Electronics in Motion and Conversion November 2016

Electronics in Motion and Conversion November 2016
ISSN: 1863-5598
Electronics in Motion and Conversion
ZKZ 64717
November 2016
Power is in our nature. Everyday we deliver full power for your success.
Quick design-to-product using state-of-the-art technologies
Extensive production experience with maximum flexibility
GvA SOLUTIONS: Short time-to-market with innovative plug&play system solutions
Vast product knowledge and consulting expertise
GvA Leistungselektronik GmbH
Boehringer Straße 10 - 12
D-68307 Mannheim
Phone +49 (0) 621/7 89 92-0
[email protected]
Viewpoint ........................................................................................... 4
Novemberfest in Munich
Events ................................................................................................ 4
News ............................................................................................. 6-14
Blue Product of the Month ............................................................. 16
Dual Channel “Smart” Power Amplifier Uses Mixed-Signal Processing
By APEX Microtechnology
Guest Editorial ................................................................................ 18
Magnetics: The Eternal Battle with Myths, Mysteries and Black Magic
– or Maybe Really all Just a Matter of Basic Principles and Physics?
By Alexander Gerfer, CTO Würth Elektronik eiSos
VIP Interview .............................................................................. 20-21
Electronics with Diamonds
By Henning Wriedt, Bodo’s Power Systems, corresponding editor
Cover Story ................................................................................ 22-27
Peak Current Proof Input Filter with Multilayer Power Suppression
By Ranjith Bramanpalli, Würth Elektronik eiSOos
IGBTs .......................................................................................... 28-31
Control Method for a Reverse Conducting IGBT
By Daniel Domes, Infineon Technologies AG
IGBT Modules ............................................................................ 32-34
Innovative 7in1 IGBT Packages for Scalable and Easy Design of
Industrial Drives and Inverters
By Thomas Radke and Narender Lakshmanan,
Mitsubishi Electric Europe B.V.
Technology ................................................................................. 36-41
By Dr. Vladimir Scarpa, Field Application Engineer Power,
ROHM Semiconductor GmbH
High Power Switch .................................................................... 42-44
Reverse Conducting IGCT Platform optimized for
Modular Multilevel Converters (MMC)
By Tobias Wikström, ABB Switzerland Ltd - Semiconductors
MOSFETs .................................................................................... 46-48
Technology Trends Raising Power-Conversion Efficiency
By Michael Piela, Toshiba Electronics Europe GmbH
Measurement ............................................................................. 50-55
Identification of PMSM Motor Parameters with a Power Analyzer
By Kunihisa Kubota, Hajime Yoda, Hiroki Kobayashi,
and Shinya Takiguchi; HIOKI E.E. Corporation
Automotive Power ..................................................................... 56-59
Igniting the Spark
By Ashutosh Tiwari, Shailendra Vengurlekar and Namrata Dalvi,
Microchip Technology
Power Management ................................................................... 60-62
A Fast Track to Complex Power System Designs
By Arthur Jordan - Applications Engineer, Vicor Corporation
Motion Control ........................................................................... 64-65
Embedded Designs Drive Tomorrow’s Solutions
By Michele Portico, Product Marketing Manager, Vincotech GmbH
Magnetic Components .............................................................. 66-68
Innovative Integrated Magnetics for Hybrid and
Electrical Vehicles Onboard Battery Chargers
By Patrick Fouassier, PhD.-Eng.,
Inductive Components R&D Manager, PREMO Group
DC-DC Converter ....................................................................... 70-83
The Flyback DC-DC Converter: The Best Bet for Most SMPS
By Dr.-Ing. Artur Seibt, Vienna
New Products............................................................................. 84-96
㔀䴀倀䄀 匀攀爀椀攀猀
䈀攀猀琀 䌀栀漀椀挀攀 昀漀爀 䠀椀最栀 䌀甀爀爀攀渀琀 䌀愀爀爀礀椀渀最 䄀䌀 䄀瀀瀀氀椀挀愀琀椀漀渀
㔀䴀倀䄀 䘀䔀䄀吀唀刀䔀匀
䰀漀眀 氀漀猀猀Ⰰ 搀爀礀 ǻ氀洀 挀漀渀猀琀爀甀挀琀椀漀渀
伀瀀攀爀愀琀椀渀最 琀攀洀瀀攀爀愀琀甀爀攀 爀愀渀最攀㨀
ⴀ㔀㔀뀀䌀 琀漀 ⬀㠀㔀뀀䌀
䄀瘀愀椀氀愀戀氀攀 漀昀昀 琀栀攀 猀栀攀氀昀
嘀椀猀椀琀 甀猀 漀渀氀椀渀攀 愀琀 眀眀眀⸀攀挀椀挀愀瀀猀⸀挀漀洀
䌀漀渀琀愀挀琀 唀猀
一漀爀琀栀 䄀洀攀爀椀挀愀㨀 猀愀氀攀猀䀀攀挀椀挀愀瀀猀⸀挀漀洀
䔀甀爀漀瀀攀㨀 猀愀氀攀猀䀀攀挀椀挀愀瀀猀⸀椀攀
The Gallery
Bodo´s Power Systems®
November 2016
HIOKI @electronica2016
Booth #179, Hall A1
[email protected]
Improve Power Conversion Efficiency and Minimize Loss
Enhance the Development of SiC, GaN and IGBT Inverters
Diverse array of sensors from 10mA to 1000A
6CH per unit, 12CH when synchronizing 2 power analyzers
±0.02% rdg. basic accuracy for power
5MS/s sampling and 18-bit A/D resolution
DC, 0.1Hz to 2MHz bandwidth
CMRR performance of 80dB/100kHz
Completely simultaneous
digital processing
Accurate period detection
Wideband power analysis
(MATLAB is a registered trademark of Mathworks Inc.)
A /D
FFT analysis up to 2MHz
Compensate current sensor phase error with 0.01° resolution
Harmonic analysis up to 1.5 MHz
Dual motor analysis
Large capacity waveform storage up to 1MWord x 6CH
MATLAB toolkit support
Digital AAF
Assess efficiencyand loss
at a glance
Analyze power waveforms
without an oscilloscope
Harmonic analysis -critical
for linking systems
FFT analysis oftarget waveforms
Harmonic analysis
Waveform analysis
Digital AAF
FFT analysis
Fast, simultaneous calculations with Power Analysis Engine II
A Media
Katzbek 17a
D-24235 Laboe, Germany
Phone:+49 4343 42 17 90
Fax: +49 4343 42 17 89
[email protected]
Publishing Editor
Bodo Arlt, Dipl.-Ing.
[email protected]
Junior Editor
Holger Moscheik
Phone + 49 4343 428 5017
[email protected]
Senior Editor
Donald E. Burke, BSEE, Dr. Sc(hc)
[email protected]
UK Support
June Hulme
Phone: +44(0) 1270 872315
[email protected]
Creative Direction & Production
Repro Studio Peschke
[email protected]
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 €
[email protected]
print run 24 000
Printing by:
Brühlsche Universitätsdruckerei GmbH
& Co KG; 35396 Gießen, 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
Toulouse, France November 2-4
Novemberfest in Munich,
Those that missed Oktoberfest in Munich can
now come to the electronica - the Novemberfest for electronics. The whole world of
electronics comes to Munich bi-annually for
the electronica, a big party ! For almost a
full week, Munich is at the center of world of
electronics. Two weeks later, in Nuremberg,
the SPS IPC Drives will take place. November will be a busy time.
As I write this, I am surprising a lot of my industry friends with visits along the East Coast
of the USA, and up into the New England
area. Manhattan left a strong impression,
as I had once been up in the World Trade
Center, looking out over the city. This year I
was in the Freedom Tower, overwhelmed by
how fast buildings have replaced the wounds
at Ground Zero. New York is always worth
a visit, if for nothing else, just to take in its
vibrant life and activity. But it is especially
moving to look down at the Statue of Liberty.
It was a gift to Americans from the French
people. A good opportunity to take a rest,
think about freedom, and the Thanksgiving
holiday soon to come. Each of us can form
our worlds and work for peace and freedom
– you will have my full support.
Respect for people is so important and it
seems surprising to have a candidate for
President denigrating so many people including parents who have lost their son in war.
It does not matter what religion the parents
have. Their son gave his life for the USA and
for freedom! My wish is that the voters will be
mature enough to recognize trash-talking for
what it is, and opt for a mature, experienced
lady. In Germany we have Chancellor Angela
Merkel and it would be very good for the
USA to choose Hillary Clinton for the next
no longer an issue, as reductions show up
at the system level. SiC and GaN are more
efficient in all applications, with transportation
and computer systems leading the way. We
are moving towards more electric automobiles. Power is being produced by green
resources; water, wind or solar. Coal, gas
and nuclear will be the past, as pollution and
efficiency become paramount.
I now look forward to seeing you at the
electronica or sps ipc drives. Bodo’s Power
Systems reaches readers across the globe.
If you are using any kind of tablet or smart
phone, you will now find all of our content on
the new website www.eepower.com.
If you speak the language, or just want to
have a look, don’t miss our Chinese version:
My Green Power Tip for November:
Turning off every street light during daylight
should be possible. If you observe one not
responding, have a talk to the town officials.
That will save a lot of energy.
Best Regards
Everywhere I visit, innovation is moving
fast. Wide band gap devices are taking over
critical design activities for systems. Cost is
Power Integrity Seminar 2016,
Munich Germany, November 3,
electronica 2016,
Munich, Germany, November 8-11
sps ipc drives 2016,
Nuremberg, Germany, November 22-26
Bodo´s Power Systems®
November 2016
S/ ves 20
SP Dri A.
LF xx10
Current transducer range
Pushing Hall effect technology to new limits
To save energy, you first need to measure it! To maximise energy savings, you need to
measure the current used accurately!
By using the most advanced materials available, LEM’s new LF xx10 transducer range
breaks new ground in accuracy for Closed Loop Hall effect transducer performance.
LEM ASIC technology brings Closed Loop Hall effect transducer performance to the level of
Fluxgate transducers and provides better control and increased system efficiency, but at a
significantly lower price.
Available in 5 different sizes to work with nominal currents from 100 A to 2000 A, the LF xx10
range provides up to 5 times better global accuracy over their operating temperature range
compared to the previous generation of Closed Loop Hall effect current transducers.
Quite simply, the LF xx10 range goes beyond what were previously thought of as the limits of
Hall effect technology.
• Overall accuracy over temperature range
from 0.2 to 0.6 % of IPN
• Exceptional offset drift of 0.1 % of IPN
• Fast response time less than 0.5 μs
• Higher measuring range
• 5 compact sizes in a variety of mounting
topologies (flat or vertical)
• Immunity from external fields for your
compact design
• 100 % fully compatible vs LEM previous
• -40 to +85 °C operation
At the heart of power electronics.
EPE ECCE Call for Papers
The Power Electronics community will gather in Warsaw, Poland, from
11 to 14 September 2017, to exchange views on research progresses
and technological developments in the various topics described hereunder. On Monday 11 September a number of tutorials will be organised and several technical visits are planned on Friday 15 September.
The 19th Conference on Power Electronics and Applications (and
Exhibition), EPE’17 ECCE (Energy Conversion Congress and Expo)
Europe is co-sponsored by the EPE Association and IEEE PELS. It
will take place at the Gromada Airport Hotel and Conference Centre.
EPE ECCE Europe is the place for specialists in power electronics,
systems and components, to present papers and attend sessions
on state-of-the-art technology in this challenging and evolutionary
sector. The conference aims to be a meeting forum for researchers,
developers and specialists from academia and industry. Papers are
encouraged on all topics described hereunder for interdisciplinary
discussions of new ideas, research, development, applications and
the latest advances in the field of power electronics and adjustable
speed drives.
Especially, wide bandgap materials (SiC & GaN) are a serious chance
to big step forward in broad range of applications.
Several tutorials will be held prior to the conference. Authors willing to
propose a tutorial at EPE‘17 ECCE Europe are invited to send a proposal to Brigitte Sneyers at the scientific secretariat (EPE Association,
c/o VUB-IrW-ETEC, Pleinlaan 2, B-1050 Brussels, Belgium, e-mail:
[email protected]) before 15 January, 2017.
The Future of Mobility at lectronica 2016:
From Data Security to Autonomous Driving
In the future, mobility will be shaped by developments in automotive electronics: Smart lighting, autonomous driving and connected
vehicles—to name just a few catchwords—are closely related to electronic components and software developed around the world now and
in the future. electronica, the World’s Leading Trade Fair for Electronic
Components, Systems and Applications to be held in Munich from
November 8–11, 2016, will shed light on the latest developments.
to the German Electrical and Electronic Manufacturers’ Association
(ZVEI), global demand for semiconductors for automotive electronics was worth nearly 35 billion dollars in 2014. Today, electrical and
electronic components account for 30 percent of an automobile’s production value—with an upward trend. The ZVEI expects the industry
to grow at a rate of 4.5 percent over the next five years. No wonder,
because according to the association, 80 percent of innovations in automobile manufacturing are driven by microelectronics and software.*
Expert knowledge at the electronica Automotive Conference
“Automotive” will be a focal point of electronica again in 2016. More
than 800 exhibitors from this sector have already registered for the
fair. All in all, more than 2,800 companies will present their products
and services at the fair. On November 7, the day before the fair
begins, leading managers and experts will meet at the electronica
Automotive Conference to discuss the industry’s key issues. The
main topics of this year’s lectures are safety, automated driving and
interior electronics. Steve Nadig from Daimler Trucks North America
has agreed to give a keynote address on “Autonomous Trucks: A
Global Perspective.” In another keynote address titled “Automotive
meets CE,” Dr. Ludger Laufenberg (Kostal) will explain how autonomous driving is effecting interior electronics. And Dr. Reinhard Ploss
(Infineon Technologies) will examine the current role of electronics in
the automotive industry in a keynote titled “Semiconductors as a key
enabler for the transition of the automotive industry.”
Experts will also discuss the industry’s latest challenges and developments at the Automotive Conference and the Automotive Forum.
Automotive electronics is on the rise around the world: According
Redundant Power Supplies using GaN FETs
Telcodium, a leader in power supply design, in collaboration with
Transphorm Inc. released the industry’s first redundant power supplies using gallium nitride (GaN) field-effect transistors (FETs).
Telcodium’s AC Series replaces a typical three-module power supply architecture (two power supply bricks and one intermediate bus
converter (IBC)) with a single power module with redundant AC feeds.
Telcodium’s power module operates at 94 percent True System Efficiency (TSE)* or higher—reducing average energy loss by 13 percent
or more. To achieve the same TSE with the typical three-module
power supply, the bricks and IBC would each need to yield a 97 per-
Bodo´s Power Systems®
cent efficiency—which exceeds the 80Plus Titanium specification and
has yet to be demonstrated by any power supply manufacturer.
Further, the new module is 30 percent smaller than the abovementioned two bricks and eliminates the standalone IBC—freeing considerable, critical space inside a host system.
November 2016
SCALE-iDriver™ Family
Single-Channel Gate Driver ICs with Reinforced Isolation
• 2.5 A, 5 A, 8 A output
current devices
• Drives 1200 V IGBTs and
MOSFETs up to 110 kW
• Revolutionary FluxLink™
communication technology
• Low profile with 9.5 mm
creepage and reinforced
igbt drivers
20 Years of Effective Collaboration
Toshiba Electronics Europe and EBV Elektronik, a leading European
distributor of semiconductors, are celebrating the 20th anniversary
of their highly successful working relationship. In this time, collaboration between the two companies has grown from strength to strength
and now sees EBV as one of Toshiba’s most important world-wide
distribution partners.
The association, which began on the 1st September 1996, originally covered selected countries in central Europe. However, after
a successful, initial five months it was expanded to cover the whole
of Europe. This partnership quickly saw yearly sales figures surpass
$20 million, with a particular focus on couplers and MCU/Echelon
interception systems.
Over the last 20 years EBV has been quick to introduce their customers to products and technologies pioneered by Toshiba, such as
SRAMs and NAND Flash memories, backed up with joint customer
visits and presentations. This has been continued with the latest
ranges of high-end photocouplers and MOSFETs, where opportunities for collaborative development and customer feedback have been
taken by both parties to improve the experience of the end-user.
Picture: Klaus Michel, General Manager Distribution Sales, Toshiba
& Slobodan Puljarevic, President EBV Elektronik
Promoting New Digital Age in India
STMicroelectronics, a global semiconductor leader serving customers across the spectrum of electronics applications, has showcased
its latest technologies and products for Smart Driving, Smart Cities
and Homes, and Smart Things at electronica India in the Bangalore
International Exhibition Center.
ST’s Smart Driving solutions focus on increasing safety and security
with Assisted Driving, enhancing engine efficiency and accelerating
vehicle electrification with Green Driving, and improving comfort and
convenience with telematics and infotainment in Connected Driving.
Similarly, ST’s offering for Smart Cities and Homes addresses energy
efficiency along the entire power lifecycle through high-efficiency power supplies and energy-management technologies, smart (LED) lighting, and home automation. ST also provides all the building blocks
for applications that make up the vast universe of Smart Things in our
homes, offices, streets, and cars.
Non-Isolated Digital Point-of-Load Standard for 60 A Current
The Architects of Modern Power® (AMP Group) consortium announced an additional standard aimed at establishing common mechanical and electrical specifications for the development of advanced
power conversion technology for distributed power systems. The
‘gigaAMP™’ standard, introduced to provide a higher current option
in a land-grid array (LGA) footprint, builds on the previously-released
‘picoAMP™’ standard, published in September 2015, which defined
standards for a lower range of non-isolated platforms ranging from 6 A
to 18 A. The 60 A ‘gigaAMP’ standard defines a compact footprint of
25.1 x 14.1 mm in an LGA format.
The new ‘gigaAMP’ standard also adds to the previously released
‘teraAMP™’ standard for non-isolated digital point-of-load (POL)
dc-dc converters, released in February 2015, and the ‘microAMP™’
and ‘megaAMP™’ standards released during electronica in November
2014. The first products to meet with this new ‘gigaAMP’ standard will
be announced by AMP Group® members later in the year.
Completing Acquisition of Fairchild Semiconductor
ON Semiconductor Corporation and Fairchild Semiconductor International, Inc. jointly announced that ON Semiconductor has completed
its announced acquisition of Fairchild.
