theory simplifies the PMSM motor equations, by
making a change of variables that refers the stator quantities abc
(spatially displaced by 120°) to qd components, directed along axes
each 90° apart, rotating at an arbitrary speed ω.
Open-loop flux estimator equations
In particular, if the qd reference frame is stationary (the so-called
Clarke transformation on the αβ frame, ω=0) we’ll get to:
where Ls is the synchronous inductance, Ls = Lls + 3Lms/2,
Let’s define the back-EMFs (Electro-Motive Forces) as:
Substituting the flux linkages equations in the voltage equations and
solving:
These equations give us the possibility of an open-loop estimation of
the rotor flux that goes through known motor parameters and meas-
ured terminal voltage and currents. As can be seen, rotor angular
position information is held directly and can be extrapolated by
means of an arctangent function, recursive algorithm, or PLL (Phase
Locked Loop).
Going into time-discrete domain, it gives:
and whose block diagram is depicted below:
The observer
To understand how a state observer works and why it can be consid-
ered a closed loop approach; let’s start from the definition of ‘observ-
ability.’
In control theory, a system is said to be ‘observable’ if it is possible to
fully reconstruct the system state starting from its output measure-
ments. The ‘state observer’ (or simply the ‘observer’) is then the sys-
tem that provides the estimation of the observed system internal
state (X) given its input (U) and output (Y) measurements.
The state observer thus allows the reconstruction of the internal state
of a system (on condition that it is observable) without having direct
access of it but simply starting from the knowledge of system inputs
and outputs.
As shown in the above figure, the observer is composed of two
blocks. The first block is constituted by a model of the plant whose
internal state we want to know. Due to unavoidable inaccuracies on
plant model and to parameters’ deviations or uncertainties, the states
would be destined to diverge from real state X if we computed it
by simply applying input U to our plant model. On the contrary, this
can be avoided by ‘closing the loop’ - that is, by feeding back the
error between the measured and observed outputs.
In fact, multiplying this error by the observer gain K and reporting it
as input to our plant reproduction allows a continuous adjustment of
the plant model guaranteeing the coherency between X and . X̂
X̂
X̂
( )
( ) )()()()(
)()()()(
1
1
1
1
kiLThirhvk
kiLThirhvk
ss
k
hsPM
ss
k
hsPM
ββββ
αααα
λ
λ
−⋅−=
−⋅−=
∑
∑−
=
−
=
( )( ) ββββ
αααα
θλ
θλ
iLdtirv
iLdtirv
ssrmPM
ssrmPM
−−=Φ=
−−=Φ=
∫∫
cos
sin
rrmmr
rrmmr
dtde
dtde
θωθ
θωθ
β
α
sin)(cos
cos)(sin
Φ−=Φ
=
Φ=Φ
=
⎩⎨⎧
Φ⋅+=Φ⋅+=
⎪⎪⎩
⎪⎪⎨
⎧
+=
+=
mrs
mrs
s
s
iLiL
dtd
irv
dtdirv
θλθλ
λ
λ
ββ
αα
βββ
ααα
cos
sin;
m
r
r
r
abc
mlmm
mml
m
mmml
abc
abcabcsabc
s
ss
ss
s
ss
s
ss
ss
s
s
ss
i
LLLL
LLL
L
LLLL
dtd
irv
Φ
⎥⎥⎥⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢⎢⎢⎢
⎣
⎡
⎟⎠⎞
⎜⎝⎛ +
⎟⎠⎞
⎜⎝⎛ −+
⎥⎥⎥⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢⎢⎢⎢
⎣
⎡
+−−
−+−
−−+
=
+=
3
2sin
3
2sin
sin
22
22
22
πθ
πθ
θ
λ
λ
M O T I O N C O N T R O L
Figure 2: PLL (Phase Locked Loop) block diagram
Figure 3: The state observer
For all these reasons, a state observer can be considered a closed
loop system.
We can now apply these concepts to PMSM equations in order to get
an estimation of its internal state starting from its input and from the
measurement of its output.