On September 16, 2016, ON Semiconductor received confirmation
that clearance related to the completion of its proposed acquisition of
Fairchild from the Ministry of Commerce in the People’s Republic of
China had been obtained and that ON Semiconductor was entitled
to close the transactions under PRC law. As such, the conditions to
the acquisition of Fairchild relating to the termination or expiration of
required waiting periods, and receipt of required approvals, under
applicable antitrust laws were fully satisfied and ON Semiconductor’s
Bodo´s Power Systems®
tender offer to purchase all of the outstanding shares of common
stock of Fairchild for $20.00 per share in cash (the “Offer”) expired as
scheduled one minute following 11:59 p.m., New York City time, on
September 16, 2016 and was not extended.
As a result of the Offer and the merger, Fairchild ceased to be a
publicly traded company, its common stock will no longer be listed on
NASDAQ, and Fairchild became a wholly owned subsidiary of ON
November 2016
Allegro Stepper Motor Driver ICs
The Simplest Microstepping Control In The Industry
Allegro MicroSystems offers a full line of stepper motor driver and
pre-driver ICs. These devices feature easy to use two wire step and direction
translator interfaces as well as industry standard parallel or serial control.
• Microstepping
• Parallel, serial and
step/direction interfaces
• PWM current control
• Robust protection features
• Diagnostic outputs
• Stall detect assistance
• Package options include exposed
pad QFN, TSSOP, SOIC, and DIP.
All packages are (Pb) free.
Applications include:
Office Automation
- Printers
- Scanners
- Copiers
- 3D Printing
- Throttle control
- Transmission
- Headlamp leveling
- Vent position control
- Sewing machines
- Closed circuit television
- CNC milling machines
- Ticketing
- Vending
- Robotics
94616 Rungis Cedex, FRANCE
Tel: +33 (0) 1 56 70 03 80
E-mail: [email protected]
Allegro MicroSystems Germany GmbH
Adlerweg 1, D-79856 Hinterzarten, GERMANY
Phone: +49-(0)7652-9106-0
Fax: +49-(0)7652-767
E-mail: [email protected]
Consystem S.r.l.
I-20144 Milano, ITALY
Tel: +39 02 4241471
Website: www.consystem.it
E-mail: [email protected]
Maximize Productivity for Industry 4.0 Applications
Significantly increase manufacturing productivity with the Pocket
IO programmable logic controller (PLC) development platform from
Maxim Integrated Products, Inc.. The platform provides customers
with the ability to achieve the smallest form factor and highest power
efficiency for next-generation PLC designs.
Lost productivity is a common concern for Industry 4.0 designers
challenged with keeping a manufacturing line running 24 hours a day,
7 days a week. Without intelligent data available at their fingertips,
factory operators do not have the insight to make informed, real-time
decisions which can significantly improve uptime, revenue, and gross
margins. In addition to capturing real-time data, PLCs require fan-less
operation due to harsh industrial environments. As a result, highly efficient power solutions are required to minimize heat dissipation.
Maxim’s Pocket IO PLC development platform redefines how factories
operate and enables Industry 4.0 applications. To maximize productivity, it provides real-time intelligence to quickly and effectively make
decisions, adaptive manufacturing to avoid potential downtime, and
distributed control to provide redundancy. The Pocket IO provides the
following key advantages to increase productivity:
• Real-time intelligence: Fast data processing provides the necessary data to make intelligent decisions quickly and effectively to
optimize yield.
• Adaptive manufacturing: Manufacturing flexibility allows for realtime changes and adjustments to avoid potential downtime.
• Distributed control: Ultra-small footprint of less than 10 cubic
inches and smart energy consumption brings PLC down to the
manufacturing line, re-distributing intelligent control and providing
When compared to the Micro PLC Platform from two years ago, the
Pocket IO decreases form factor by an additional 2.5x and reduces
power consumption by another 30%.
Visit Maxim at electronica 2016 in Munich Hall A4, Stand 279.
High-Performance Embedded Design with STM32F7 MCU Lines
STMicroelectronics has introduced STM32F7 microcontroller lines
and added accessories and options to the development ecosystem,
easing access to high-performance embedded design based on the
ARM® Cortex®-M7 core.
The latest STM32F722 and STM32F723 microcontrollers in the
very high-performance STM32F7 series reduce memory footprint by
integrating value-added features including code-execution protection and high-speed USB physical-layer (PHY) circuitry that streamline development of connected applications. The STM32F732 and
STM32F733 variants come with extra cryptographic features on-chip,
such as an efficient AES256 HW engine. There are versatile package
options from a 64-pin LQFP up to 176-pin LQFP or UFBGA for projects demanding high I/O count, and 256KB or 512KB of on-chip Flash
memory with 256KB RAM.
Extensive pin, package, and software compatibility between the
new devices and higher-end STM32F7 variants with 256KB to 2MB
Flash memory and 256KB to 512KB RAM in packages up to 216-pin
TFBGA simplifies design scaling for product differentiation and futureproofing. STM32F7 projects can also be readily ported throughout the
large STM32 microcontroller family, which contains over 700 devices
that cover all 32-bit Cortex-M cores and offer a wide range of peripheral, pin-count, and memory options.
World’s Biggest Trade Fair for the Electronics Industry
electronica 2016 - Munich - 8 to 11 November, once again, electronica
will bring the global electronics industry together in Munich for the
World’s largest trade fair dedicated to every aspect of electronics
technology. electronica 2016 will offer visitors the chance to see all
the major players from key industry sectors including semiconductors,
power, interconnection, embedded and test and measurement. With
the addition of an extra hall, electronica 2016 will have 13 halls full of
innovations as well as an extensive programme of conferences and
supporting events to provide visitors with the chance to learn more
Bodo´s Power Systems®
about developments that are considered key to the industry’s future.
Topics will include automotive, embedded solutions, LED technology,
Internet of Things, cyber security, healthcare and wearables.
The program of conferences and forums will explore these exhibition
topics in greater depth. By using the links in this newsletter, you can
find out all there is to know about electronica 2016 and you can also
save time and money by pre-registering.
November 2016
Upcoming SMTA Space Coast Expo & Tech Forum
Alpha Assembly Solutions, the world leader in the production of
electronic soldering and bonding materials, will be an exhibitor at the
SMTA Space Coast Expo & Tech Forum to be held on November 17th
in Melbourne, Florida.
Alpha will showcase their vast product portfolio of innovative materials
and solutions for the electronics assembly industry with particular
focus on ALPHA® Paste & Preforms for optimized process solutions
and ALPHA® Wave Solder Alloy & Chemistries and Cored Solder
Wire for optimal performance. ALPHA® Recycling Services will also
be highlighted at the show to help customers realize their options for
responsibly disposing of solder dross and waste.
“We are seeing more and more stringent specifications from our customers making it ever more critical for our product technology to yield
consistent results while lowering overall costs,” said Don Hosman,
Southeast District Sales Manager for Alpha, a part of the MacDermid Performance Solutions group of businesses. “Our team works
hard to exceed quality, efficiency and budget expectations to provide
advanced products for a host of markets and applications.”
In addition, Alpha will feature material set combinations to include
paste, tacky flux, cored wire and wave soldering flux that are tested
with one another to achieve greater reliability. These pairings, which
were tested to IPC TM650 004B requirements, will
save customers a great deal of trial and error while increasing their
throughput and time-to-market.
To receive information on successful material set combinations, be
sure to visit Alpha during the SMTA Space Coast Expo or visit our site
at AlphaAssembly.com.
PhD Course in “the Smart Transformer”
The chair of power electronics of Kiel University offers a three-day
industrial / Ph.D. course in “the Smart Transformer, its Impact on the
Electric Grid and Technology Challenges”. The Smart Transformer
(ST), a power electronics-based transformer, can provide ancillary
services to the distribution grids to support the grid management,
in addition to the voltage adaptation. The Smart Transformer is a
natural connection point for hybrid (AC and DC) grids both at MV and
LV levels. The Smart Transformer is supposed to reach a market of
US$204.3 million by 2020. This course investigates its possible services to the electric grid plus the power electronics technologies which
could enable these services.
This course consists of lectures (2 ECTS) and laboratory experiments
(2 ECTS). and takes place from February 22 till February 24, 2017 in
faculty of engineering at Kiel university, Kiel, Germany. The registration is open till December 5, 2016 (“[email protected]”).
The instructors are Prof. M. Liserre from University of Kiel; Prof. C.
Vournas from University of Athens, and Dr. G. Buticchi form University
of Kiel.
For more information about the course contents please visit:
or contact: [email protected]
Innovative Products and Concepts for Power & Energy
At the world’s leading show for electronics, electronica in Munich
(Nov. 8-11, 2016/Hall A5-Booth 542) ROHM Semiconductor will
showcase cutting-edge power management solutions for numerous
applications in the automotive, industrial and residential/home arenas.
Utilising the latest SiC and Si technologies, proprietary processing
and packaging technologies, these devices maximise efficiency and
compactness, paving the way for cost and component reduction
while delivering optimum performance. Combining ROHMs knowhow
in analogue and digital power technology, they represent the most
advanced developments of ROHM’s global R&D centers and joint
designs with industry partnerships. All these products are manufactured in the company’s fully owned, vertically integrated production
sites. Among others, key topics will include 3rd Gen SiC MOSFETs,
Schottky barrier diodes (SBDs) and Modules, LED-Drivers for exterior
Lighting, Design Kits for wireless power delivery and the technology
partnership with Formula E-car developer Venturi.
The 3rd Gen of SiC Schottky Barrier Diodes (SBD) realise lowest
forward voltage (VF) and lowest reverse leakage current (lR) over
the entire temperature range among all of the SiC SBDs currently
available on the market. In addition to this, they feature high surge
Bodo´s Power Systems®
current capability which is ideal for power supply applications. Adding
to the recently announced TO220AC devices at 650V/6, 8 and 10A,
ROHM will introduce D2pak devices also adding lower current options, 2A and 4A to the family. ROHM’s new full SiC modules including
a chopper type module for converters integrating both, Trench SiC
November 2016
Energy Efficient and Sustainable
Systems with SiC
ROHM Semiconductor, a leading enabler of SiC, has been focused on developing SiC for use as a
material for next-generation power devices for years and has achieved lower power consumption
and higher efficiency operation.
Full Line-up
SiC Wafer
Discrete and Modules
Leading Technology
ROHM is the first semiconductor supplier
worldwide who succeeded to provide
SiC Trench MF Technology
in mass production
Full Quality and Supply Chain Control In-House integrated manufacturing system
from substrate to module
Full System Level Support
Local system specialists provide
comprehensive application support
ECPE Workshops
Thermal and Reliability Modelling and Simulation of Power
Electronics Components and Systems’ and ‘Model Predictive
Control in Power Electronics - Expectations and Applications 30 November - 1 December 2016, Fuerth/Nuremberg, Germany
Chairmen: Prof. E. Wolfgang (ECPE),
Prof. B. Wunderle (TU Chemnitz), M. Thoben (Infineon)
Model Predictive Control in Power Electronics
- Expectations and Applications 7 - 8 December 2016, Nuremberg, Germany
Chairmen: Prof. R. Kennel (TU Munich),
Dr. T. Geyer (ABB Corp. Research)
18 - 19 October 2016, Nuremberg, Germany
Chairmen: Prof. E. Wolfgang (ECPE),
Prof. U. Scheuermann (Semikron)
ECPE Tutorial ‘Failure Mechanisms of Insulating Materials
27 October 2016, Kassel, Germany
Chairmen: Prof. A. Claudi (University of Kassel), Dr. R. Bayerer (Infineon)
For the ECPE Calendar of Events 2016 and more Power Electronics
Conferences & Events supported by ECPE please visit:
ECPE Tutorial
‘Thermal Engineering of Power Electronic Systems - Part II
(Thermal Management and Reliability)
Delivering Industry’s First 1000V SiC MOSFET
Wolfspeed, a Cree Company and a leader
in silicon carbide (SiC) power products,
has introduced a 1000V MOSFET that
enables a reduction in overall system cost,
while improving system efficiency and
decreasing system size. The new MOSFET, specially optimized for fast charging
and industrial power supplies, enables a
30 percent reduction in component count
while achieving more than 3x increase in
power density and a 33 percent increase in
output power.
“Supporting the widespread implementation of off-board charging stations, Wolfspeed’s technology enables smaller, more
efficient charging systems that provide
higher power charging at lower overall
cost. This market requires high efficiency
and wide output voltage range to address the various electric vehicle
battery voltages being introduced by automotive suppliers,” explained
John Palmour, CTO of Wolfspeed.
“Wolfspeed’s new 1000V SiC MOSFET offers system designers ultrafast switching speeds with a fraction of a silicon MOSFET’s switching
losses. The figure-of-merit delivered by this device is beyond the
reach of any competing silicon-based MOSFET,” Palmour added.
With the introduction of its 1000V SiC MOSFET, Wolfspeed leads the
market with the industry’s most complete device portfolio. Wolf-
Sales Manager in Denmark
TDK-Lambda Germany has appointed a
new sales manager in Denmark where it
opened a subsidiary in April 2013. The global power supplies manufacturer established
a local presence 3 years ago in response to
an increase in demand in Scandinavia.
In order to be well prepared for current and
future business and to promote cooperation
between direct and distribution sales, Mr
Allan Jakobsen has been appointed as the
Bodo´s Power Systems®
speed was the first company to release a commercially qualified SiC
MOSFET in 2011 and remains a leader today, committed to delivering
great technology and value.
Designers can reduce component count by moving from siliconbased, three-level topologies to simpler two-level topologies made
possible by the 1000 Vds rating of the SiC MOSFET. The increase
in output power in a reduced footprint is realized by the ultra-low
output capacitance ―as low as 60pF― which significantly lowers
switching losses. This device enables operations at higher switching
frequencies, which shrinks the size of the resonant tank elements
and decreases overall losses, thus reducing heatsink requirements.
Wolfspeed has determined these proof-points by constructing a 20kW
full-bridge resonant LLC converter and comparing it to a market-leading 15kW silicon system.
Wolfspeed offers a 20kW full-bridge resonant LLC converter reference design, listed as part number
CRD-20DD09P-2. This fully assembled hardware set allows designers to quickly evaluate the new 1000V SiC MOSFET and demonstrate
its faster switching capability, as well as the increased system power
density the device enables.
The LLC converter's reference design files, which include full schematics, bill of materials, simulation files, and detailed a user guide,
can be found online at go.wolfspeed.com/referencedesigns. The full
hardware is available for purchase on demand from Wolfspeed.
new “Sales Manager Nordic”, serving Denmark, Sweden, Norway,
Finland and the Baltic countries. “With his experience in direct customer service management, Mr Jakobsen will continue to serve our
clients in the best possible way. He will also support the success of
TDK-Lambda’s distributor network,” explained Ulrich Schwarz - Sales
Director of TDK-Lambda Germany GmbH.
[email protected]
November 2016
8-11 November
Power & energy A2.547
Passives/modules B5.107
Murata IPDiA A4.518
Enabling your next
power innovation
Mono-block converter
UMAL energy device
Low power DC-DC
DC-DC converter with world’s
highest power density
Low-profile high capacity energy
Murata’s DC/DC converters can
meet the needs for miniaturization,
Ultra small surface mount
package 10.5 x 9.0 x 5.6mm
Low profile,
Low profile,
Quick response to load change
Large capacity (12 mAh),
High efficiency
Outstanding thermal derating
Low internal resistance (200 mΩ)
Long life
Low noise
Continuous discharge
Constant voltage charge
High reliability /
Telecom grade
[email protected]
MGJ6 Low profile
Dual Channel “Smart” Power Amplifier Uses Mixed-Signal Processing
to Generate Multi-Pulse Waveforms
Used in Industrial Ink Jet Printing
Apex Microtechnology is making the design of power analog drive
circuitry for industrial ink jet printer nozzles less of a challenge by pairing a high performance power amplifier with a digital front end. The
MP113 is a complete solution for implementing multi-pulse waveforms
required to drive a changing number of print nozzles with quality and
accuracy in large format printing applications.
provides thermal efficiency with a power dissipation rating of up to
100W per channel.
Key performance specifications include:
The MP113 features onboard firmware that controls a dual channel
amplifier that allows the user to define the desired waveform for their
specific end system. The amplifier is rated at >10A PEAK per channel
with a voltage supply operation of 180V and up to 135V of output
“The MP113’s graphical user interface provides a simplified solution
that is easy to implement with an existing drive system or an entirely
new design,” explains Apex Strategic Marketing Manager Jens Eltze.
“The MP113 supports expanded printing capabilities including the use
of grayscale and the ability to change droplet sizes for greater intensity which is needed to print higher resolution images.”
The modular design of the MP113 is in a broader sense its own
evaluation kit. The embedded firmware allows users to test various
parameters such as the size and speed of droplet formation and to
optimize these specifications for specific types of ink. The MP113 also
is designed with protection and diagnostic features to enhance the
overall reliability of the system. In terms of electrical performance, the
output for the MP113 is referenced to ground and not the negative
supply rail, thus eliminating the need for a high current negative supply. A low current negative supply (≤50mA) allows input to zero volts
without violating the common mode input range. The MP113 also
Pricing, Availability and Evaluation Tools
Sample units of the MP113 are available now for qualified applications, with production volumes targeted for Q4. Consult the factory
for per unit pricing. Complete product information is online at www.
apexanalog.com/apex-products/mp113. For technical support, contact
Apex applications engineering at 800-546-2739, or [email protected]
Apex Microtechnology is an industry
leader in high power analog components,
designed to meet the performance and
cost design targets of our customers’
precision control applications. Apex Microtechnology is headquartered in Tucson,
AZ, USA. More information about Apex
Microtechnology is available at www.
Apex Microtechnology,
Apex and the Apex logo are trademarks
of Apex Microtechnology, Inc.
Bodo´s Power Systems®
November 2016
Hall A4 Booth 219
TI Power at electronica 2016
Meet our power experts for in-depth technical discussions
This year you can join TI at electronica 2016 in person or
online. Register before November, 10th to get access to our
#TIexperts online sessions and ask Bernd Geck your
questions about power management at ti.com/electronica.
Watch an exclusive session explaining loop bandwidth considerations for flyback in CCM regarding RHPZ after November, 10th.