Let’s consider PMSM motor equations:
In order to make this system linear, we introduce two new state vari-
ables:
Supposing now the mechanical quantities slowly changing versus the
electrical ones, we can write that:
This is, under the position x = [iα iβ eα eβ ]t, u = [vα vβ]t, y = [iα iβ]t,
the canonical ISO representation of a linear time-invariant system:
It is possible to demonstrate that the PMSM, so represented, is an
‘observable’ system; then starting from the measurement of its cur-
rents iα iβ and knowing the applied voltages vα vβ, it is possible to
reconstruct the B-emfs values eα eβ:
The feedback given by the error between estimated and measured
currents allows adjusting the motor equation, strongly decreasing the
sensitivity to motor parameters’ uncertainties and inaccuracies.
In order to have a clear idea of the computational complexity of this
approach, let’s write the observer equations:
Going into time-discrete domain, it gives:
Comparative Flux estimator vs Observer
In order to compare the performances of the flux estimator with ones
of the state observer, two aspects have been taken into considera-
tion: the computational complexity and the robustness to motor
parameters’ uncertainties and inaccuracies.
⎪⎪⎪⎪
⎩
⎪⎪⎪⎪
⎨
⎧
−−+=+
+−+=+
+−−+−=+
+−−+−=+
TkepkikiTKkeke
TkepkikiTKkeke
kvLTke
LTkikiTKki
LTrkiki
kvLTke
LTkikiTKki
LTrkiki
r
r
sss
s
sss
s
)(ˆ))()(ˆ()(ˆ)1(ˆ
)(ˆ))()(ˆ()(ˆ)1(ˆ
)()(ˆ))()(ˆ()(ˆ)(ˆ)1(ˆ
)()(ˆ))()(ˆ()(ˆ)(ˆ)1(ˆ
2
2
1
1
αββββ
βαααα
βββββββ
ααααααα
ω
ω
⎪⎪⎪⎪⎪
⎩
⎪⎪⎪⎪⎪
⎨
⎧
−+−=
−+=
−++−−=
−++−−=
)ˆ(ˆˆ
)ˆ(ˆˆ
)ˆ(ˆˆˆ
)ˆ(ˆˆˆ
2
2
1
1
ββαβ
ααβα
ββββββ
αααααα
ω
ω
iiKepdted
iiKepdted
iiKLv
Le
Lir
dtid
iiKLv
Le
Lir
dtid
r
r
sss
s
sss
s
⎩⎨⎧
=+=)()(
)()()(
tCxtytButAxtx&
⎪⎪⎪⎪⎪
⎩
⎪⎪⎪⎪⎪
⎨
⎧
−=
=
+−−=
+−−=
αβ
βα
ββββ
αααα
ω
ω
epdt
de
epdt
deLv
Le
Lir
dtdi
Lv
Le
Lir
dtdi
r
r
sss
s
sss
s
)sin(
)cos(
tppetppe
rrm
rrm
ωωωω
β
α
Φ−=Φ=
⎪⎪⎪
⎩
⎪⎪⎪
⎨
⎧
=
+Φ
+−=
+Φ
−−=
pdt
dLv
tppLL
irdtdi
Lvtpp
LLir
dtdi
rr
srr
s
m
s
s
srr
s
m
s
s
ωθ
ωω
ωω
βββ
ααα
)sin(
)cos(
42 Bodo´s Power Systems® July 2009 www.bodospower.com
Figure 4: Feedback given by the error between estimated and meas-ured currents
Figure 5: The block diagram of the observer equations
M O T I O N C O N T R O L
To evaluate the computational complexity of the two algorithms, the
number of mathematical operations which compose the two schemes
has been taken into account:
As reported in the table, it is possible to see that the number of oper-
ations required by the proposed state observer algorithm is about
double of those required by the VI flux estimator.
However, it has to be pointed out that such an advantage of the VI
flux estimator can not be completely exploited in terms of a smaller
Ts as both the rotor position sensor-less detection algorithms are
often executed with the same sampling rate of the vector control (i.e.
with the same or with an integer sub-multiple of PWM frequency).
Now, let’s evaluate robustness of both of the algorithms to motor
parameters’ uncertainties. Different simulations have been carried out
to measure the error on rotor position angle introduced by a poor
understanding of motor parameters. The figure 6 shows to this pur-
pose the error on rotor position angle when the stator resistance uti-
lized by the algorithm is the double the real one:
The graph in Figure 6 shows a clear and strong sensitivity of VI flux
estimator to stator resistance deviation especially at low speed while
the closed loop approach utilized by the state observer algorithm
allows a full compensation of the stator resistance uncertainty.