To learn about TI Power broad portfolio
visit ti.com/power
The platform bar is a trademark is a registered mark of Texas Instruments. © 2016 Texas Instruments Incorporated
International Exhibition and
Conference for Power Electronics,
Intelligent Motion, Renewable
Energy and Energy Management
Nuremberg, 16 – 18 May 2017
» Connecting
Global Power
You are the expert, we provide
a platform for you. Join
top-class industry leaders
and present your expertise.
Magnetics: The Eternal Battle
with Myths, Mysteries and
Black Magic – or Maybe
Really all Just a Matter of
Basic Principles and Physics?
By Alexander Gerfer, CTO Würth Elektronik eiSos
Is magnetism an
esoteric science of the
18th century or perhaps
even black magic?
Are magnetics really
a closed book for the
Or EMC ferrites an
unknown entity?
We observe that the majority of engineers,
including our own junior engineers, are not
well trained in the basics. Much is omitted in
training and higher education, but it already
starting in high school. A classic example,
Pythagoras’ theorem, is no longer included
in many curricula. It appears as though this
is simply too trivial? Yet, without a foundation
of well understood (!) basic knowledge, only
superficial knowledge proliferates, or in the
worst case, ignorance is not overcome.
xhibitor and
Become an e
etent, intermeet a comp
rt audience.
national expe
Follow us
# pcimeurope
More information at:
+49 711 61946-820
[email protected]
This example demonstrates exactly where
the problem lies with magnetics too. Regardless of which country’s college and university
curriculum you look at, EMC (Electromagnetic Compliance) and inductive components
are often just side topics. In addition, the
basic knowledge taught is not only diluted,
but often outdated. Today’s inductive components are based on the same physical
principles – but the diversity of core materials
has expanded rapidly; they work in quite
different frequency ranges and in many other
application areas than those described in
the text books as “in practice”. For instance,
Dr. Ray Ridley impressively described at
APEC2016 that students could not even
recognize a transformer on a photo, but they
could certainly identify the robotic comic
figure, “Bumble Bee”, as a “Transformer” on
another photo!
Competitiveness in the market depends on
companies being in a position to achieve
a fast time-to-market. Whoever is faster to
the market with their solution has a decisive
influence on prices and market shares.
Competitiveness also depends on whether
this solution is achieved without interference
emission and with interference immunity,
whether it is efficient and of highly durable
and reliable quality!
Speed to market is compromised if the basic
principles on the origins of interference are
not understood and then the devices have to
be made EMC compliant in laborious EMC
tests and redesigns. Inductive components
are blamed for lack of efficiency in power
supplies in over 60% of cases. However,
going into more detail reveals that much
would have been resolved by choosing the
right topology, correct component selection
and understanding, e.g. of current density
and the associated limit of miniaturization.
Service life depends largely in whether the
design stays cool. High operating temperatures are a curse for all electronics!
So we can only hope that the young engineers actively keep abreast of their continuing education. Not only based on Wikipedia,
but also in the numerous good seminars
offered by the manufacturers.
We at Würth Elektronik eiSos have offered
handbooks for many years and we provide
enlightenment on this subject in seminars.
But how was it in Kant’s famous definition,
“Enlightenment is man’s emergence from
his self-incurred immaturity.” I can only appeal: Keep on learning! Don’t be scared by
black magic, demystify EMC and inductive
November 2016
SEMiX ®3 Press-Fit
It’s Your Choice!
Integrated Shunts
Driver Bundle
Standard Industry Package
125kW up to 250kW
Available in 1200V/1700V from 300A to 600A
Standard Industry Package, optional with integrated shunt resistors
Integrated bundle saves R&D time, material and production cost
Plug-and-Play driver with isolated current/voltage/temperature signal
Pre-applied Phase Change Material with optimized thermal performance
Urban Transport
Phase Change Material
Elektronics with Diamonds
Since quite some time, Silicon is still the workhorse of the electronics industry, but
semiconductors with a broad bandgap (WBGS) are already winning ground in certain
applications faster and faster. For instance Akhan Semiconductor concentrates on the
diamond technology. Diamond is unmatched in its ability to diffuse heat, perform as a
semiconductor, and create smaller and more powerful electronics. Our correspondent
Henning Wriedt talked with its CEO Adam Khan about these WBGS.
By Henning Wriedt, Bodo’s Power Systems, corresponding editor
Henning Wriedt: Mr. Khan, where do Wide Band Gap Semiconductors (WBGS) have their place in the semiconductor technology field
and on the application side?
Adam Khan: Wide Band Gap Semiconductors have the potential to
address nearly the entirety of the materials and device components
of the global semiconductor market but perhaps nowhere more imminently than power electronics. Any user of technology understands
heat is a major issue where cellphones, laptops, tablets, and the like,
are warm to the touch even after limited usage. The performance
demands of consumer technologies have forced the approach of the
physical limitations of the existing silicon based technologies.
Henning Wriedt: Which are the current WBGS-Materials?
Adam Khan: The existing WBGS materials such as Silicon Carbide and Gallium Nitride have been in extensive development since
the 1980s & 1990s, respectively, and have seen traction in power
electronics applications, such as high voltage switching and power
inverter systems, as well as wide usage in optoelectronics applications (such as LED in the case of Gallium Nitride).
Henning Wriedt: Which are the most important specifications of
WBGS compared to other semiconductor materials?
Henning Wriedt: What are the pros and cons of the current WBGS?
Adam Khan: Despite higher processing temperatures (700C) and
dependence on costly substrates, the long maturation Silicon Carbide
based technologies have made it the cheapest of the WBGS materials. Silicon Carbide is currently utilized in power devices, and is starting to play an important role in electric and hybrid electric vehicles
(EV/HEV) markets due its higher operating temperature at higher
Less mature than Silicon Carbide, but more so than diamond, Gallium
Nitride devices have had widespread deployment in primarily LED,
but also Radio Frequency and high power electronics primarily due
to the materials direct bandgap and high-frequency performance.
Despite its advantages, Gallium Nitride materials remain more costly
than both nanocrystalline diamond and Silicon Carbide with restrictions on large area wafer availability.
Diamond shows the best theoretical performance of all WBGS
exceeding the performance capability of the competing materials by
several orders of magnitude. Advances in synthesis and semiconductor doping have since opened the materials capability to address current and emerging market needs, but as it is the least mature of the
WBGS materials, the availability of component devices is presently
Henning Wriedt: Can WBGS somehow extend Moore’s Law, since
current technologies seem to reach a density plateau?
Adam Khan: Very much so. From the design perspective, WBGS materials are capable of faster switching speeds as compared to Silicon
which allow for higher frequencies of operation translating to faster
clockspeeds for logic circuits. More importantly perhaps, diamond is
capable of direct integration with the existing silicon based chips, allowing direct cooling of chip junctions and increasing the performance
capability of the existing silicon based circuitry.
Figure 1: Physical characteristics of Si and the major WBG semiconductors (Credit: Oak Ridge National Laboratory)
Adam Khan: WBGS materials can operate at higher voltages, higher
frequencies, and power densities as compared to silicon, allowing
more power to be delivered using smaller and simpler components
from fewer chips. Also unlike Silicon, they can both operate at higher
temperatures and dissipate heat more efficiently, where diamond has
the highest thermal conductivity. This enables cost efficiency, energy
efficiency, and the next generation of design capability for semiconductor.
Bodo´s Power Systems®
Henning Wriedt: Do the R&D community and the Government support the WBGS technology adequately?
Adam Khan: More recently, there has been a marked increase in
the grants supporting programs involving WBGS materials amongst
the various U.S. based agencies. Further, with ambitious efforts on
the part of the U.S. Department of Energy through programs like the
Power America Initiative and the Materials Genome Initiative, the
public sector R&D climate is very supportive.
November 2016
Henning Wriedt: Your company focuses on one WBGS. Please
explain your Miraj™ Diamond Platform
Adam Khan: At its core, the Miraj Diamond™ platform consists of two
breakthrough innovations in diamond semiconductor research-- the
CMOS compatible Nanocrystalline Diamond synthesis process, developed by Argonne National Laboratory, and the co-implantation n-type
diamond process pioneered by AKHAN Semiconductor. Combined
with related process intellectual property, the platform provides innovative solutions to the automotive, aerospace, telecommunications,
consumer electronics, and military & defense markets rendering more
reliability and efficiency in existing systems and allowing new design
capabilities in next-generation systems.
Dean Technology offers a wide range of power electronics in all
common packages, with ratings to cover most applications. As
with everything we offer custom specifications, packages, and full
assemblies are a specialty - we pride ourselves on helping customers
get exactly what they need for their success.
Contact us today for a catalog, quote, or to begin a new design!
Visit us at
in Booth A5.232
Figure 2: Maximum breakdown voltage of a power device at the same
doping density normalized to Si.
(Credit: Oak Ridge National Laboratory)
Henning Wriedt: What are the manufacturing and integration aspects
of this platform?
AKHAN Semiconductor, Inc.
940 Lakeside Drive
Gurnee, IL 60031, USA
Tel: +847.855.8400
Adam Khan: AKHAN has recently entered the marketplace with its
innovative thermal and optical Miraj Diamond™ products, with pilot
commercial scale manufacturing capability of our initial product line
with licensing product availability for vertically integrated/ high volume
customers. The platform is currently compatible with 200mm process
lines, normative to power electronics wafer sizes. Further the platform can be directly integrated with existing Silicon and a variety of
Glass substrate processes.
Henning Wriedt: Your company is based in Illinois - no need to move
to the Silicon Valley?
Adam Khan: The Illinois innovation ecosystem is rather unique as it
possesses an existing skilled and largely untapped labor force--the
byproduct of having two national laboratories, several world-class
universities, and prolific technology companies such as Motorola. The
state itself has also been supportive of AKHAN’s efforts-- recognizing
the importance of advanced materials manufacturing and the resultant
creation of S.T.E.M. jobs-- crafting an incentive package totaling over
$5 million (USD). The company does maintain a remote presence in
Silicon Valley with both myself and the company’s CTO, Bill Alberth
being based out of our San Francisco office.
Picture : Adam Khan, Founder
and CEO of AKHAN Semi
Adam Khan is the Founder
and the Chief Executive Officer
of AKHAN SEMI. Previous to
AKHAN SEMI, Mr. Khan studied
both physics and electrical
engineering at the University of
Illinois at Chicago before pursuing research at the graduate
level at Stanford University’s
Stanford Nanofabrication Facility
Mr. Khan has authored several patents, technical publications, is
invited to technical talks regularly, and also co-chaired the 2014
New Diamond and Nano Carbons conference (Materials Research
Society). Mr. Khan is co-inventor of the Miraj Diamond™ Platform,
the world’s first CMOS Compatible N-type Diamond Materials and
November 2016
Bodo´s Power Systems®
Peak Current Proof Input
Filter with Multilayer Power
Suppression Bead WE-MPSB
Power supplies are often designed for steady state operation, with transient conditions
mainly considered as an afterthought. In practice, transient conditions such as startup,
shutdown, and load transients are often far more stressful on the components of the power
supply than operation in steady state. To suppress high frequency noise, chip bead ferrites
are mainly placed at the input and output of power supplies.
By Ranjith Bramanpalli, Würth Elektronik eiSos
There are two good examples of transients that are often overlooked, but requires merit careful attention. The inrush current occurs
when a power supply first starts up or when PWM used for variable
loads such as dimming of LED drivers. Chip bead ferrites are often
positioned at the inputs and outputs of power supplies where they
must endure heavy transient currents, and this creates a need for
compact, cost effective devices that are also highly reliable. Such
ferrites are placed at the input and output because they are very
effective at filtering the high frequency noise in switching regulators.
The high frequency noise results from rapid switching currents ringing
with parasitic inductance and capacitance. Such noise tends to occur
at frequencies from 50 MHz to 500 MHz and is known as “ringing”,
“spikes” or “periodic and random deviation noise” (PARD noise). Figure 1 shows PARD at the origin, the switching node, and also shows
how PARD noise shows up at the input and output of the switcher.
In general, chip bead ferrites should be always placed as close as
possible to the converter as source of noise. However, one effective
method for preventing PARD noise from getting into the input and
output leads of a switching power supply is to place chip bead ferrites
in series with the inputs and outputs. These should be placed as close
as possible to the edge of the PCB and/or to the connectors, and
footprints for ferrites should be placed in series with both the positive
and the negative of each connector.
Again, in general ferrites should be placed as close as possible to
the source of the noise because the noise couples into the unfiltered traces and cables. But be aware that it is very likely that high
frequency noise can couple around a ferrite via parasitic capacitance
to GND and earth planes. Most EMC standards begin limiting radiated EMI at 30 MHz, so preventing this unwanted antenna effect of
input and output leads is highly important. When a ground plane or
a shielded enclosure is present, noise can couple around a ferrite
placed towards the interior of a PCB, as shown in Figure 2.
Ground Plane
PARD Noise Loop
PARD Noise Loop
Figure 2: PARD noise gets around chip bead ferrite beads L1-L4 by
coupling capacitive through the ground plane and earth to the input
and output connectors
Figure 1: PARD noise without Chip Bead Ferrites starts at the switching node of a buck converter (blue) and contaminates the input voltage (yellow) and the output voltage (green)
Zooming in and measuring the frequency of PARD noise in Figure 1
reveals a frequency of 170 MHz. Conducted noise like the waveforms
in Figure 11 will generate radiated noise if it leaks onto input and
output wiring harnesses.
Bodo´s Power Systems®
Würth Elektronik eiSos has recently developed a family of chip bead
ferrites that feature high average / RMS current ratings, low DC
resistance and are also tested and specified for high current pulses.
This peak current proof series, the WE-MPSB Multilayer Power Suppression Bead family, is especially suitable for use in positions where
short-duration currents far exceed the average currents.
Inrush Currents at Turn-On
At the moment when a power supply is turned on, any capacitors
connected to the input bus will begin to charge. In some very rare
cases, a soft-start of the input supply controls the ramp in a smooth,
November 2016
VIN = 12.0 V -+
180 μF
8 mΩ, 0.3 µH
8 mΩ, 0.3 µH
monotonic behavior, but in most cases, the input voltage ramps up
very quickly. For example, if the 12 VDC power bus in Figure 3 is
already up and running when a mechanical switch connects it to the
buck converter, the ramp slope is only limited by the source resistance and the resistance and parasitic inductance of the leads/PCB
traces/switch. For this application note the resistance and inductance
of a 30 cm banana-to-banana test cable was measured and came out
to 8 mΩ and 0.3 µH, respectively. In practice all voltage sources are
current limited, but if the 12 VDC bus had a large amount of output
capacitance, a fact for the laboratory DC power supply used in this
application note, then the charging current when the mechanical
switch closed could easily exceed 30 A as shown in Figure 4.
L1 2.2 µH
WE-XHMI 74439358022
10 μF
10 μF
100 nF
400 kHz
pulses of current in a short time. The same 30 cm cables were used
to connect the 5.0 V output to a load that draws the maximum of 8 A
of output current, and Figure 5 shows that when the 8 A load is connected with a fast rise time the current transient comes close to 25 A.
5.0 V, 8 A max
Figure 3: Schematic of the test buck converter showing source
resistance, input lead resistance and inductance along with all input
Figure 4: Input inrush current of 33 A for a 12 VDC bus with a nearinstantaneous connection charging 20 µF of ceramic and 180 µF of
polymer aluminum input capacitance
Figure 5: Outrush current for a 5 VDC bus with a near-instantaneous
connection to an 8 A load with 200 µF of ceramic and 660 µF of
polymer aluminum output capacitance
Using WE-MPSB Multilayer Power Suppression Bead
Problems with Steady State Current Ratings
The WE-MPSB family was designed to provide a similar range of
impedances as the standard, WE-CBF family of chip bead ferrites.
The WE-CBF family provides RMS current ratings, but like nearly
all other chip bead ferrites from any manufacturer, no peak or pulse
current ratings. In this example, in order to handle a 33 A pulse with
steady state specifications multiple WE-CBF family devices would
be needed, since the highest RMS current rating in this family is 6 A,
for the 1806 or 1812-sized devices. Just one WE-CBF family device,
for example the 4 A-rated
742 792 150 with a 1206
case size and rated 80 Ω
at 100 MHz would handle
the steady state current, but
repeated startup transients
could lead to failures such
as the ones depicted in
Figure 6.
Six such devices would
ut inrush current of 33 A for a 12 VDC bus with a near-instantaneous connection charging 20
be needed for the positive
Figure 4 shows a pulse that peaks at approximately 33 A and settles
nd 180 µF of polymer aluminum input capacitance >>
input line and another six
after around 100 µs to the 5 A current limit of the laboratory power
for the negative input line,
supply used as the input source. It then takes another 200 µs to
a pulse that peaks at approximately 33 A and settles after around 100 µs to the 5 A current
and this is not practical
charge the input capacitors up to the target 12 V. Compare this waveatory power supply used as the input source. It then takes another 200 µs to charge the input
for several reasons: First,
form to the steady state input source current:
the target 12 V. Compare this waveform to the steady state input source current:
chip bead ferrites can be
Vout · Iout, max
paralleled for continuous
Isource, max =
= 3.7 A
η · Vin, min
0.95 · 11.4 V
currents and their positive
of 95%)
temperature coefficient
is is
of 95%)
will ensure that they share
e facing the The
designer becomes
any inputbecomes
filter components
must be able
to more or less evenly.
facing theevident:
circuit designer
evident: any
rrent pulses input
each filter
time components
the convertermust
is switched
but selecting
ferrites rated
such current
be ableon,
to handle
heavy current
pulsesto handle
leads to overdesign
state.is switched on, but selecting ferrites rated to
sharing is neither tested nor
each time
guaranteed for short-durahandle the full pulse current leads to overdesign for steady state.
tion pulse currents. Second,
ts at Turn-On
placing several components
Outrush Currents at Turn-On
s will be placed at the output. The converter has two polymer aluminum 330 µF output
parallel with impedThe
will be
at the
The converter
20 mΩ of ESR
two 100
µF placed
with has
3 mΩ
Figure 6: Melted and burned chip bead
his capacitor bank is capable of supplying large pulses of current in a short time. The samethat is dominated by
ferrites due to overcurrent and overheatresistance
ere used to connect the 5.0 V output to a load that draws the maximum of 8 A of output
causes the inductance, the ing
ESR the
bank is
ure 5 shows mΩ
8 AThis
is connected
a fastofrise
time thelarge
current transient
25 A.
Bodo´s Power Systems®
November 2016
䘀椀氀洀 䌀愀瀀愀挀椀琀漀爀 䐀攀猀椀最渀攀搀 䘀漀爀
一攀砀琀 䜀攀渀攀爀愀琀椀漀渀 䤀渀瘀攀爀琀攀爀猀
㄀ 洀洀
䰀䠀㌀ 䘀䔀䄀吀唀刀䔀匀
䰀漀眀 䔀匀䰀ⴀ 氀攀猀猀 琀栀愀渀 ㄀ 渀䠀
伀瀀攀爀愀琀椀渀最 琀攀洀瀀攀爀愀琀甀爀攀 琀漀 ⬀㄀ 㔀뀀䌀
䠀椀最栀 刀䴀匀 挀甀爀爀攀渀琀 挀愀瀀愀戀椀氀椀琀礀 ⴀ 最爀攀愀琀攀爀 琀栀愀渀 㐀 䄀爀洀猀
䤀渀渀漀瘀愀琀椀瘀攀 琀攀爀洀椀渀愀氀 搀攀猀椀最渀 琀漀 爀攀搀甀挀攀 椀渀搀甀挀琀愀渀挀攀
嘀椀猀椀琀 甀猀 漀渀氀椀渀攀 愀琀 眀眀眀⸀攀挀椀挀愀瀀猀⸀挀漀洀
一漀爀琀栀 䄀洀攀爀椀挀愀㨀 猀愀氀攀猀䀀攀挀椀挀愀瀀猀⸀挀漀洀 䔀甀爀漀瀀攀㨀 猀愀氀攀猀䀀攀挀椀挀愀瀀猀⸀椀攀
resistance and the impedance to drop, making them far less effective
at filtering the desired noise. Third, six components cause high costs
and need much PCB space.
Figure 7: Enter the pulse length, peak current and number of pulses
Select the proper WE-MPSB
In situations where peak currents exceed average current by ratios
from 3:1 to nearly 10:1 WE-MPSB are typical area of use. A first pass
for selecting chip bead ferrites is to review all parts that can handle
the RMS current of 3.7 A.
Peak Current Proof Ferrites for the Input
In our application we are expecting 10 000 switching cycles during
lifetime, so 10 000 pulses with 33 A will stress the WE-MPSB of the
input filter and needs to be survived. The first step and most comfortable way is to enter these data into the pulse designer of REDEX-
PERT. There are 9 parts left, which we took all in the product storage
for easy comparison.
Validation of Effective Resistance
From the 9 left WE-MPSB we now select the one with the highest
resistance (not total impedance) at the noise frequency. In general
chip bead ferrites have their highest resistance at the frequency of
their highest total impedance, but for other frequencies there is no
general approximation possible. The fastest way to find the best part
is using REDEXPERT from Würth Elektronik (www.weonline.com/
redexpert). As a registered user you can place the chart slider at 170
MHz (see Figure 8), and read directly the resistance values of each
part out of the grid, and even sort descending to get the part with
highest resistance.
Considering all of above parameters the red highlighted part WEMPSB 742 792 245 51 seems to be the best suitable component for
our application. Its current rating is 4.0 A, and it can withstand about
18 700 pulses of 33 A and 8 ms length. Keeping in mind that this 8 ms
is much longer than the initial pulse of 500 µs and the short peak of
100 µs, gives it a plenty safety margin. From all suitable components,
it is the one with the highest resistance at 170 MHz.
Pulse-Stable Ferrites for the Output
The output RMS current is the same as the average output current of
8.0 A. Following the same guidelines, there are five candidates rated
for greater than 8.0 A:
All five parts can handle more than 10 000 pulses and have more
than 8 A rms current rating, so the final selection will require actual
EMC testing to determine which part filters the most noise. The
smaller parts are less expensive but provide less reduction of noise.
Testing of the selected components
For final lab testing we added the above mentioned WE-MPSB
742 792 245 51 to the input and WE-MPSB 742 792 251 01 to the
output. You can easily see, that the green output voltage is now
already silent.
Performing radiated EMI scans proves that the chip beads successfully suppress the PARD noise. Especially in the range of the 170
MHz PARD ringing, the EMI is significantly improved.
Further Considerations
Influence of RDC to overall efficiency
The used WE-MPSB 742 792 245 51 has a DC resistance of 35 mΩ,
which adds additional conduction losses and therewith reduces the
efficiency. The measurements in our lab shown just a slight decrease
Figure 8: Determine the best suitable WE-MPSB with REDEXPERT
leads to 742 792 245 51
Bodo´s Power Systems®
November 2016
Wir bieten Lösungen
n aus!
W ir stelle and 130
Halle B6
Passive und elektromechanische Bauelemente
Gewinner der Silbermedaille bei der
Wahl zum Distributor des Jahres 2014
in der Kategorie Passive Bauelemente:
Technische Kompetenz & Support
of the efficiency from 95% downwts
for each used chip bead
electronic components GmbH · Fon: + 49 (0) 51 30 / 58 45-0 · Fax: + 49 (0) 51 30 / 37 50 55
Your Powerful
ferrite. To investigate it in detail, we calculate the efficiency as follow:
[email protected]
Influence of DC Bias to Impedance Characteristic
As all magnetic parts, also chip bead ferrites follow the physics
principle of elementary magnets. With increasing DC current they
will continuously saturate up to the level of complete saturation. This
saturation effect shifts the impedance curve, as shown in figure 10.
The peak inductance value remains almost constant with a drop of
just 40% of its initial value, whereas the impedance at lower frequencies significantly drops by rates of up to 90%. In low frequencies the
inductive part is dominating, which saturates with DC current. Above
the SRF the capacitive part is dominating, which is not effected by the
DC current.
wts // electronic is specialized in distribution of
powerful passive and electro-mechanical components. And in your success. Contact us!
Visit us in 16
Nov 8-11,
oth 126
Hall B6, Bo
Phone: +49 (0) 5130 / 58 45-0 · Fax: +49 (0) 5130 / 37 50 55
[email protected] · www.wts-electronic.de
* Winner of the silver medal in the election Distributor of the
Year 2015 in the category “Technical Expertise and Support“.
Figure 9: Lab Results with Chip Bead Ferrites. The green output voltage is almost silent
However, if you do your EMI measurements at full load current, this
(worst case) impedance shift is already considered in your measurements, and you don’t have to care about this effect. Important to know
is, that the bigger the size of the chip bead ferrites, the lower is the
shift of impedance caused by DC current.
Z vs. f vs. IDC
Impedance [Ohm]
0.0 A
1.0 A
2.0 A
3.0 A
4.0 A
Frequency [MHz]
Figure11: Impedance of WE-MPSB 742 792 245 51 with DC bias current from 0A to 4A
Figure 10: Lab Results of EMC-Testing. The user WE-MPSB chip
beads significantly improve the radiated EMI in the range from 100 to
250 MHz
Chip bead ferrites are the best components for reducing high frequency noise above 10 MHz. In power supply layouts they must be
placed as close as possible to the source of noise which can be the
input and output connectors, to properly filter conducted EMI from
the input and output wiring harnesses. This prevents conducted EMI
from becoming radiated EMI. Being the first components and the last
components in the chain exposes chip bead ferrites to heavy transient
currents, and circuit designers can now select parts that filter noise
with minimal impact to power efficiency and will handle large current
pulses with excellent reliability.
November 2016
Bodo´s Power Systems®
Control Method for a Reverse
Conducting IGBT
When IGBT and diode functionality is combined into a single piece of silicon, a reverse
conducting IGBT (RC-IGBT) is created. This allows a standard IGBT/diode-module to
be built on a single silicon chip. This results in enhanced current carrying capability
without increasing the foot print of the module and – depending on the device technology
– allowing the diode´s electrical performance to be influenced by the control state of the
IGBT gate. However, in order to manage the losses in the combined RC-IGBT, special
control approaches need to be considered.
By Daniel Domes, Infineon Technologies AG
Device Introduction
Reverse conducting IGBTs can be built by partially interrupting the
p-doped collector area with n-doped regions. This creates the diode
functionality, yet there remains sufficient area for the IGBT to inject
minority carriers into the drift region for low forward voltage (VCE(sat)).
With this approach, the diode functionality is dependent on the state
of the gate control. Devices of this type are designed for hard switching applications, and are known as Reverse Conducting IGBTs with
Diode Control (RCDC-IGBT).
Figure 1: 6.5 kV RCDC-IGBT static diode performance as a function
of gate voltage. In cross section: red color is p-type-doped, green
color is n-type-doped. Tvj=125 °C
Loss Optimal RCDC-IGBT Performance
The RCDC-IGBT gate state has a significant impact on the forward
characteristics of the diode. From the static loss perspective, in diode
conduction mode the gate needs to be turned off. The lowest VF can
be achieved when VGE=-15 V, this is a little higher when VGE=0 V.
Since VF corresponds to the carrier density inside the chip, for the
lowest dynamic losses and thus lowest Qrr, VF should be selected to
be a high value.
Deciding how to drive the gate in diode conduction mode will depend
on the pulse frequency of the application and the ability to desaturate
the diode prior to it turning off.
Special Gate Drive Aspects
A gate driver for low loss RCDC-IGBT operation needs to be able to:
• detect the diode conduction mode and prevent turn on of the
• desaturate the RCDC-IGBT diode by driving VGE to 15 V prior to
diode turn off
Bodo´s Power Systems®
Figure 2: Flow chart of the RCDC gate driver control scheme
• drive VGE to 0 V in diode conduction mode in the case of a typical
6.5 kV inverter pulse frequency and limited diode desaturation time
• detect a load current zero crossing in diode mode and turn on the
RCDC-IGBT gate for smooth current transition from the diode to the
IGBT of the same switch
• detect the load current zero crossing in IGBT mode and turn off the
RCDC-IGBT gate for low loss diode operation
Detect the Diode Conduction Mode
In a classic inverter, a forward conducting IGBT is turned off at the
start of the interlock time period. For the opposite diode, this means
that first the blocking voltage decreases and then the current starts
to rise. Once the interlock time period is over, the antiparallel IGBT
gate of the diode is turned on. For an RCDC-IGBT, the turn on of the
conducting diode´s antiparallel IGBT needs to be prevented by the
gate driver logic.
November 2016
It is recommended to monitor the VCE of the switch before executing the turn on command from the control side. In this scenario, the
voltage across the diode switch is low before the interlock time ends,
clearly indicating that the diode is conducting.
The high voltage detector is a simple frequency-compensated voltage
divider. In high voltage applications, this circuit is often present in the
gate driver stage for desaturation detection purposes and adds no
additional parts to the Bill of Material (BOM).
Figure 3: Diode detection without desaturation pulse
Figure 4: Diode detection with desaturation pulse
For diode desaturation purposes, the interlock time is calculated for
each gate driver individually. Consequently, the high and low side gate
driver input signals will change at the same time. Falling edges of
the control signal are executed immediately, turning off the LS-IGBTs
gate. The IGBT turns off normally and the voltage across the high side
switch decreases. A voltage detector checks whether the VCE of the
high side switch drops below a defined threshold (displayed as “VCE
low”). In this case, the high side switch will go into diode conduction
mode and the gate (VGE) is switched from -15 V to 0 V as soon as the
detector output “VCE low” changes.
Diode Desaturation
Detecting the diode conduction state and keeping the corresponding
switch gate in the off-state ensures a high carrier density inside the
device thus maintaining low VF-values. However, for dynamic loss
reduction this condition is not desired as high carrier density causes
high Qrr and hence high IGBT turn on and diode turn off losses.
If the diode switch gate is turned on before the diode is turned off, the
operation point is shifted from a low to a high VF output curve and
the diode carrier concentration is reduced with a strong effect on the
IGCTs. Highest
power density for
most compact
The IGCT is the semiconductor of choice for
demanding high-power applications such as
wind power converters, medium voltage
drives, pumped hydro, marine drives,
co-generation, interties and FACTS. ABB
Semiconductors’ range of 4,500 to 6,500
volt asymmetric, reverse conducting and
reverse blocking IGCTs deliver highest
power density and reliability together with
low on-state losses.
ABB Switzerland Ltd. / ABB s.r.o.
[email protected]
November 2016
Bodo´s Power Systems®
dynamic losses. Typical desaturation time for a 6.5 kV RCDC-IGBT is
20 to 100 µs.
For practical implementation, the driver needs to accurately predict
the point in time at which the diode turns off. This corresponds to the
opposite IGBT turning on, which (based on the signal definitions) is
executed after the IGBT switch control signal changes from low to
high and the interlock time tinterlock is over.
This approach is illustrated in Figure 4. The high side switch diode
conduction state was detected and the gate switched to VGE=0. Now,
the high side and low side gate input signals change synchronously.
The low side gate driver counts the interlock time and when it is over,
turns on the low side IGBT.
The diodes switch gate driver creates the desaturation pulse by driving VGE to 15 V. No active switching in the half bridge occurs until the
interlock timer is over. The gate driver of the diode switch remains
with VGE at 15 V for the desaturation time (tdesat). The duration of tdesat is shorter than tinterlock since the remaining locking time tlock must
be added. The locking time should be kept small to prevent the diode
saturating again, reducing the effect of the desaturation. A typical
value of tlock for a 6.5 kV RCDC IGBT is 0.5 µs.
With this approach, the diode desaturation duration corresponds
to the maximum interlock time tolerated by the application. A long
interlock time ensures best device performance but decreases the
dynamic response of the system. Using a very small gate resistor
applies the shortest time constants for the desaturation pulse and
gives the best desaturation result. In Figure 2 references this resistor
as RGD, whereas the nominal gate resistors are named RGI(on) and
Considering a practical 6.5 kV traction inverter system with a frequency of several hundred Hertz and maximum interlock time of 20µs, the
RCDC-IGBT performs best if the gate runs at 0 V in diode conduction
mode. In this case, the static diode losses are slightly higher than with
operation at VGE=-15 V. Total losses are minimized as Qrr is lower
than with VGE=-15 V diode operation. For other frequencies and longer desaturation times the optimal operation timing will be different.
Load Current Zero Crossing Approach: Diode to IGBT
If, in a classical inverter approach, a diode is conducting then the load
current can change polarity as the antiparallel IGBT is normally turned on
via the gate. For an RCDC-IGBT, this situation must be detected and the
gate turned on immediately to avoid interrupting the load current.
If a PN-diode conducts and the current decreases to zero, the diode
remains flooded with carriers allowing the load current to reverse
direction even though the antiparallel IGBT gate is not turned on. In
Figure 5a, the load current (IL) changes direction at t4 but, as IC(HS),
still flows through the diode. The high side IGBT gate remains in an
off state, as its control signal is low. As soon as the carriers in the
diode are depleted by the load current, the voltage across the diode
reverses at time t5. The load current di/dt is small compared to di/dt in
a hard switching event.
The gate driver must check for positive VCE while the diode is conducting. As soon as VCE becomes positive, the gate is immediately
turned on. The detection circuit must be able to react to low positive
VCE voltages, to avoid the output voltage change becoming unnecessarily high. In Figure 5a, at time t5 this effect is exaggerated. It is
advised to use a classical desaturation detection circuit with a high
voltage diode chain, a current source and a comparator.
Figure 6: Load current zero crossing, commutating from diode (IC<0)
at VGE=0 into the antiparallel IGBT (IC>0) with VGE=15 V; very small
increase in VCE (see inset) at the time the detector recognizes the
zero crossing event, load current without interruption
Figure 5: a) simplified RCDC-IGBT half bridge system, b) schematic
waveforms for demonstration of load current zero crossing at t4, current IC(HS) goes through the diode for t2 ≤ t < t5 and changes into
IGBT for t5 ≤ t < t6
Bodo´s Power Systems®
Figure 6 shows the load current commutation from the diode to the
IGBT by means of RCDC IGBTs in an H-bridge configuration. The
gate driver circuit detects the small increase in VCE (inset) and turns
on the RCDC-IGBTs gate. The load current changes polarity without
interruption or excessive voltage distortion.
November 2016
0Load Current Zero Crossing Approach: IGBT to Diode
As well as the transition of the load current from the diode to the
IGBT, the current can also change its direction to flowing from the
IGBT into the antiparallel diode. This does not risk interrupting the
load current as the gate remains in an on state and the diode sinks
the current. If VGE remained at 15 V, VF would be unnecessarily high
and thus the static losses increase until the next control command is
received. It is recommended to use the proposed desaturation circuit
again, detecting a small VCE voltage across the RCDC-IGBT. Since
VF is initially high, the voltage difference in VCE from IGBT to diode
conduction also becomes high and can easily be detected.
Drive Scheme
Figure 2 shows the complete RCDC-IGBT gate driver control scheme.
The state machine is able to handle all basic RCDC-IGBT gate drive
requirements including diode conduction mode detection, diode
desaturation, load current zero crossing from diode to IGBT and vice
Figure 7 shows the gate driver used. If IGBT switching is required, the
gate resistors RGI(on) and RGI(off) are used. If minimum time constant switching is required to desaturate the diode, a comparatively
small RGD is used. The advanced H-bridge concept allows VGE to be
driven to 0 V when the diode is conducting.
Figure 7: Schematic gate driver circuit for operating a RCDC-IGBT in
a conventional inverter system, from the inverter control stage only
the ctrl signal needs to be provided, all RCDC-IGBT specific information is generated and processed inside the gate driver circuit.
In high voltage IGBT gate drivers a high voltage divider is commonly
used for desaturation detection. The RCDC-IGBT gate driver has a
desaturation circuit consisting of a high voltage diode chain, comparator and current source. Logically, three binary input signals “ctrl”,
“VCE” and “HV desat” are processed by the state machine.