On the contrary, as shown in Figure 7, both algorithms are pretty
sensitive to stator inductance uncertainty even if the error introduced
by the state observer is still a bit lower (5% at low speed):
Conclusions
Due to its performance in terms of robustness to motor parameters’
deviation (e.g. stator resistance), the state observer described in this
article had been chosen for implementation on the STM32 microcon-
troller and included in the STM32 FOC PMSM firmware library
The STM32 takes advantage of the powerful CortexTM-M3 computa-
tional capabilities.
In fact, the highly efficient 3-stage pipeline with branch speculation,
the single cycle 32bit multiply, and the hardware divide instructions
allows keeping, in spite of the number of mathematic operations, the
overall state observer routine execution time low - around 3.5μs
(CPU running at 72MHz), thus limiting the CPU load contribution of
the sensor-less algorithm to just 3.5% at 10 kHz PWM frequency.
Moreover, because of the high density of the Thumb2 instruction set,
the code size of the overall state observer routine is quite reduced.
www.st.com
43www.bodospower.com July 2009 Bodo´s Power Systems®
Figure 6; The error on rotor position angle
Figure 7: Both algorithms are pretty sensitive to stator inductance
Mul/Div Add/Sub Total
Flux estimator 6 6 12
State Observer 10 14 24
High Junction Tj: 150C°
17mm height package
High power cycling!
Transfer mold package
Metal Base Plate
SanRex® Semiconductors
New! SanRex Thyristor & Diode Modules Generation!
the"TECHNO-BLOCK´S"
Sansha Electric Japan & SanRex Europe GmbH / Germany introducing:
Thyristor + Thyristor modules
70A up to 110A
up to 1800V
Thyristor + Diode modules
70A up to 110A
up to 1800V
Diode + Diode modules
70A up to 110A
up to 1800V
please contact us!!
more information's: www.sanrex.com
please contact us!!
e-mail: [email protected] GmbH, Dessauerstr. 6, 80992 München /Tel: 089 311 20 34 /Fax: 089 316 16 36
Digital Power Workshop based on TI´s F28x family.
For more information and your free drill hole stencil please visit
www.biricha.com
Biricha Digital Power offering
STM32 FOC PMSM FW library v2.0
State observer execution time 3.5μs
State observer code size 0.5kB
44 Bodo´s Power Systems® July 2009 www.bodospower.com
N E W P R O D U C T S
Mitsubishi Electric introduces a new version
Dual In-line Package Power Factor Correc-
tion (DIPPFCTM) developed for high power
air conditioner and general inverter use.
In the DIPPFCTM series (PS5178x) low
thermal resistance was realized by using an
insulation sheet structure with high heat dis-
sipation transfer capability. As a result, the
thermal resistance was improved by about
35% compared with the present large
DIPPFCTM packaging technology. Because
of low thermal resistance, package size of
the new Mini DIPPFCTM is reduced to
approximately 70% compared with the con-
ventional Large DIPPFCTM and the input
amperage rating is expanded up to 30Arms
under its standard application conditions.
Further extension of the current capability to
above 30Arms will be possible in future by
the development and continuous improve-
ment of thermal interface material properties
and lower power device loss.
The internal circuitry of Mini DIPPFCTM
comprises two IGBT chips which are
designed for high speed switching with a
trench gate structure CSTBTTM and four
diode chips in which high-side diodes are
designed as fast recovery type and the low-
side diodes are designed as low forward
voltage drop types. Moreover, a LVIC is
designed and implemented with necessary
functions such as IGBT drive, control power
supply under voltage (UV) lockout circuit. In
contrast to the conventional PFC circuit
topology using a classical boost structure, in
the new Mini DIPPFCTM, rectifier function
and boost operation have merged: The high-
side diode of the diode bridge hold the func-
tion of boost diode concurrently, and IGBT is
added in parallel with the low-side diode of
the diode bridge.
The result is the new Mini DIPPFCTM series
developed by optimized power chips togeth-
er with the innovative dissipation insulation
sheet.
www.mitsubishichips.com
A Novel Transfer Molding PFC
Everlight Electronics Co. Ltd. (TSE:2393)
announces the introduction of single channel
and dual channel 8-pin SOP phototransistor
optocoupler families offering the best avail-
able combination of key specifications
sought after by designers: low-profile (3.63
mm) package, high isolation voltage (3750
Vrms), minimum creepage distance (5 mm),
wide operating temperature range (-55 °C to
+110 °C), multiple current-transfer-ratio
(CTR) bins and high breakdown voltage (80
V). The single channel EL2XX series is com-
prised of 10 devices while the dual channel
ELD2XX series is comprised of 6 devices.
They are all well suited for use in DC-DC
converters, battery chargers, general-pur-
pose switching circuits and programmable
controllers. Compared with conventional
SOP optocouplers, the devices offer impor-
tant advantages, including: Smaller form-
factor (3.63 mm max. package height and
1.27 mm lead pitch) to accommodate
designs with limited board real estate; High-
er temperature operation (110 °C vs. 100 °C)
to improve system reliability; Selectable CTR
ranges characterized at low LED currents to
increase design flexibility; Higher isolation
voltage (3750 Vrms vs. 2500 Vrms) to
improve insulation characteristics and * Pb-
free and RoHS compliant to meet green
environmental initiatives.
www.everlight.com
8-Pin SOP Optocouplers Offer Best Combination of Features
Fairchild Semiconductor addresses a critical
need in the high brightness (HB) light emit-
ting diodes (LED) market with its primary
side regulation (PSR) pulse-width modula-
tion (PWM) controllers that simplify design,
reduce board space and provide important
performance advantages. The FSEZ1016A
is an EZSWITCH™ that integrates a PSR
PWM controller with a power MOSFET and
the FAN100 is a PSR PWM controller.
Through this integration, these controllers
achieve the most accurate constant current
(CC) through their built-in proprietary TRUE-
CURRENT™ technology and tight constant
voltage (CV) without using secondary-side
feedback circuitry. By tightening the constant
current over a wide voltage range, the same
circuit can accommodate different numbers
of LED units in a string, increasing design
flexibility, accelerating time-to-market and
stretching the lifetime of HB LEDs. With this
high level of integration, these PSR PWM
controllers conserve board space, accommo-
dating the form factor of lamp cases that
continue to diminish in size.
The FSEZ1016A and FAN100 feature a pro-
prietary green-mode function that provides
off-time modulation to linearly decrease the
PWM frequency under light-load conditions.
They also minimize power consumption
(standby power at no load condition <0.15W)
by reducing the secondary-side feedback cir-
cuitry and components.
www.fairchild.com
Primary Side Regulation PWM Controllers
45www.bodospower.com July 2009 Bodo´s Power Systems®
N E W P R O D U C T S
Avago Technologies announced that it has developed a new ultra-low
power optocoupler technology that will pave the way for a new gener-
ation of optical isolators that can operate as much as 90 percent less
power than standard optocouplers available today. The total power
consumption of this new innovative optocoupler design is below 2mA
compared to 15mA for a standard optocoupler and 4 to 5 mA for
magnetic isolators. These new ultra-low power optocouplers target
designers of communication interfaces (RS485, CANBus, and I2C),
microprocessor system interfaces, and digital isolation for A/D, D/A
conversion applications. Avago is a leading supplier of analog inter-
face components for communications, industrial and consumer appli-
cations.
Avago’s innovative ultra-low power
optocoupler design features an opti-
cally coupled gate that combines a
high efficiency light emitting diode
(LED) and an integrated high gain photo detector enabling designers
to interface the optocoupler input directly from a microcontroller out-
put – avoiding the use of buffers to drive the LED. Moreover, these
next-generation optocouplers support both 3.3V and 5V supply volt-
ages and provides reliable system performance in industrial tempera-
tures ranging from –40 to +105 degrees C.
www.avagotech.com/optocouplers
Ultra-Low Power Optocoupler Technology
Ideal for embedded computing, storage, tele-
com, datacom and networking, the ZL2008
Requires up to 50% fewer components and
half the PCB space of competitive solutions
Intersil has introduced the ZL2008, a high-
performance synchronous step-down DC-DC
converter with pin-strap compensation and
current sharing. Taking advantage of Zilker’s
patented Digital-DC™ technology, the
ZL2008 is designed for digital power supply
module and system designers who demand
easy-to-use board-level power supply config-
uration to keep ahead of rapidly changing
market requirements.