Testing Static and Dynamic
Power Behaviour
NEW! Motor Drive Analyzer
8 Ch, 12-bit, up to 1 GHz
Complete Test Capabilities for Power
and Embedded Control Debug
FREE Power Poster
Learn more
new Power Basics Poster:
November 2016
Bodo´s Power Systems®
Innovative 7in1 IGBT Packages
for Scalable and Easy Design of
Industrial Drives and Inverters
Industrial drive applications require scalable IGBT modules to simplify the design and
to provide the possibility of utilizing the same components like driver boards and bus
bars for different inverter power ratings. The Mitsubishi NX-series 7in1 IGBT packages
provide such scalable solutions combined with high power density and simplified inverter
By Thomas Radke and Narender Lakshmanan Mitsubishi Electric Europe B.V
Different motor control inverters require different IGBT power module
package sizes. Conventionally, for smaller inverter power ratings, the
6in1 IGBT power modules are used. For higher ratings, inverters are
conventionally built by using three half bridge power modules. As a
result the inverter construction is different for different power ratings,
thus the driver boards, bus bars and heatsinks cannot be reused for
achieving higher power ratings. This necessitates higher development
expenditure alongside the requirement for complicated logistics for
the sourcing of new and unique components. In several motor drive
applications, reactive power required for the motor operation has to
be considered and an additional brake unit has to be incorporated. To
address the requirements of scalability and reducing the number of
components, the NX-series 7in1 IGBT packages have been developed. The NX-series 7in1 packages include a three phase inverter
bridge plus an additional brake chopper IGBT with a current rating
between 75A and 300A in the 1200V class.
existing standard IGBT power module housings and therefore various compatible rectifier modules from different manufactures are
available. To provide full scalability for small and large packages, the
same arrangement of power and control terminals are designed. This
offers the possibility of using the same components like driver boards
and bus bars for different power ratings. This approach - multiple
utilization of components is a key factor in minimizing the development effort, time and costs. The separation and the orthogonal
arrangements of the dc-terminals (P/N) and the ac-output allows for
a simplified inverter construction. In cases where the brake chopper
is not required, the unused IGBT can be turned-off by short circuiting
the gate and emitter pins. Therefore, depending on the requirement,
the 7in1 module can also be easily utilized as a three phase inverter
without the brake chopper.
NX-series 7in1 IGBT Modules
The NX-series 7in1 IGBT modules contain a three phase inverter
bridge and a brake chopper as shown in the internal circuit diagram
in Figure 1. A thermistor is implemented to monitor the baseplate
Figure 2 Multiple use of components enabling scalability
Figure 1: Line-up and circuit diagram of NX-series 7in1 IGBT Modules
Two different packages were developed in order to cover the whole
line-up from 75A to 300A. An inverter with the high power density
can be achieved by implementing these current ratings into the
122x62mm² footprint (small pkg.) and the 122x122mm² footprint
(large pkg). The 17mm module height is already compatible to other
Bodo´s Power Systems®
Inverter Benchmark
Conventionally, the 6in1 IGBT modules are available in a package
with 122x62mm² footprint with a current rating up to 200A. Usually pin terminals are used which limit the current capability of the
contacts. An additional limitation is the high heat concentration in
air cooled heat sinks due to the small package footprint. Due to this
reason, the inverter power rating based on 6in1 modules is limited to
about 55kW. Inverters for power ratings above 55kW conventionally
employ three half bridge modules with an additional optional brake
November 2016
STSPIN™ motor drivers:
Leading in integration, performance
and efficiency
• Scalable and flexible for all motor types with wide operating voltage, current and temperature range
• Extensive diagnostics and fully-protected to reduce the number of external components, cost and
• Up to 256 micro-steps per step for unprecedented smooth and silent motion
• Multiple package options to meet a variety of design challenges from board space to thermally harsh
• Comprehensive development environment for fast and easy design
Motor type
Typical power
range (W)
Part number
20 - 1000
L6480, L6482, L6506, L297
< 10
5 - 90
L6472, L6470, L6474, L6228, L6208
40 - 600
< 10
5 - 90
L6227, L6226, L6225, L6207, L6206, L6205, L6203, L6201, L298, L293, L2293Q
< 10
5 - 60
L6229, L6235, L6234, L6230
Brushed DC
Drivers with
For further information
and full design support,
visit us at www.st.com/stspin
chopper module. This change from a compact 6in1 IGBT module to
three half bridge modules translates into a heatsink size increase of
about 210%, greater complexity in design and an expensive inverter
construction. In place of just 1 power module four power modules
have to be assembled. The compact NX-series 7in1 package offers a
power range comparable to the conventional 6in1 module. The large
package version of the NX-series 7in1 module allows for an extension of up to 90kW in the inverter power range. Therefore the large
package version of the 7in1 modules offer an intermediate solution
between the compact 6in1 / 7in1 modules and the high power half
bridge modules. The high power density of the large 7in1 package
is achieved by using low loss CSTBTTM IGBT chips combined with
a thermally optimized package structure and additionally utilizing the
maximum allowed junction temperature of 175°C. A thermal simulation by using Mitsubishi’s Melcosim software ([3] publically accessible
Ver. 5.3) has been performed and the result is shown in Figure 3. This
simulation considers typical conditions for motor control applications
with a switching frequency of 4 kHz. The heatsink temperature is
considered as 100°C which is typical for air cooled applications. As
demonstrated, one 200A / 1200V 7in1 module (CM200RXL-24S in the
large package) has performance comparable to three 2in1 modules
with 225A /1200V rating (CM225DX-24S1). This alternative 7in1
solution instead of three, (or with brake chopper - four) 2in1 modules
reduces the size by about 50% as shown in Figure 4. The 300A /
1200V 7in1 (CM300RXL-24S1) module is able to deliver an even
higher performance. The 6in1 module, with 200A rated current has
thermal limitations and cannot provide the inverter output power equal
to an inverter employing the 7in1 or the 2in1 modules as shown in
Figure 3. Additionally an air cooled heat sink design which maintains
the sink temperature (Ts) below 100°C will be quite difficult to achieve
using the 6in1 module (considering the relatively small foot print of
122x62mm²). Considering an optimistic assessment where a heat
sink temperature of 100°C can be maintained, the maximum output
current of the 6in1 200A module is limited to 120Arms at an IGBT
junction temperature of about 125°C. Considering the same condition
(125°C junction temperature), the 7in1 200A/1200V module is able to
deliver an output current of about 150Arms. The heat sink design for
the baseplate of the 7in1 will be much simpler since the base plate
area of the 7in1 module is two times greater than that of the 6in1
module. Therefore it is reasonable to consider operating the heat sink
at 100°C while utilizing the large 7in1 module.
The NX-series 7in1 IGBT modules provide an optimized package for
the requirements of motor control inverters. Taking advantage of this
demonstrated scalability offered by the large and small packages, the
same inverter construction concept can be utilized for developing an
extended inverter series capable of delivering higher power levels.
The efforts required for designing and implementing the extension
of the inverter power range compared to the conventional approach
(using 6in1 modules) is greatly reduced because components like bus
bars and driver boards can be reused. The low loss CSTBTTM chips
in combination with the superior thermal performance offered by the
thermally optimized package structure in the large package delivers a
high power density.
Therefore the 7in1 modules offer an optimized solution to design a
scalable cost effective motor drive inverter.
Figure 4: Heatsink size and inverter power comparison
[1] NX-series IGBT modules Application Note
[2] Datasheet CM200RXL-24S
[3] Melcosim Version 5.3 (https://www.mitsubishielectric.com/semiconductors/ssl/php/members/login/login_s.php)
Figure 3: Thermal performance comparison
Bodo´s Power Systems®
November 2016
huge peak
strong bead!
electronica Munich Hall B6 Booth 404
The WE-MPSB series is the world’s first ferrite bead that specifies how it performs under high current transients.
 muscular peak current capability
This unique bead protects and extends the life of your application. It features an ultra-low RDC which delivers the
 high rated currents
lowest self-heating in its class at high currents. The WE-MPSB is ideal for DC/DC applications requiring high
 ultra-low RDC to minimize
efficiency. For more information, please visit www.we-online.com/WE-MPSB
unwanted losses
 effective broadband filtering
Design your filter with our free online software – REDEXPERT. You will have the best guidance to ensure your circuit
can withstand transient peak currents using our peak-specified WE-MPSB ferrite beads: www.we-online.com/
Products in original size:
A stone step for smooth replacement in Industrial Applications
Features and benefits of silicon carbide (SiC) have been extensively demonstrated. It has
been proven to be the suitable material for high voltage, high frequency power switches.
As shown in Figure 1, several SiC based devices have been launched in recent years.
By Dr. Vladimir Scarpa, Field Application Engineer-Power, ROHM Semicondcutor GmbH
Quality and Reliability
One of the main concerns from development engineers is the reliability of SiC devices. This is understandable, since SiC does not have
yet same time on the field than equivalent IGBTs and MOSFETs made
out of silicon.
In an initial phase, in early 2000s, Schottky barrier diodes (SBD) out
of SiC have been implemented. SiC SBDs have been widely used in
power factor correction (PFC) stage of single-phase switched mode
power supplies (SMPS). They provided very low reverse recovery
charge, saving significant amount of dynamic losses in the diode
itself and in the counterpart switch. Lately, further applications, like
photovoltaic inverters and uninterruptable power supplies (UPS) also
adopted SiC SBDs.
1st SiC SBD
1st Full SiC
6“ SiC Wafer
As it can be shown in Table 1, qualification processes of both silicon
and SiC devices are following same standards and test conditions.
ROHM also does innumerous further tests to verify the quality and
lifetime of its products. Results of these tests can be directly shared
with customers. In addition, when customer requires special tests
outside standard conditions, they can be performed by ROHM after
1st SiC JFET
1st SiC Planar MOSFET
Gate oxide lifetime
As shown in Figure 1, ROHM recently launched the 3rd Gen of SiC
MOSFETs, first devices with trench gate, different from planar gate
from former generations. This new structure practically eliminates the
resistance from the parasitic JFET inside the structure. As a consequence, the amount of on resistance per area unity has been reduced
by half [2].
1st SiC Trench
Figure 1: Timeline with main market events in the area of SiC devices
The trench structure adopted by ROHM, also known as U-structure,
is shown in Figure 2 (a). As it can be seen, the electrical field around
the interface SiC- SiO2 during blocking time is much lower than in the
conventional trench structure. This guarantees long lifetime of gate
oxide, also for operation voltage levels close to nominal.
With respect to active switches, in 2010 the first normally-off silicon
carbide switches were launched based on MOSFET structure. These
devices simplified the gate driver circuitry, and enabled a smoother
transition from standard silicon IGBTs and super
IEC Standard
junction MOSFETs. As consequence, many
High Temperature
electronic companies from different application
Reverse Bias
fields started to see SiC switches in their shortterm horizon, and began development projects
High Temperature
which include SiC MOSFETs. More recently,
Gate Bias
after extensive qualification procedure in system High Humidity
level, part of these projects turned out into prodHigh Temperature
Reverse Bias
ucts that were then introduced into the field.
High Temperature
Remarks SiC2
1000 h @ 95% Vds,max,
Tamb = 125..145°C
@ 100% Vds,max ,
Tamb= Tj,max =175°C
1000 h @ ±VGS,max,
Tamb= Tj,max
1000 h @ Vds,max= 80V
85% RH, Tamb= 85°C
1000 h @ T
There are however some particularities in the
adoption of SiC MOSFETs. In order to get the
Low Temperature
1000 h @ TSTG,min
maximum of their benefits, some changes in
system design maybe required. This article will
100 cycles
Thermal Cycle
-describe some of the issues faced before and
TSTG,max - TSTG,min
during development phase. For each of these issues, some technical hints and, in some cases, 1 Based on industrial standard tests for product qualification, as in „Semikron Application
practical examples will be shown in order to
Manual“ [1]
overcome these issues.
2 From ROHM qualification of discrete parts, following JEITA ED-4701.
Table 1: Qualification tests of switches based on silicon and SiC materials.
Bodo´s Power Systems®
November 2016
electrical engineering soft ware
Fast simulation of complex systems
Code generation
Frequency analysis
Available as standalone program
or Simulink blockset
Get a free test licen
w w w.plexim.com/tri
In order to turn the MOSFET on, voltage is applied between gate and
source. As consequence, electrical field occurs inside the gate oxide.
In order to predict the lifetime of gate oxide, an accelerated test is
performed, where higher voltage is applied to the gate. By collecting
certain amount of points with different gate voltages, it is possible
to extrapolate the gate oxide lifetime when typical gate voltage is
used. Figure 3 depicts results from accelerated test of 3rd Gen SiC
MOSFET from ROHM. It is possible to see that, for recommended
gate values, lifetime of gate oxide is expected to be much longer than
typical requirements industrial applications.
some ultrafast IGBT devices have also none or very reduced SCWT.
Nonetheless, they have been widely used in industrial applications.
Robustness against Cosmic radiation
The phenomenon of cosmic radiation has been well described in literature [3]. It can represent a critical issue, especially for applications
which operate at voltage levels very close to the nominal voltage of
the switch. This is the case for instance of photovoltaic inverters and
UPS systems.
Failures due to cosmic radiation occur randomly during the lifetime
of a semiconductor device. In accelerated tests, it is possible to
predict the failure in time (FIT) rate of a device. This value is normally
increasing exponentially with the blocking voltage, as shown in the
test results in Figure 4. It shows the comparison between devices
from silicon, namely IGBTs and MOSFETs, and SiC MOSFETs from
ROHM semiconductor, all rated for same current and voltage. It is
possible to observe that SiC MOSFETs have a FIT rate up to 3 orders
of magnitude lower than silicon counterparts.
Figure 2: Double trench structure and simulated electrical field, compared to conventional trench
Considering further events, like short-circuit and avalanche, they have
been proved through laboratory tests of SiC MOSFETs at ROHM.
Definition of parameters like short-circuit withstand time (SCWT) and
avalanche energy is now under preparation.
Figure 3 – Accelerated gate oxide lifetime test of 3rd Gen SiC MOSFET from ROHM, at Tj=175 °C.
In the meantime short-circuit and avalanche shall be avoided, or
solved in system level. This is already done for instance with silicon
IGBTs, that inherently do not have avalanche robustness. Moreover,
Bodo´s Power Systems®
Figure 4 - FIT rate due to cosmic radiation for same current, same
voltage rated devices from silicon (IGBTs and MOSFETs) and SiC
MOSFETs. [3]
The reason of this huge improvement is due to some characteristics
of SiC MOSFETs, which guarantee their robustness against cosmic
radiation. Among them there is the smaller chip area of the device
for same current rate and the higher margin of electrical field during
blocking state, with respect to breakdown field of the material.
Price-performance ratio
It is well known that the production cost of silicon carbide wafers is
comparatively higher than those out of silicon. This is related, among
other factors, to the energy involved in the growing process. This cost
difference is expected to reduce sensitively in next years, driven by
factors like:
• Increase of wafer size;
• Reduction of defect density on wafer level and those caused by
• Economy of scale, following the increasing adoption of SiC
Even facing higher price per device, it is already possible to build up
systems based on SiC which offer economic advantages. Next section will discuss two examples of systems whose cost can be reduced
by introducing SiC switches.
Auxiliary power supply
Industrial systems need an auxiliary power supply, which provides
November 2016
interference (EMI) in the frequency range starting from 150 kHz. Frequency range and emission limits in dBµV are depicted in Figure 5.
energy required to feed peripheral components, like micro-controller,
gate drivers, fans, LCD displays, etc. For this application, Flyback
converter is typically used. In 3-phase systems, where the input
voltage can reach 480 Vac phase-to-phase, devices with blocking
voltages above 1000 V are normally required.
Let a photovoltaic inverter be considered, operating under 20 kHz,
typical frequency for silicon IGBTs above 600 V blocking capability.
The squared waveform voltage coming from the inverter needs to be
filtered before reaching the grid. This waveform can be described according Fourier transformation as a series of sinusoidal waves, whose
frequencies are the fundamental i.e. 20 hHz, and the odd harmonics,
i.e. 60 kHz, 100 kHz, and so on. It is possible to demonstrate then,
that the first frequency to fall into the regulated range is the 9th harmonic, which is naturally damped by almost 20 dBµV. This has been
well depicted in [5], and is reproduced in Figure 6.
Before availability of SiC switches, auxiliary systems were implemented with silicon MOSFETs and BJTs of up to 1500 V blocking capability. These devices present big chip area and still high on resistance.
This results in high static and high dynamic losses, making the use of
heat-sinks almost mandatory and increasing system cost.
With the use SiC, 1700 V rated MOSFETs offer on-resistance as low
as 1 Ω. In addition, due to small chip size, dynamic losses are sensitively reduced. As a consequence, it is possible to operate at switching frequencies up to 120 kHz ‑ thus reducing transformer size, - and
still avoid the use of any heat-sink. This is demonstrated in [4], which
also introduces the quasi-resonant controller for Flyback converters
from ROHM.
Industrial power supply
Switched mode power supplies (SMPS) are widely used in industry to
feed DC loads, like low voltage batteries. These systems are typically
composed of two stages: an AC-DC rectifier followed by an isolated
DC-DC converter.
Figure 5 shows two industrial power supplies, disposed side by side.
The system on the left is based on silicon IGBTs. It operates with
switching frequencies of 20 kHz (AC-DC) and 40 kHz (DC-DC). Its
achieved efficiency is 83% at the nominal power of 20 kW.
On the right side there is the new system, based on SiC MOSFETs.
The much lower dynamic losses of these devices allowed switched
frequency increase. The AC-DC stage operates at 40 kHz, and the
DC-DC at 150 kHz. Considering all the consequent reduction in magnetic components, total system volume is 40 % smaller than reference
system. Furthermore, there was no system cost increase, while the
nominal power could be increases 30%, up to 26 kW.
150 kHz
Figure 6 – Filtering effort for a grid-connected photovoltaic inverter, for
different switching frequencies. As reference, the standard CISPR11
“B“ [5] modified.
If frequency is now increased to f.i. 100 kHz, the 3rd harmonic will
already fall into the regulated area. As a result, there will be around
18 dBµV less natural damping, which must be compensated by passive filter. In order to avoid system cost increase due to extra filtering
effort, it is recommended to operate at
Load switching frequency just below 50 kHz
(3rd harmonic out of regulated range)
or still slightly below 150 kHz (1st
harmonic out of regulated range).
20 kHz
40 kHz
50 kHz
150 KHz 3
20 kW
26 kW
2.5 losses
If higher switching frequency is desired, according to [5] the next sweet
spot will fall above 400 kHz. Above this
value, the size of the passive filter will
be again smaller.