Key features and benefits: Easy-to-use pin-
strap compensation and current sharing.
The compensation and transient response of
the ZL2008 can be optimized for load cur-
rent and output capacitance by adjusting
resistor pin-straps. Current sharing of up to
eight devices in parallel with individual phase
enable/disable pins is easily configurable.
Pin-strap power management configuration.
Advanced power management features such
as digital soft-start delay and ramp,
sequencing, tracking and margining are fast
and easy to implement. Power supply relia-
bility and availability can be improved
through real-time monitoring using the
I2C/SMBus interface with the PMBus proto-
col. www.intersil.com
Space-saving Digital Power Management IC
Toshiba Electronics Europe has launched a
new range of power MOSFETs that is opti-
mized for motors used in fans, pumps and
other automotive motion control applications.
The new MOSFETs combine industry-lead-
ing on-resistance and input capacitance rat-
ings with a package design that offers better
thermal dissipation and power cycling capa-
bilities than previous automotive MOSFETs.
Available with maximum voltage (VDSS) and
current (ID) ratings of 60V and 150A, Toshi-
ba’s new MOSFETs are based on the com-
pany’s latest U-MOS trench semiconductor
technology. This technology contributes to
typical RDS(ON) specifications as low as
1.7m? and typical input capacitances (Ciss)
down to just 4500pF. As a result, the devices
offer the industry’s smallest RDS(ON)* gate
charge (Qg) ‘figure of merit’, which ensures
optimum switching speed and efficiency.
All of the new MOSFETs are supplied in
Toshiba’s TO220SM(W) package. This uses
copper connectors and a wide source termi-
nal to drive down RDS(ON) and package
inductance, reduce thermal resistance, and
ensure high current carrying capacity. The
package is qualified to AEC-Q101 at a chan-
nel temperature of 175ºC. A thickness of just
3.7mm means that it is 21% thinner than
existing TO-220SM package technology.
www.toshiba-components.com
Power MOSFETs for Automotive Motion Control
Aware that thermal management is an
important part of environmental care, and
that anything which doesn’t require the use
of electricity in order to maintain acceptable
temperatures, is likely to be welcomed by
design engineers, Ohmite have launched the
first of a range of heatsinks which not only
meet these requirements, but also save time
and money.
Specifically designed for use with TO-220
and TO-247 packages, including Ohmite’s
own TAH20, TBH25, TCH35, TEH70, TK20
and TN15 power resistors, the heatsinks fea-
ture an integrated clip for securing the com-
ponent to the heatsink without the need of a
screw.
Pressure is directed on to the center of the
heat dissipating device, maximising the effi-
cient transfer of heat and avoiding the can-
tilever effect which can result from over
torqueing the screw in conventional designs.
The torsion spring clip also has solderable
feet to secure the assembly to the PCB.
www.ohmite.com
Heatsink Design Enhances Thermal Management
N E W P R O D U C T S
46 Bodo´s Power Systems® July 2009 www.bodospower.com
These devices prevent damages caused by
transient overvoltages in industrial electrical
equipment and installations.
These damages can be disruptive, dissipa-
tive or destructive whatever their origin:
power or equipment disturbances, switching
on and off, lightning strikes, utility grid
switching.
The core of Surge Trap® is a TPMOV®:
Thermally Protected Metal Oxide Varistor.
TPMOV® is a patented technology devel-
oped by Ferraz Shawmut. It leads to fail-safe
surge protective devices needing no fuse for
overcurrent protection.
Surge Trap® are: Easy to install, they are
DIN-rail mountable, easy to retrofit: when
using the Pluggable model the user has only
to replace the plug unit without removing the
base. Pluggable is the premium offering in
the Surge Trap® line when Modular is an
economic alternative,
Global: with UL and IEC models and for pho-
tovoltaic applications as well, communica-
tive: a front visual indicator communicates
the status of the TPMOV®. Remote signal-
ing is an option via an auxiliary microswitch.
Surge Trap® is a cost-effective solution
against overvoltage transients requiring no
fuse holder and no additional wiring. It pro-
vides substantial cost savings and reduced
installation time as well.