Parasitic elements as stray inductances, capacitances and resistances from
the surrounding may block operation
Figure 5 – comparison between SMPSs made out of silicon IGBT (left) and SiC MOSFETS (right).
at such high frequencies. They may
Table aside shows circuit parameters and most important achievements with SiC based system.
come, for instance, from packaging, isolation foils, capacitor series resistance, PCB design, among
Electromagnetic Compatibility
others. In order to overcome this issue, resonant topologies maybe
SiC switches is to allow high frequency, high speed switching. This
an interesting option. Circuits like series resonant LLC [6] avoid hard
kind of operation results in very low dynamic losses, but can also
switching operation, reducing thus the emitted electromagnetic radiabring some challenges in accomplishing electromagnetic compatibility
tion. In addition, switching speed can also be controlled. This makes
(EMC) requirements.
the voltage waveform more trapezoidal-like, and increases the natural
dumping of the high frequency harmonic content.
The standard CISPR11, - and specially its “B” version, dedicated to
grid connected systems – establishes limits for the electromagnetic
Bodo´s Power Systems®
November 2016
SiC MOSFETs brings significant technical improvement with respect
to equivalent silicon devices. In addition, SiC MOSFET devices have
already achieved required quality level for most common industrial applications. This is proven by results of standard qualification process,
and accelerated tests of gate oxide lifetime. In addition, SiC MOSFETs have improved robustness against some events, like cosmic
Economic advantages brought by SiC MOSFETs have also been
demonstrated in some applications, even with the price difference as
it is today. In near future, factors like wafer increase and economy of
scale will decrease price of SiC devices and make it more and more
competitive and practically all applications inside electronic industry.
Against Terrestrial Neutron-Induced Single-Event Burnout in SiC
MOSFETs” IEEE Trans. Nucl. Sci., vol.61, 2014.
[4] Application Note “BD768xFJ-LB series Quasi-Resonant converter
Technical Design”. Available online.
[5] A. C. Schittler, D. Pappis, P. Zacharias: “EMI filter design for high
switching speed and frequency grid-connected inverters”, EPE
[6] Y. Nakakohara, et al “Three Phase LLC Series Resonant DC/DC
Converter Using SiC MOSFETs to Realize High Voltage and High
Frequency Operation”, IEEE Transactions On Industrial Electronics.
[1] „Semikron Application Manual“, 2nd edition, page 117. Available
online https://www.semikron.com/dl/service-support/downloads/
[2] “ROHM Semiconductor shows 3rd Generation SiC MOSFETs with
Trench Gate Structure”. Available online http://www.powerguru.
[3] Source : H. Asai, I Nashiyama, K Sugimoto, K Shiba, Y Sakaide,
Y Ishimaru, Y Okazaki ,K Noguchi, T Morimura “Tolerance
November 2016
Dr. Vladimir Scarpa
Field Application Engineer-Power
ROHM Semicondcutor GmbH
Bodo´s Power Systems®
Reverse Conducting IGCT
Platform optimized for Modular
Multilevel Converters (MMC)
ABB has developed a new Reverse Conducting
Integrated Gate Commutated Thyristor (RCIGCT) platform and recently the first products
such as the 4.5kV / 3kA and a 6.5kV / 2.15kA
RC-IGCT have reached the production stage. The
development was driven by high voltage converters for
STATCOM applications with Multi-level Modular Converter
(MMC) topology. The IGCT was chosen because of the low switching
frequency with subsequent potential for low losses, as well as the straightforward
inclusion of redundant cells for continuous operation after faults.
By Tobias Wikström, ABB Switzerland Ltd - Semiconductors
Performance / On-state
In an MMC application with comparably low switching frequency, the
on-state losses are crucial for the system’s efficiency. This is one of
the most important advantages of the IGCT over the IGBT for this
type of application.
Figure 2: Output of thermal simulations of the existing (left) and new
(right) RC-IGCT technology at identical power density in the active
parts. The maximal silicon temperature is 30°C lower using the newly
developed package. The color scale indicates the temperature in °C.
The heated (reddish) parts is the silicon device, stacked with molybdenum disks and copper pole-pieces. The cathode sides are directed
upwards. Both devices are stripped of thermally irrelevant features,
such as flanges, rubber, plastic and ceramic parts.
Figure 1: IGCT on-state voltages for the 4.5kV and 6.5kV IGCTs. The
“minimal” curves are what can be achieved without reducing the carrier lifetime
Figure 2 shows a simulation of the thermal response of the developed
RC-IGCT compared to the existing platform, at the same power level
and using the same color scale for the temperature. The new housing
omits the pole-piece trenches needed for conveying the gate signal,
and those that incorporate the gate spring system over the active
area. This has a profound effect on the maximal junction temperature
and especially on surge-current capability. Additionally, the wafer
was shifted from the axial center of the package, closer to the anode
thermal contact which reduces RTH for the device in general.
Bodo´s Power Systems®
Figure 3: Sample waveforms from a 4.5kV RC-IGCT, switching off
above 4500A at 3.2kV DC-Voltage
November 2016
The power to
The Power of Things™
You are reinventing the way individuals play and interact. You are designing the products
that will preserve natural resources and biodiversity. You are behind the technology that
will improve personal, home and corporate security.
While you are working on the applications that will shape the future, we are here to
support you with our proven power products and collaborative design approach.
Power Hall A2, Booth 613
Maximal controllable current
The HPT+ platform was incorporated in the new RC-IGCT for improving the high-temperature current controllability. Together with the
improvements in gate-circuit impedance – it was reduced by more
than a factor of three – the current-handling capability of the system
could be significantly improved. Both the diode and GCT parts can be
switched much beyond the SOA-specification at 135°C as shown in
Figure 3. At lower junction temperatures, the maximum controllable
current increases to over 6 kA.
The diode is extremely robust. It is governed by the current rate of
change during reverse recovery. In IGCT applications, the dI/dt is
moderated by dimensioning a reactor together with the maximal cell
voltage. For an appropriate selection of cell DC voltage and size of
the choke, there is almost no limit to how much forward current the
diode can recover from. For these new products, the limit has not
even been established exactly, because it so much higher than the
specified requirement, that the actual capability is of academic interest only.
A challenging task is to ensure nice behavior at low forward current
and high cell voltage over the whole temperature range. The most
straight-forward way to improve the switching behavior is to increase
the diode thickness, albeit with obvious loss drawbacks. It is particularly important for RC-IGCTs to minimize the diode thickness, due to
the integration of both the switch and the diode on the same Silicon
wafer. In order to keep snap-off within acceptable limits at a maximal
cell voltage of 3.2kV, it was necessary to use a double proton peak for
the 4.5kV device. The 6.5kV device prevailed with the single proton
peak, with the maximal cell voltage of 4.6kV.
As a new device platform, the devices have successfully passed the
full range of accelerated reliability testing such as temperature- and
load cycling. Hence, ABB is confident that the newly developed technology platform will display the same excellent reliability track-record
as its predecessor.
Including an adequately large number of levels in an MMC system
leads to a virtual disconnection of the higher-order harmonic output
from the switching frequency of the semiconductors. In theory, this
can be as low as the output’s fundamental frequency, which is 50
or 60 Hz in grid applications. In contrast, for inverter topologies with
fewer levels, the harmonic content decreases with increased semiconductor
switching frequency. The possibility of
low switching frequency favored the
selection of the IGCT as the inverter’s
main switch, thanks to its inherent
low on-state voltage compared to the
IGBT. Another significant advantage
of the hermetic press-pack IGCT is
that it can potentially handle large fault
currents without rupturing. Hence,
choosing the IGCT also facilitated
securing the cell integrity, including
safety of equipment, at very low extra
effort and cost.
Figure 4: The Power-Electronic Building Block containing two MMC cells
with four RC-IGCTs each
Bodo´s Power Systems®
The existing RC-IGCT platform was developed further as both the
size of the diode as well as the current handling capability needed increasing for the intended application. A positive side-effect of platform
development is the opportunity to review all design aspects. In that
endeavor, it was possible to improve on existing thermal bottlenecks
that impede particularly the surge-current capability. One of the basic
requirements was maintaining the backwards mechanical compatibility to the existing platform. As a result, the new device is identical on
the outside, with the same outer dimensions as its predecessor.
The incorporation of an outer ring gate structure in combination with
the integration of HPT+ technology on the RC-IGCT platform addressed both future requirements on current handling capability as
well as performance at high junction temperatures. The gate circuit
impedance was reduced further by wrapping the gate leads close to
the cathode pole-piece. The gate contact was moved to the periphery of the device. The move increases the area consumption of the
gate contact infrastructure by necessity, if nothing else is changed.
To counteract the loss of active area, the centering tolerance of the
gate contact on the wafer was much improved. Thus, despite the new
contact has a larger diameter, the space consumption for the gate infrastructure could be reduced. Combined with increasing the maximal
device diameter capability of the production facility, the active area
could be significantly increased without increasing the size of the raw
silicon wafer. All of the active area gain was given to the integrated diode, which almost doubled in size. Nevertheless, the improvements in
wafer technology, lowering the gate circuit impedance and removing
thermal bottlenecks, the IGCT performance could be improved too,
without increasing the size of the IGCT part.
The 4.5kV diode receives local lifetime control in two axial positions;
the 6.5kV in just one position, due to the switching behavior at low
current and high voltage. The double-position lifetime control has a
drawback: it limits the minimal on-state that can be achieved, so it will
be a target for future developments to incorporate a different technology, for example FCE, to improve the switching behavior of the 4.5kV
A 4.5kV / 3kA and a 6.5kV / 2.15kA RC-IGCT optimized for low
frequency applications, such as grid applications employing the MMC
topology. In order to fulfil the requirements, a new technology platform
for RC-IGCTs was developed. This platform will serve as the base for
future technology large area IGCTs from ABB, also for applications
with higher switching frequency.
November 2016
About the author
Tobias Wikström has developed power semiconductors for ABB since 2001. As a matter of
course, he has specialized on IGCT technology,
and was the development project manager for
the new RC-IGCT platform until its transfer to
Digital Telemetry
& Programming
3-18V, 12A Regulator with I2C
120k 80.6k
470 nF
I2C Bus
22 nF
Key Highlights
• 5% Accuracy Output Voltage and
• Enterprise SSDs
Output Current Monitoring
• Networking
• Programmable Current Limit
• Servers
• Programmable Frequency
• Set-Top Boxes
• Selectable PWM/PFM Mode
• Smart TVs
• 1% Internal Reference Accuracy
• I2C Programmable Output Range from 0.6V
Follow us:
© 2016 Monolithic Power Systems, Inc. Patents Protected. All rights reserved.
Technology Trends Raising
Power-Conversion Efficiency
The latest advances in superjunction MOSFETs and silicon-carbide rectifiers give
designers extra freedom to optimise performance and efficiency in cost-sensitive
power-conversion applications
By Michael Piela, Toshiba Electronics Europe GmbH
Power Supply Design Demands Efficiency Gains and More
In the drive to continue increasing energy efficiency in switching
power-conversion systems such as PFC and switching power supplies, superjunction MOSFETs and wide-bandgap silicon-carbide
(SiC) diodes have become favoured solutions for energy-conscious
designers. Both technologies have allowed smaller die sizes in relation to key parameters such as MOSFET on-resistance and diode
reverse voltage, enabling designers also to reduce circuit size and increase current density. As market adoption of these device technologies continues to grow, new demands are coming to the fore, such as
improved noise performance.
capacitance, superjunction MOSFETs also exhibit lower switching
losses than conventional silicon transistors.
Figure 2a shows the structure of early superjunction devices, which
have traditionally been fabricated using a multi-epitaxial process. Rich
doping of the N-region illustrated allows much lower on-resistance
than is achievable in conventional planar transistors. The P-type regions bounding the N channel are architected to achieve the desired
breakdown voltage.
Reducing electromagnetic noise emission is desirable in high-end
power supplies for equipment such as LCD TV, LED lighting, medical
power supply, notebook power adapters and power supplies for tablets. Resonant switching topologies, such as the LLC converter with
zero-voltage switching, are popular for these types of applications for
their inherently low electromagnetic emissions. Primary-side switching
in an LLC circuit as shown in figure 1 (MOSFETs Q1 and Q2), is often
now handled by superjunction transistors to achieve a compact and
energy-efficient power supply.
Figure 2a:
Figure 2b.
The N- and P-type structures of these devices have been fabricated
using multi-epitaxial processes that have resulted in dimensions that
are larger than ideal and have an associated impact on overall device
size. The nature of the multi-epitaxial fabrication also restricts engineering of the N-channel to minimise on-resistance.
Figure 1: Primary-side superjunction transistors boost the efficiency of
high-end LLC-resonant PSUs.
Superjunction Transistor Progress
The superjunction MOSFET has enabled power supply designers
to benefit from significantly lower conduction loss for a given die
size than is achievable using conventional planar silicon MOSFETs.
Because the device architecture also allows low gate charge and
Bodo´s Power Systems®
Figure 3: Single-epitaxial fabrication has enabled a flatter on-resistance/temperature characteristic. The TK12A60W represents DTMOS
IV and TK290A60Y DTMOS V generation.
November 2016
Improved fabrication processes, such as
deep trench filling that enables singleepitaxial fabrication, now give designers
greater freedom to optimise the aspect ratio
of N- and P-channels and so further minimise
on-resistance while also reducing MOSFET
size. Figure 2b illustrates Toshiba’s fourthgeneration DTMOS IV family, which takes
advantage of single epitaxy to achieve a 27%
reduction in device pitch at the same time as
reducing on-resistance per die area by 30%.
Also DTMOS V is based on the deep trench
process, with further improvements at cell
structure level.
The single-epitaxial process also enables
superjunction MOSFETs to deliver more
stable performance in relation to temperature
change. Ultimately, this helps to counter the
typical reduction of efficiency experienced in
power converters at higher operating temperatures. Figure 3 shows how the temperaturerelated change in normalised on-resistance
is significantly reduced in devices using the
latest-generation technologies, resulting in
12% lower on-resistance at 150°C.
DTMOS V FETs Meet Demands for Lower
With the arrival of fifth-generation DTMOS
V devices, designers can now choose
superjunction MOSFETs that deliver lownoise performance suitable for use in power
converters. DTMOS V FETs also display a
well-balanced ratio of lower noise performance and switching performance. This is
achieved through a modified gate structure
and patterning, which results in increased
reverse transfer capacitance seen between
the gate and drain (CRSS or CGD).
Emitted noise is comparable to that experienced with competing low-EMI devices,
while at the same time the devices deliver
the superior on-resistance that characterises
superjunction technology. Figure 4 compares
the level of EMI emitted by fourth- and fifthgeneration N-channel, 0.38mΩ-class 600V
devices used in the PFC circuit of a television
power supply, showing a significant reduction
interference from the later technology.
Rectifier Diodes Toughen Up with SiC
Complementing the high efficiency and current density of deep-trench superjunction
power switches, new generations of silicon
carbide (SiC) diodes combine inherently
superior energy efficiency compared to standard silicon devices with increased current
density, higher current ratings and greater
robustness, and enhanced cost-performance
Recap on SiC Advantages
The properties of silicon carbide (SiC) enable SiC Schottky Barrier Diodes (SBDs) to
deliver fast and temperature-stable reverserecovery comparable to that of conventional
silicon SBDs, which ensures energy-efficient
turn-off performance, without suffering the
conventional SBD’s relatively high and
temperature-dependent leakage current that
can result in thermal instability if reversevoltage derating is not applied. In addition,
the wide-bandgap property of SiC allows the
device to have a higher voltage rating in relation to die size, enabling 650V and 1200V
devices to be housed in industry-standard
surface-mount and through-hole packages.
This combination of characteristics makes
SiC diodes ideal for applications such as
power-factor correction when used as shown
in figure 5, in conjunction with a high-speed
superjunction MOSFET such as a DTMOS IV
X-type device.
We invented the MANGANIN® resistance alloy 125 years ago. To this day, we produce the MANGANIN® used
in our resistors by ourselves.
More than 20 years ago, we patented the use of electron-beam welding for the production of resistors, laying the
foundation for the ISA-WELD® manufacturing technology (composite material of Cu-MANGANIN®-Cu). We were
the first to use this method to manufacture resistors. And for a long time, we were the only ones, too.
Today, we have a wealth of expertise based on countless projects on behalf of our customers. The automotive
industry’s high standards were the driving force behind the continuous advancement of our BVx resistors. For
years, we have also been leveraging this experience to develop successful industrial applications.
The result: resistors that provide unbeatable excellent performance, outstanding thermal
characteristics and impressive value for money.
Electronica 2016
Messe München // November 8 - 11 // Hall B5 // Booth 143
Innovation by Tradition
Isabellenhütte Heusler GmbH & Co. KG
Eibacher Weg 3 – 5 · 35683 Dillenburg ·Phone +49 (0) 2771 934-0 · Fax +49 (0) 2771 23030
[email protected] · www.isabellenhuette.de
November 2016
Bodo´s Power Systems®
Figures 6a and 6b illustrate the enhanced architecture of the SiC SBD
in comparison with the standard silicon SBD architecture.
Figure 4: Improved noise performance displayed by fifth-generation
superjunction technology
To increase the surge-current capability and hence deliver more
robust devices for power-switching applications, the first-generation
architecture has been modified to minimise modulation of the conductivity (as measured using the diode forward-voltage, VF) thereby allowing higher maximum forward surge current, IFSM. Figure 7 shows
how this has been achieved by optimising the area of the P+ region.
Changes to the diode architecture have modified the relation between
current density and VF, raising the voltage at which conductivity modulation begins to occur, as shown in figure 8. This allows the device
to have higher IFSM. As a result, the second-generation architecture
permits IFSM to be increased above the reach of first-generation
Figure 5. The latest SiC diode technology can be used in conjunction
with a high-speed superjunction MOSFET, to boost the efficiency of
PFC circuitry.
The Emerging Generation
The key targets for the latest generation of 650V SiC SBDs have
been to increase performance in relation to device cost, and to raise
the maximum forward-current surge capability and thus deliver more
robust devices that are capable of surviving harsh exception conditions.
As with LSI semiconductors, power-semiconductor die size is a key
determinant of device cost. Development of the second-generation
SiC SBD architecture has focused on reducing the die thickness.
The result has been to reduce thickness by two-thirds, bringing an
attendant cost saving, while also raising current density by a factor of
up to 1.5.
Figure 8: Conductivity modulation starts at a higher VF in secondgeneration devices
Power supply designers are under pressure to satisfy unrelenting
demands for greater energy efficiency, reliability and miniaturisation,
within increasingly tight cost constraints. Moreover there is less time
available to look at EMI suppression during the design process.
Success depends on taking advantage of the latest power-semiconductor technologies that deliver lower on-resistance and noise
performance in the case of power MOSFETs, and reduced leakage
with greater temperature stability in the case of rectifier diodes. The
latest-generation superjunction MOSFETs and SiC diodes deliver
these advances, as well as improved switching performance, greater
robustness and reliability, and increased current density, at a price
that can make economic sense for cost-sensitive applications.
Figure 6a: Basic architecture of standard silicon SBD
Figure 6b: Architecture of SiC SBD
Figure 7: Optimising the SiC P+
region in the second-generation
650V SiC SBD
Bodo´s Power Systems®
November 2016
Identification of PMSM Motor
Parameters with a Power Analyzer
By Kunihisa Kubota, Hajime Yoda, Hiroki Kobayashi, and Shinya Takiguchi
HIOKI E.E. Corporation
1. Introduction
Recent years have seen permanent magnet synchronous motors
(PMSMs) and related control technologies rapidly permeate into the
advanced power electronics landscape and markets.
These developments reflect the advent of high-performance, high-efficiency designs thanks to progress in permanent magnet materials as
well as the advantages of PMSMs relative to other motors in terms of
quiet operation and simplicity of maintenance1). Recently, PMSMs are
being adopted in hybrid and electric vehicles in addition to household
electronics and industrial machinery, and their entry into widespread
use is expected to accelerate in the future2).
In general, PMSM analysis and control are based on the equivalent
circuit model for a motor expressed on the d- and q-axes. A variety of high-performance control methods have been proposed for
PMSMs, and these control algorithms are based on d-q equivalent
circuits, making it extremely important to identify the equivalent circuit
constants—in other words, the motor parameters (d-axis and q-axis
inductance, Ld and Lq)—with a high degree of precision.
Of these motor parameters, Lq exhibits a particularly high degree of
current dependence due to magnetic saturation3, 4), making it difficult
to implement high-performance control while using low-precision motor parameters measured in a simple manner with an LCR meter or
other instrument while the motor is in the stopped state.
In this equation, vd and vq represent the d-axis and q-axis components of the armature voltage for each phase; id and iq, the d-axis and
q-axis components of the armature current for each phase; R, the
armature resistance for each phase; p, the differential operator (d/dt);
Ld and Lq, the d-axis and q-axis self-inductance; ω, the rotation angle
(electrical angle) speed; and ϕa (=Ke), the RMS value of the permanent magnet’s flux linkage with the armature (i.e., the induced voltage
Fig. 2.1 illustrates the result of assuming a stationary state (so that
time-derivative terms can be ignored) and expressing Eq. (2.1) as a
d-axis and q-axis vector diagram. In the figure, v1 and i1 represent
the fundamental components of the phase voltage and phase current,
and θv and θi represent the fundamental phase angle of the phase
voltage and phase current, respectively. Based on Fig. 2.1, the d-axis
and q-axis voltage equations can be formulated as follows:
Solving these for LD and Lq yields the following equations:
This paper introduces a method by which a power analyzer can be
used to identify motor parameters easily and with a high degree of
precision while the target motor is operating. In addition, it provides
results (motor parameters) obtained through the actual use of this
2. Method for identifying motor parameters
This chapter provides a brief description of the principles employed
to identify PMSM motor parameters using a power analyzer and of a
procedure for doing so.
2.1 Principles
If we assume the following with regard to the voltage equation for a
PMSM expressed on the d-q coordinate axis, we arrive at Eq. (2.1)3) .
i) The spatial distribution of magnetic flux in the gap between the stator and rotor takes the form of a sine wave moving along the gap.
ii) The harmonic components of the voltage and current can be
iii) Core loss can be ignored.
Bodo´s Power Systems®
Figure 2.1: PMSM vector diagram
2.2 Identification procedure
This section describes a procedure by means of which a power analyzer can be used to identify motor parameters.
November 2016
Home Appliances
Power Transmission
One of our
key products:
LV100 and HV100 High Voltage
Power Modules
Power Devices from
Mitsubishi Electric.
LV100 and HV100: The new high voltage power modules of Mitsubishi Electric
for a safe and greener tomorrow. The newly developed dual module structure
is reducing the thermal stress applied to Si- and SiC-power chips, enabling a
low internal package inductance and allowing good scalability for flexible power
electronics solutions. Latest proved technologies are applied to satisfy reliable
operation and long life time requirements in demanding applications as Railway,
Wind generators and MV-drives.
More Information:
[email protected]
Scan and learn more
about this product
series on YouTube.
- IGBT chips with latest generation
CSTBT TM technology and RFC diodes
- Robust main terminals suitable for
doubling the rated module current in case
of using SiC-chips
- Eliminating of substrate solder by new MCB
(Metal Casting direct Bonding) technology
- High robustness/resistance against environmental influences due to the newly developed SCC (Surface Charge Control) process
- Module case material suitable for high
pollution degree and fire inhibition according
to EN45545
Although this specific procedure uses a Hioki Power Analyzer
PW6001, motor parameters can be identified using a similar procedure with any power analyzer that provides an electrical angle measurement function that is equivalent to that offered by the PW6001.
2.2.1 Measuring the armature resistance R for each phase
Measure the armature resistance R for each phase using a resistance
meter or other suitable instrument in advance.
2.2.2 Performing phase zero-adjustment and identifying the
induced voltage constant Ke
After placing the motor terminals of the PMSM being measured in the
open state (id = iq = 0), connect the motor terminals to the “CH 1”,
“CH 2” and “CH 3” voltage inputs of the Power Analyzer PW6001. Additionally, connect the encoder’s A-phase pulse output to “CH B”, its
B-phase pulse output to “CH C”, and its Z-phase pulse (origin signal)
output to “CH D” (Fig. 2.2).
2.2.3 Identifying the motor parameters Ld and Lq with user-defined functions
The d-axis and q-axis self-inductance Ld and Lq can be identified
using R as measured in Section 2.2.1 and Ke as identified in Section
2.2.2. First, connect the drive inverter output to the motor terminals
that were left open in Section 2.2.2 and operate the motor (Fig. 2.3).
At this time, the following equations will obtain based on Fig. 2.1:
By configuring the instrument’s user-defined functions (UDFs) with
these equations as well as Equations 2.4 and 2.5, it is a simple matter
to identify Ld and Lq while monitoring vd, vq, id, and iq. See reference5)
for specific examples of settings for the Power Analyzer PW6001’s
user-defined functions.
Configure the Power Analyzer PW6001’s settings by setting the motor
analysis operating mode to “Single,” the measurement parameter
to “Torque Speed Direction Origin,” and “CH B” input to “Pulse.” In
addition, set the wiring connection for “CH 1”, “CH 2” and “CH 3” to
“3P3W3M,” the synchronization source to “Ext1,” and Δ conversion to
“ON.” Setting the synchronization source to “Ext1” allows the voltage
and current phase angles to be measured using the inputted encoder
pulse as the reference, and setting Δ conversion to “ON” allows the
line voltage to be converted to, and measured as, a phase voltage.
In this state, drive the motor from the load side to generate an
induced voltage and perform phase zero-adjustment on the Power
Analyzer PW6001. As a result of this step, θv and θi will represent the
phase angle expressed using the phase of the induced voltage generated in the q-axis direction as the reference—that is, the electrical
At this time, Eq.(2.4) can be rewritten as follows since the induced
voltage vq is equal to v1, allowing identification of Ke.
Figure 2.3: Wiring connections when identifying the Ld and Lq motor
3. Measurement example
This section presents the results of using the procedure described in
Section 2.2 to actually identify motor parameters.
3.1 Measurement conditions
Tables 1, 2, and 3 describe the specifications of the inverter (Fig. 3.1),
drive-side motor, and load-side motor (Fig. 3.2) used in the procedure.
In this equation, f1(=ω/2π) represents the frequency of the phase voltage’s fundamental wave.
Rated output capacity
10.0 kVA
Rated output voltage
400 Vrms AC
Rated output current
14.5 Arms AC
Rated input voltage
700 V DC
Rated input current
15.1 A DC
Maximum input current
18.6 A DC
Input voltage range
0 V DC to 800 V DC
Switching frequency
Up to 200 kHZ
Switching element
Myway Plus Corp.
Table 1: Inverter specifications
Figure 2.2 Wiring connections when performing phase zero-adjustment and identifying the induced voltage constant Ke
Bodo´s Power Systems®
Table 4 describes the measuring instruments that were used. The
Hioki Resistance Meter RM3544 noted in the table was used to measure the armature resistance R of the drive-side motor listed in Table
2 for each phase (Section 2.2.1).
November 2016
Design & debug your power supply
in minutes, not weeks
Digital multiphase power solutions
for 10A to 450A applications
Easy-to-use digital multiphase controllers coupled with
our smart power stage technology provide scalable
solutions from 10A to 450A
• Full telemetry data, monitoring and control with
PMBus™ 1.3 and AVSBus™ support for smart, energyefficient systems
• Patented synthetic current control tracks phases
with zero latency to respond to any load with precise
current and voltage positioning, requiring 30% less
capacitance than competitive devices
• PowerNavigator™ GUI software tool allows you to
quickly optimize your configuration and easily
store it in non-volatile memory
Flexible configurations to meet any rail requirements
Dual Output
Output Phase
X+Y ≤ 7
X+Y ≤ 4
X+Y ≤ 7
X+Y ≤ 4
X+Y ≤ 7
X+Y ≤ 4
X+Y ≤ 7
X+Y ≤ 4
IMVP8 & VR13
X+Y ≤ 7
PMBus, VR13
PMBus, VR13
X+Y ≤ 4
PMBus, VR13
X+Y ≤ 4
General Purpose
For datasheets, samples, videos and more, visit
3.2 Identifying the induced voltage constant Ke
The induced voltage constant Ke was identified using the procedure
described in Section 2.2.2.
For reference, Fig. 3.3 illustrates the induced voltage (phase voltage)
waveforms for the drive-side motor and A/B/Z phase pulse waveforms
for the encoder during the identification process.
RM86A20-2-E8 DC brushless motor with encoder
Rated voltage
100 V DC
Rated current
Rated rpm
2500 rpm
Rated output
120 W
Armature resistance for each phase
0.89768 Ω
Number of poles
Number of pulses per rotation
Figure 3.3 Drive-side motor induced (phase) voltage and encoder’s
A/B/Z phase pulse waveforms during identification of the induced voltage constant Ke
Table 2: Drive-side motor specifications
SS60E80-6 DC motor
Rated voltage
100 V DC
Rated current
4.8 A
Rated rpm
2500 rpm
Rated output
350 W
Table 3: Load-side motor specifications
Figure 3.4: Relationships between the motor rpm n, the RMS value
v1 of the fundamental component of the drive-side motor induced
(phase) voltage, and the identified induced voltage constant Ke
Fig. 3.4 illustrates the relationships between the motor rpm n, the
RMS value v1 of the fundamental component of the drive-side motor
induced (phase) voltage, and the identified induced voltage constant
Ke. The measured v1 value varies proportionally with n, while the
identified Ke value remains roughly constant, without regard to n. In
this way, the relationships between these three values can be seen to
satisfy the relationships described in Eq.(2.6).
Figure 3.1: Inverter
Ke exhibits a small amount of variability during low-speed operation
due to the more pronounced rotating unbalance of the motor in that
operating regime.
Figure 3.2: Drive-side motor (left) and load-side motor (right)
Power Analyzer
Current Sensor
Resistance Meter
Table 4: Measuring instruments
Bodo´s Power Systems®
3.3 Identifying the Ld and Lq motor parameters
The d-axis and q-axis self-inductance Ld and Lq were identified using
the procedure described in Section 2.2.3. For reference, Fig. 3.5
illustrates the inverter’s secondary-side phase voltage and phase
current as well as the encoder’s A/B/Z phase pulse waveforms during
Fig. 3.6 illustrates the relationships between (a) the d-axis current id
and the identified d-axis self-inductance Ld and (b) the q-axis current
iq and the identified q-axis self-inductance Lq. Ld remains roughly
November 2016
constant, without regard to id. By contrast, Lq exhibits a high degree
of current dependency due to magnetic saturation and varies significantly with iq. These characteristics make it clear that it is not possible
to use an LCR meter or similar instrument to identify Ld with a high
degree of precision while the motor is in the stopped state. Instead,
the value must be identified while the motor is operating.
The variability in the Ld and Lq values when the id and iq values are
small is also likely to be caused by rotating unbalance of the motor
during low-speed operation.
Fig. 3.6 illustrates the results of identifying the Ld and Lq motor parameters while the motor’s rpm is varied while holding the current phase
angle constant, showing the current dependence of Ld and Lq. The
current phase angle dependence of the motor parameters can also be
verified by applying this identification method.
Figure 3.5: Inverter secondary-side phase voltage and phase current
and encoder’s A/B/Z phase pulse waveforms during identification
of the Ld and Lq motor parameters (when driving the motor with the
4. Conclusion
This paper has introduced a method for identifying PMSM motor
parameters easily and with a high degree of precision using a power
analyzer. It also presents the results of using the introduced method
along with a Hioki Power Analyzer PW6001 to identify actual motor
parameters. It must be noted that the method introduced in this paper
presumes the use of an analytical model that posits that core loss can
be ignored. That said, by measuring mechanical loss and identifying
the equivalent core loss resistance in advance, it would be possible
to further develop the described method in order to identify motor
parameters while taking into account core loss.
The identification of PMSM motor parameters introduced in this paper
is only one example of an application for power analyzers, which can
be used effectively in numerous other settings in the power electronics field. The authors look forward in the future to actively introducing
other applications in which power analyzers can be effectively.
Figure 3.6: Relationships between (a) the d-axis current id and the
identified d-axis self-inductance Ld (shown in red) and (b) the q-axis
current iq and the identified q-axis self-inductance Lq (shown in blue)
1) Shigeo Morimoto : “Trend of Permanent Magnet Sychronous Machines”,
IEEJ Trans, Vol.2 (2007), pp.101-108.
2) Investigating R&D Committee on industry applications of PM motors : “Trend
in the latest technologies and applications of permanent magnet synchronous
motors”, IEEJ Technical Report (2009), No.1145 (in Japanese).
3) Shigeo Morimoto, Yoji Takeda, and Takao Hirasa: “Method for Measuring a
PM Motor’s dq Equivalent Circuit Constants”, IEEJ Transactions on Industry
Applications, Vol.113-D (1993) No.11, pp.1330-1331 (in Japanese).
4) A. Soualmi, F. Dubas, D. Depernet, A. Randria and C. Espanet : “Inductances
estimation in the d-q axis for an interior permanent-magnet synchronous
machines with distributed windings”, Proc. XX ICEM (2012), pp.308-314.
5) HIOKI E. E. Corp. : “Identification of PMSM Parameters with the Power
Analyzer PW6001” (White paper), retrieved from https:// www.hioki.com/ en/
products/ detail/?product key= 5796.
www.hioki.com/ en/
November 2016
Bodo´s Power Systems®
Igniting the spark
The angular timer of a microcontroller can help implement capacitor discharge
ignition in an internal combustion engine
By Ashutosh Tiwari, Shailendra Vengurlekar and Namrata Dalvi, Microchip Technology
The capacitor discharge ignition (CDI) system in an internal combustion engine can be implemented on single profile ignition pickup (PIP)
systems using standard peripherals on an 8bit microcontroller, but
there are challenges associated with such designs.
In an internal combustion engine, the air and fuel chemical mixture is
burnt and extreme heat is generated expanding the exhaust gases,
which force the cylinder piston to move, causing the camshaft to
rotate and create a kinetic energy. This kinetic energy is coupled to
the vehicle’s wheels by gear trains to convert the angular motion into
linear motion.
the spark firing angle. To fire the spark correctly and accurately, a
separate module known as the ignition control mechanism, is used.
Ignition systems
There are two types of ignition systems: inductive discharge ignition
(IDI) or transistor controlled ignition (TCI); and CDI. The CDI system
uses high-voltage capacitor discharge current output to fire the spark
plug, see Figure 2. It can be implemented using core independent
peripherals (CIPs) found on Microchip PIC microcontrollers. These
include angular timer (AT), signal measurement timer (SMT), maths
accelerator and configurable logic cell (CLC).
A four-stroke cycle engine that uses four distinct piston strokes –
intake, compression, power and exhaust – to complete one operating
cycle is shown in Figure 1.
Figure 2: Basic capacitor discharge ignition (CDI) system
Figure 1: Four stroke engine with engine parts
The top dead centre (TDC) is the highest position of the piston near
the spark plug and the bottom dead centre (BDC) is the lowest
position near the camshaft. After the spark plug fires into the power
stroke, the air-fuel mixture needs time to completely burn; this burning
process is progressive in nature in that the mixture at the top burns
first and quickly moves towards the bottom.
Therefore, to burn the air-fuel mixture completely and produce the
maximum pressure wave, the spark plug should be precisely fired
moments before the piston reaches the TDC and at the proper angle,
which is determined by the engine piston speed. There are other
factors, such as temperature and throttle position that also determine
Bodo´s Power Systems®
There are two types of CDI system – alternating and direct current
capacitor discharge ignition, or AC-CDI and DC-CDI. In an AC-CDI
system, an alternator or stator (magneto) generates enough power
for all electronic systems including the CDI. The capacitor is charged
through the rectified output of the magneto AC supply, which is 200
to 400V DC. When the engine is cold (not running), a kick-start is
required to rotate both the engine and the magneto. This does not
generate sufficient power from the magneto to charge the capacitor completely for a high-voltage spark. For very low RPM, the firing
angle is always constant. Hence, analogue firing is used to fire at
negative PIP output from the magneto-flywheel pulser coil without
calculating the RPM.
In a DC-CDI system, a constant 12V DC power is always available
from the battery. It requires an additional DC-DC converter to raise
the 12V DC to 200-400V DC. This additional circuitry makes the CDI
module slightly larger than an AC-CDI system. When the engine is not
running, it can be started easily at a precisely calculated firing angle,
as the input DC supply is always available.
November 2016
Power Electronic Stacks
When Time to Market Counts
You want to...