With Pluggable and Modular models Surge
Trap® is a flexible high-tech protection solu-
tion.
www.ferrazshawmut.com
Fail-Safe Surge Protective Devices
LEM has introduced the ITL 4000-S current
transducer for non-invasive measurement of
currents up to 4000ARMS in conductors of
up to 268mm diameter. The new transducer
allows the isolated measurement of AC, DC
and pulsed currents, up to three times the
nominal value for peak measurement at fre-
quencies up to 50kHz (+/-1dB).
Using closed-loop Fluxgate technology, high-
ly accurate measurements of +/-0.1% of IPN
are achieved over the operating temperature
range from -40°C to +70°C. This high level
of accuracy also allows the measurement of
small DC currents in the presence of large
AC components, which is particularly useful
in applications such as transformer protec-
tion. For example, it is possible to monitor
+/-10A DC over a 4000ARMS AC current
with an uncertainty of +/-1A over an operat-
ing temperature range from -25°C to +50°C.
The technology also offers very good offset
and gain thermal drift performance.
The large aperture of the ITL 4000-S makes
it particularly suitable for measurements on
high-voltage DC systems, which use large-
diameter cables. It features high insulation
for working voltages up to 1.5kVRMS in
accordance with the EN 50178 standard.
www.lem.com
Current Transducer Accurately Measures up to 4000A
Since the first announcement of their colla-
boration Fuji is now coming up with MiniSKi-
iP-PIM-Modules from 8 to 100A and a blok-
king voltage of 1200V.
At the same time Fuji has started the mass-
production of the newest IGBT-generation
(V-series) applying FS-Trench-Technologies
and providing lower total-losses and impro-
ved EMI-characteristics at the same time.
V-series IGBTs will be used for the complete
range of Fuji Mini-SKiiP-Modules.
www.Semikron.com
www.fujielectric.de
Fuji to team-up with Semikron for MiniSKiiP®Spring-contact Power Modules
In today’s world and for the future environ-
mentally friendly technologies is a necessity!
Consumer and manufacturers have turned
their attention towards energy saving elec-
tronic devices.
Various semiconductor manufacturers
already offer simple ICs with which competi-
tive SMPS can be designed. Würth Elektron-
ik has developed two transformer series,
which are designed and manufactured to the
requirements of the leading semiconductor
manufacturers STMicroelectronics and
Power Integrations. The offline transformer
WE- UNIT has an input voltage of 85 – 265
VAC and an isolation voltage of 4 kVAC. The
transformers are especially designed for uni-
versal input as well as for offline-switch
mode power supply.
An additional requirement to decrease the
energy consumption will be released by the
law. One of these requirements results in the
disappearance of the linear regulators in the
near future. They will be replaced by SMPS.
The aim of manufacturing SMPS’ is the
development of power supply for the world-
wide universal input. The advantages are
obvious: Efficiencies of 80% and more, less
weight and smaller size than linear regula-
tors as well as a low stand by-power con-
sumption.
www.we-online.com
Transformer for Energy Saving Electronic Devices
C O N T E N T S
SynQor announces the addition of a new
product family to its portfolio of Hi-Rel dc/dc
converters, the MQFL-28E series. Incorpo-
rating SynQor’s field proven high-efficiency
synchronous rectifier technology, this
advancement results in the only wide range
input dc/dc converters developed specifically
for the Military/Aerospace industry.
The MQFL-28E family handles up to 120W
while meeting the long-term over-voltage
surges in the input voltage specified by MIL
STD 704(A-F), RTCA/DO-160E, and DEF
STAN 61-5. The MQFL-28E converters
accept continuous input voltages of 16-70 V
while permitting broader transient input volt-
ages of 16-80V for 1 second.
The MQFL-28E series is available in eleven
standard single output voltages including
1.5V, 1.8V, 2.5V, 3.3V, 5V, 6V, 7.5V, 9V, 12V,
15V and 28V. Dual output converters are
also available with output voltages of +/-5V,
+/-12V and +/-15V with the capability of pro-
viding up to 80% power from either output.
These feature-rich MilQor Hi-Rel MQFL-28E
converters operate at efficiencies up to 91%
and are designed to operate from -55°C to
+125°C with multiple screening options for
stringent environmental requirements.
www.synqor.com
DC/DC Converters & Filters for Military Applications
The 5th IET International Conference on
Power Electronics, Machines and Drives
PEMD 201019-21 April 2010 | Thistle Hotel | Brighton | UK
www.theiet.org/pemd
Call for Papers
18 September 2009 Abstract Submission
27 November 2009 Notification of Acceptance
29 January 2010 Submission of Final Papers
19-21 April 2010 Date of Conference
Key Deadlines
Join a stellar list of international engineers from industry, research and
academia to discuss, debate and learn about recent developments and
future trends in this important and fast changing area.
In 2010 the conference is expanding its focus to embrace the challenges
and solutions of the applications and systems in which power
electronics, motor and drive technologies play a critical part.
Attendees at PEMD 2010, established as a major forum to showcase the
latest advances in power technology, will gain valuable insights into the
technology roadmaps of the materials and components that are integral
to driving innovation.
For a full list of topics and to submit
your paper, please visit the web site
Conference Themes:
� More/All Electric Transport
� Generation, Transmission and Distribution
� Machines and Drives
� Power Electronics
� Renewable Energy Systems
Supported by: Exhibitors: Media Partners:
IXYS Corporation
(NASDAQ: IXYS)
introduces new 170V-
300V GigaMOSTM
Power MOSFETs.
These power MOS-
FETs provide high cur-
rent capability (up to
260A), eliminating the
need for multiple components when parallel-
ing lower current MOSFET devices in high
power applications. The resultant effect is a
reduction in part count, as well as the num-
ber of required drive components, improving
over-all system reliability and simplicity.
These power MOSFETs are optimized for
superior switching performance in a broad
range of high power switching applications.
GigaMOSTM Power MOSFETs incorporate
IXYS’ Trench Technology to achieve low
Rds(on) and gate charge (Qg), while main-
taining superior switching performance and
ruggedness. Power switching capability is
further enhanced by IXYS’ proven HiPer-
FETTM process yielding a fast intrinsic recti-
fier which provides low reverse recovery
charge (Qrr) and excellent commutating
dV/dt ratings. Additional features include a
175 degree Centigrade operating tempera-
ture and avalanche capabilities. These com-
bined product attributes coupled with high
current ratings, make for an ideal device for
high current power switching applications.
The high current capability of these devices
make them suitable for electric and hybrid
car and carts and other high power battery
powered electrical equipment and tools.
http://www.ixys.com
GigaMOSTM Power MOSFETs
48 Bodo´s Power Systems® July 2009 www.bodospower.com
N E W P R O D U C T S
APEC 2010 21
Bicker Elektonik 25
CT-Concepts 3+C3
Danfoss 15
epe 33
Intersil 7
IR C4
ITPR 17
IXYS 31
LEM 1
PEMD 2010 47
Powersem 9
Sanrex 43
Semicon 23
Semikron C2
SPS/ICP/DRIVES 13
Würth Elektronik 19
ADVERTISING INDEX
We believe that solar power represents a
future for renewable energy. PV inverters
become more and more compact exposing
sensors to magnetic disturbance. In order to
offer a more efficient sensor for this applica-
tion, we have optimised the ESM range in
terms of magnetic immunity and dynamic
response.
Thanks to a wide range, from 500A to 2000A
and the possibility to customize the sensor in
function of your needs, the ESM sensor is
our high performance range for inverters.
ESM sensors have been developed in accor-
dance with the EN50178 industrial standards
and complies the EN61000-6-2 and
EN61000-6-4 standards regarding electro-
magnetic compatibility.
For all these good reasons, world’s solar
inverter manufacturers place their trust in the
performance, reliability and compliance of
ABB’s sensors.
ABB takes protection of the environment
very seriously. It is a priority for all ABB com-
panies and not least for us at ABB’s Sensors
products line, we are proud of our ISO
14001 certification.
www.abb.com
Sensors Improve the Performance of Inverters
TEWS TECHNOLOGIES announced the
TPMC317, a conduction cooled single-width
32-bit PMC module which offers six inde-
pendent channels operating as either a stan-
dard SSI interface controller, a SSI ‘Listen
only’ device, an incremental encoder, or a
general purpose counter.
The standard SSI interface controller pro-
vides a clock burst output to the absolute
encoder and receives the returned positional
data. The SSI interface controller operates
with a programmable clock rate from 1μs to
15μs and programmable data word length
from 1-bit to 32-bit.
In ‘Listen only’ Mode, the channel listens to
an existing SSI interface to observe its data
transfer using both the SSI clock and data
as inputs. The channel also has a program-
mable data word length from 1-bit to 32-bit,
and the SSI clock rate of the observed SSI
interface can be in the range of 1μs to 15μs.
In both modes the data word can be encod-
ed in Binary- or in Gray code and with odd,
even or no parity.
The 32-bit incremental encoder counter is a
preloadable up-and-down counter. The
counter is programmable for single, double
and quadruple analysis of the encoder sig-
nals. In conjunction with the isolated 24V
digital inputs it provides the possibility of
automatic preload of the counter whenever
the motion system passes a reference posi-
tion.
www.tews.com
6 Channel SSI Incremental Encoder Counter for Motion Control Applications
Tyco Electronics has announced the addition
of a 10A version of its Corcom GG series
general-purpose and HG series medical fil-
tered power-entry modules. A 5-inch (127
mm) wire lead option has also been intro-
duced for the HG medical modules to aid in
module installation. Previously, the lines
were offered in ratings of 1, 3, and 6 amps.
The GG series of power entry modules com-
bines the functions of a general-purpose RFI
filter with an IEC power cord connector and
single or dual metric fuse holder. The HG
series is the medical version of the GG
series with reduced line-to-ground capaci-
tance to meet UL 2601 patient care require-
ments. Both series are compact - they are
Tyco Electronics’ smallest filtered power-
entry modules with metric fuse holders.
The GG and HG series of power-entry mod-
ules are available with current ratings up to
10A at 250 VAC, 50/60 Hz. With the addition
of the lead wire option to the HG series, both
the GG and HG series are available with
either industry-standard .250-inch (6.35 mm)
tab terminals or 5-inch lead wires for load-
side terminations.
These modules are UL recognized, CSA cer-
tified, and VDE approved. Both GG and HG
series power-entry modules find wide use in
communications, computer and consumer
electronics, industrial, commercial, instru-
mentation and medical equipment.
www.tycoelectronics.com
Expanded Line of Power-Entry Modules
SAMPLES AVAILABLE!
CT-Concept Technologie AG, Renferstrasse 15, CH-2504 Biel, Switzerland, Phone +41-32-344 47 47 www.IGBT-Driver.com
High Frequency
Artists!
Features350kHz max. switching frequency
±1ns jitter
+15V/-10V gate voltage
20W output power
60A gate drive current
80ns delay time
3.3V to 15V logic compatible
Integrated DC/DC converter
Power supply monitoring
Electrical isolation for 1700V IGBTs
Short-circuit protection
Fast failure feedback
Superior EMC
The 1SC2060 is a new, powerful member of the CONCEPT
driver core family.The introduction of planar transformer tech-
nology for gate drivers allows a leap forward in power den-
sity, noise immunity and reliability. Equipped with the latest
SCALE-2 chipset, this gate driver supports switching frequen-
cies up to 350kHz with best-in-class efficiency. It is suited
for high-power IGBTs and MOSFETs with blocking voltages up
to 1700V. Let this artist perform in your high-frequency or
high-power applications.
1SC2060 Gate Driver
Part Number Package VOFFSET VOUT
IO+ & IO- (typical)
tON & tOFF
(typical)
AUIRS2123S SOIC8 600V 10V - 20V 500mA 140 ns & 140 ns
AUIRS2124S SOIC8 600V 10V - 20V 500mA 140 ns & 140 ns
The AUIRS212xS family of 600V, single
channel high-side driver ICs for low-,
mid-, and high-voltage automotive
applications features exceptional
negative Vs immunity to deliver the
ruggedness and reliability essential for
harsh environments and automotive
under-the-hood applications.
Features
• Designed and characterized to be
tolerant to repetitive Vs transient
voltage
• Fully operational up to 600V
• Tolerant to large dV/ dt
• Under voltage lockout
• Lead-free, RoHS compliant
• Automotive qualified per AEC-Q100
t
VS Undershoot
VS -COM
-VS
VBUS
Greater protectionagainst a “negative Vs” event
t
Rugged, ReliableAutomotive-Qualified 600V ICs
THE POWER MANAGEMENT LEADER
For more information call +33 (0) 1 64 86 49 53 or +49 (0) 6102 884 311or visit us at www.irf.com
10388AD_AUIR2123_BODOS_v1.indd 1 28/01/2009 16:40