- reduce time to market
- limit your financial investment
- control your cost of development
- save costs on manufacturing
of stack assemblies
We are the global market leader in power electronic stacks with
over 200,000 assemblies in the field. Five global stack centers
provide optimized, pre-qualified and field-tested stack assemblies.
Our stacks feature short lead times while still easy to customize
for your needs. Our global stack centers provide worldwide coordination, engineering, manufacturing and test to localize manufacturing of your stack assemblies.
Further information: www.semikron.com/power-stacks
To create the high-voltage spark in the spark plug, a high-voltage capacitor with a high-charge capacity is charged using either the output
of the DC-DC converter (DC-CDI) or using the output of the magneto,
an AC alternator (AC-CDI). The capacitor is charged to a high-voltage
supply, usually 200 to 400V.
The capacitor is connected to an ignition coil or step-up pulse transformer, which produces a very high voltage, in the range of 40kV or
The switch connects the capacitor to the primary of the ignition coil.
The switch is fired when the microcontroller gives a pulse at the gate
of the switch. The sudden rush of current in the primary of the coil
produces a very high voltage in the secondary, which generates the
spark to ignite the air-fuel mixture. Thus, the microcontroller controls
the firing angle of the switch for generating the spark.
A silicon-controlled rectifier (SCR) is most commonly used as a highpower switch in CDI. It is highly durable due to the higher operating
voltages and current ranges with moderate frequency response. The
disadvantage of the SCR is it is a one-sided switch in that the switch
can only be on. It will automatically go to off when the input is less
than the lower operating threshold.
An IGBT and MOSFET are used in modern CDI designs due to
their ability to switch on and off, as well as having better frequency
response at higher operating ranges.
engine RPM. The second positive pulse can be a reference point for
deciding the firing angle for higher speeds.
The pulser coil generates the timing signals that contain positive
and negative pulses. These pulses are in the range of ±3 to ±90V,
depending on the magnetic field strength of the magnet mounted on
the flywheel.
A signal conditioning circuit inverts the negative pulse and limits the
pulses to 0 to 5V. It is also used to filter spurious noise. The signal
conditioning circuit will provide two positive outputs, one corresponding to positive pulses and another for negative pulses. Output of the
signal conditioning circuit is connected to the microcontroller.
In a digital CDI system, the microcontroller has two major functions:
deciding the advance firing angle by reading input from the sensors,
such as the pulser coil, thermistor and throttle position, then producing the firing pulse; and setting the duty cycle of pulse-width modulation (PWM) for the DC-DC converter.
The advance angle required for getting optimum performance from
the engine is mainly dependent on the RPM. Hence, the system
must be aware of the current RPM, temperature and throttle position.
Look-up tables called maps are stored in the programme memory of
the microcontroller, which give the appropriate advance angles with
respect to the RPM.
The pulser coil – or pick-up and timing coil – is responsible for providing the timing signal to the ignition control system.
A magnet is mounted on a flywheel, which is mounted on the magneto
shaft. When the flywheel rotates, the magnet passes near the pulse
coil producing a timing pulse. There is one pulse per pole. Hence, for
each magnet there are two outputs, one positive pulse followed by
a negative pulse, generating one alternating pulse pair. For a single
PIP system, there is only one pair. For multi-pulse systems there are
multiple pulse pairs, based on the number of magnets on the flywheel.
The alternating pulses are at a fixed angle with respect to the TDC
piston position in the engine for each rotation. The period of the
pulses triggers the rotation of the engine.
Based on the number of alternating pulses from the pick-up, per engine rotation, the pulser coil system can be either a single or multiple
PIP system.
In a single PIP system, the pulser coil provides one positive pulse
followed by a negative reference pulse. The angle between the pulses
and the angle between the negative pulse and the TDC are fixed. The
firing of the spark should be at a specific angle based on the engine
operating temperature, the throttle position and the RPM. This angle
is usually between the positive and negative pulses.
The negative pulse is a reference point for firing the spark at very low
RPM. The firing angle changes to an angle between the pulses for
higher RPM.
In a multiple PIP system, the pulser coil provides more than one alternating pulse. The second negative pulse is the reference point before
the TDC. This is the minimum firing angle at which the spark should
be generated for engine speeds below a lower speed threshold.
The first positive and negative pulses can be used for calculating the
Bodo´s Power Systems®
Figure 3: Conventional method for CDI implementation
Multiple maps are stored based on different throttle positions and for
different temperature ranges. Once the controller calculates the RPM,
it can then look up the appropriate advance angle from the map.
Peripherals on PIC MCUs, such as capture compare or PWM (CCP)
and ADC, together with the interrupt pin INT can be used to determine
the firing angle control in CDI.
The output for the positive PIP signal of the signal conditioning circuit
is given to the capture module. The capture module measures the
time between two positive pulses, which is the period of the pulser coil
output. The period of the pulses gives the RPM of the engine.
An analogue-digital converter (ADC) can be used to determine the engine temperature and the throttle position if it is analogue; the throttle
position input can be either analogue or digital. In the case of a digital
throttle, the position for the wide open throttle (WOT) is one state and
the partially open throttle (POT) another. There are distinct firing maps
for different throttle positions and different temperature ranges. Figure
November 2016
3 shows the conventional method for the CDI
The conventional method of RPM calculation
uses a 16bit timer along with capture peripheral. For lower RPM values, less than 60
(corresponding to 1Hz frequency), the 16bit
timer will overflow if the timer clock frequency
is 1MHz. The timer overflow bit should be
taken into account for RPM calculations.
conventional method or by using CIPs, such
as AT, CLC, SMT, maths accelerator and
complementary waveform generator (CWG).
However, using the CIPs greatly improve the
overall performance and implementation of
The AT successfully divides the input signal
to angular division without CPU intervention,
which helps to boost performance by remov-
Figure 4: AT block diagram in single-pulse mode
The PIC16F161X 8bit microcontroller has a
Core Independent Peripheral (CIP) called
the angular timer (AT), which can be used in
internal combustion engines to fire the spark
at the exact firing angle with very little CPU
intervention. As shown in Fig. 4, the periodic
pulse input to the AT can be selected either
from the internal core independent peripherals or from the external pin.
The CIPs of PIC microcontrollers, such as
CLC, AT, SMT and maths accelerator, are
used for the firing angle control in CDI.
For the RPM calculation, a 24bit SMT is
used. The SMT is configured in windowed
measurement mode with the window input
set to AT period pulse. Whenever the AT
gives a period pulse, the SMT captures the
timer value into a register, resets its timer
count and restarts counting. The timer capture into the register generates a captured
period interrupt.
ing the need for firing angle conversion from
degree to equivalent time. These angular
divisions are also very accurate and constant
throughout the input signal range.
The performance of the CDI system can
be greatly enhanced using the AT, with significant CPU bandwidth remaining for other
calculations. In addition, with the use of the
maths accelerator, the calculations are now
more accurate and faster. The SMT with its
inherent high-bit resolution helps in tracking
low engine RPM and taking necessary action, without the need for large computations.
A similar implementation technique can be
extended to other systems, such as inductive
discharge ignition, that use the same CIPs.
Ashutosh Tiwari is ‎Senior Application
Engineer at Microchip Technology
The maths accelerator peripheral, also called
a PID module, is used to calculate the firing
angle for the current RPM.
Shailendra Vengurlekar is ‎Manager
Automotive Applications at Microchip
CDI systems can be implemented using
PIC16F microcontrollers either through the
Namrata Dalvi is ‎Senior Application Engineer at Microchip Technology
November 2016
A Fast Track to Complex
Power System Designs
Innovative power conversion architectures, taking advantage of evolving power
semiconductor technologies, have risen to the challenge, especially in terms of the high
performance modular solutions now available.
By Arthur Jordan - Applications Engineer, Vicor Corporation
Recent developments in power system components have enabled
the designer to configure power supply systems of vastly greater
performance than was possible only a short time ago. This has been
driven in part by an acute need; today’s systems are built around
microprocessors and FPGAs that demand ever-higher currents at an
unprecedented variety of lower voltages. The circuit boards on which
those components reside require multiple power rails, of high stability,
at high power levels, and often with complex sequencing constraints.
Scale up to the rack and system view, and the problem becomes
infinitely more complex.
Innovative power conversion architectures, taking advantage of evolving power semiconductor technologies, have risen to the challenge,
especially in terms of the high performance modular solutions now
available. The designer’s task has not, however, been made any simpler because those same advances in technology have expanded the
options available with which to configure a system, and to meet any
given specification. The range of modules available today have made
a change that appears subtle, but can be radical in its impact. They
have separated out the basic functions of a power supply, such as
voltage level conversion, regulation, and isolation. With the freedom
to place these functions as-needed through a power supply chain,
comes the need to explore many more options to find the optimum
More choices mean more flexibility to optimize
Innovative devices such as Vicor’s BCM provide an example of how
design choices are expanded. A BCM is a low-voltage DC-DC bus
converter module; using Vicor’s Sine Amplitude Conversion technology, it provides bi-directional fixed-ratio conversion at very high
efficiency. Packaged in the IC-style ChiP, or the easy-to-use VIA
packaging technology, the BCM gives the power engineer an isolated
DC-DC conversion function. Regulation can be done upstream of the
Figure 1: A Typical BCM Application
Bodo´s Power Systems®
BCM; feed it a stable 48V and it will directly supply a low-voltage rail
on a PCB with no further regulation required. Or, it can down-convert
a rail that has less precise regulation to a low voltage, leaving individual point-of-load (PoL) regulators to supply power to what might be
a multiplicity of on-board rails.
The BCM takes its place in a large portfolio of power components,
such as the VTM, a point-of-load current multiplier (again, fixed-ratio
conversion); PRM, a regulator for use in Factorized Power architecture, that would feed a VTM to generate PoL voltages; and the DCM,
an isolated, regulated DC-DC converter. These examples are part of
a broad portfolio of regulated and non-regulated, isolated or non-isolated, functional blocks that can be assembled in a virtually limitless
number of ways.
For the power engineer, the question becomes less, “can I achieve
the necessary power provision?” and much more, “how do I determine
the optimal solution for my power design, given the comprehensive
array of options?”
In the past, design choices were often rapidly narrowed by product
performance and by accepted practice. For example, a 48V distribution rail is frequently a standard feature of a power distribution
scheme. However, with earlier generations of buck converters, conversion efficiency would typically decline sharply at high conversion
ratios (from 48V to single-figure voltages). Therefore, and with PCBs
demanding more and more rails at voltages down to – or lower than –
1V, a further distribution level of, perhaps 12V might be interposed.
What is the cost of optimization?
With Vicor’s power components, however, converting from 48V
direct to PoL is not only feasible, it can be the most efficient option
when the losses associated with generating an intermediate bus are
eliminated. It does not take a great leap of imagination to observe that
the ‘universe’ of feasible solutions has expanded enormously. It has
become an optimisation problem; one with many dimensions and, in
all likelihood, multiple candidate solutions.
To achieve optimal efficiency, the power engineer has had no option
but to turn to the data sheets. Every data sheet presents page after
page of charts showing operating parameters and performance of
its particular device. Having decided on a suitable power system
topology, the designer has to work through it, establishing the operating point of each module, look up their conversion efficiencies and
losses, sum them by pencil-and-paper or by spreadsheet – and then
repeat the exercise for every alternative configuration that looks like a
realistic option.
November 2016
flowPACK 1 H6.5
650 V/50 A-75 A-100 A
Can highly efficient single-phase solar applications benefit from
a three-level topology? They can with this smart new alternative.
Vincotech‘s new flowPACK 1 H6.5 offers a very persuasive alternative for
single-phase solar applications. Housed in a 12 mm, 4-tower flow 1 package,
this power module features an LVRT-enabled chipset optimized for up to 25 kHz.
Main benefits
/ Engineered for single-phase solar applications
/ Features innovative H6.5 topology
/ Provides LVRT (low voltage ride through)
/ Equipped with an IGBT S5
/ Chipset optimized for switching frequencies up to 25 kHz
/ Integrated NTC
c o t e c h .c
w w w.v in
-sam le
If standby power is a consideration: repeat the process using the
quiescent power figures for every block. The option, noted above,
to place the fundamental functions of power conversion at-will along
the conversion chain offers a further degree of freedom. To a certain
extent, and given that some losses/heat dissipation are inevitable,
the engineer may adjust the supply topology to have those losses
concentrated where they are most easily dealt with.
When all of that is completed, the designer will only have explored
efficiency and losses. Other constraints may very well apply, not
least, cost. Alternative configurations that perform to specification in
electrical terms may be significantly different in bill-of-materials terms.
Another aspect of power supply design that has hardly changed over
many generations is that power provisioning may be left to fit in the
cabinet or rack space remaining when design of every other aspect is
complete. Therefore, physical footprint can become critical; not only of
major blocks such as off-line AC-DC conversion, but of point-of-load
regulators and any associated passives in their footprints, distributed
around PCBs.
A better Way to design power systems
It was against this background that Vicor created its online Power
System Designer software package, designed to allow an engineer
to create a power system in minutes. The Power System Designer is
a simple-to-use, parametric-input tool that leverages Vicor’s Power
Component Design Methodology and delivers optimized power fast.
Using Vicor’s high-performance power components, the tool asks the
designer to define a system in the simplest terms; input and outputs.
Using that data, it automatically generates a complete power supply
configuration, together with all relevant operating parameters.
Customize your solution based on you key figures of merit
Given AC or DC input source and operating range, together with
required output voltages and respective power (or current), regulation
and isolation specifications, the software automatically identifies not
only a “best” solution, but presents a range of alternative solutions,
each one accompanied by a spread of figures-of-merit. Efficiency is
presented not only as an overall conversion efficiency, but broken out
into front-end and PoL efficiencies. Other parameters include:
• lowest component count
• lowest cost (cost is presented for 1-off and 500-unit volumes)
• smallest footprint
• recommended best fit.
Once again, the designer can consider options for “moving” footprint
(physical area) along the power conversion chain as the tool returns
total, front-end and PoL footprint figures for each solution. For any
selected design, the Power System Designer can display a visual
representation of the mechanical layout of the system and generate a
complete Bill-of-Materials along with ordering and pricing information.
A more holistic, faster approach to designing your power system
The Power System Designer also reports the power utilization of the
front-end and of each PoL output – that is, how much of the capacity
of each component specified in the chosen solution, is actually being
used. The engineer thereby gains a detailed view of both design
margin and capacity utilization of a given design. For more detailed
analysis, launching Vicor Whiteboard is only a single-click away.
Figure 2: Power System Designer
Figure 3: A Typical Output
No other tool provides such simplicity of use nor the same complete
end-to-end (Source to PoL) coverage. Although the Power System
Designer is not a full circuit simulator, it is backed by comprehensive
and detailed simulation of every Vicor power component. It is closer
to a behavioural model-based approach, where the detailed circuit
operation of every component and module has been simulated and/or
measured in real-world operation and tabulated, in a form the tool can
rapidly access.
The Power System Designer is a significant advance over simulators
and design aids that have been offered up until now, in that it takes
a holistic view of the power provision function from input to output,
rather than focussing on front-end parameters or point-of-load performance in isolation. Allowing exploration of the full range of solutions
possible with Vicor’s comprehensive families of field-proven, high
density modular power components, that are capable of addressing
virtually any front end or power train requirement, fast-tracks the route
to superior power density and quicker time-to-market, while minimizing design risks.
Added to that is a degree of rule-based behaviour; in the broad span
of solutions available with Vicor’s power components, there are
some choices that are inevitably more productive than others, and
that knowledge is built into the software. Due its mode of operation,
the Power System Designer is fast and precise, delivering results
instantaneously. When the time comes to verify a solution with a more
detailed parametric analysis, the package acts as a “front-end” to
Vicor’s fully-editable Whiteboard tool.
Bodo´s Power Systems®
November 2016
Advanced and Broad Power
Management Portfolio
Microchip Technology has an expansive offering of power management solutions
to fit virtually every type of design criteria. From the smallest form factor needed for
mobile devices to complex industrial power management designs to automotive
standards, you are sure to find a highly integrated solution to meet your needs. If
you are looking for greater flexibility in your design, Microchip’s digitally enhanced
power analog devices integrate a microcontroller (MCU) or digital signal controller
(DSC) for a fully programmable and flexible solution.
LDO and switching regulators
System supervisors
Charge pump DC/DC converters
Voltage detectors
Power MOSFET drivers
Voltage references
Digitally enhanced and
Li-Ion/Li-Polymer battery chargers
PWM controllers
The Microchip name and logo and the Microchip logo are registered trademarks Microchip Technology Incorporated in the U.S.A. and other countries. All other trademarks are the property of their registered owners.
© 2016 Microchip Technology Inc. All rights reserved. DS20005389A. MEC2066Eng04/16
Embedded Designs Drive
Tomorrow’s Solutions
Embedded drive solutions for motion control simplify integration, enhance performance
and speed up time-to-market. Vincotech’s power modules lineup represents the best fit
for highly reliable, low-cost motor controls that deliver higher performance in a smaller
By Michele Portico, Product Marketing Manager, Vincotech GmbH, Unterhaching
Embedded drive solutions for Industrial motion control
Discrete drives are standard solutions designed to control a wide
range of motion applications. Nevertheless, higher integration and
more complex subsystems are some of the current trends in the
Industrial market and more and more Companies provide embedded
drive systems with different level of customization.
the market today, representing the best solution for such space-constrained mechanical environments.
Embedded drive systems integrate drives and electric motor to reduce
the space occupancy thanks to their compact and hermetical design.
Since they are dedicated to specific applications, design engineers
can optimize them to reduce the size and cost of the final product and
increase the reliability and performance.
Protection circuitry is tuned to match the power device’s capability
and factory-tested to improve the system’s reliability.
Discrete drives feature filters, connectors, and cables to be assembled and tested for UL and /or CE certifications. This increases the assembly time and the overall cost of the system. On the other hand, in
embedded drives the overall system’s size, cost, and time to market
can be slashed by considerably increasing the level of integration.
Mass-produced embedded drive systems benefit considerably from
economies of scale.
Vincotech’s product portfolio for embedded drives
Space is tight in embedded drive systems, and their compact, hermetical design makes it difficult to dissipate the heat generated by so
many electronic components.
The overall system’s size, cost, and time to market can be slashed by
integrating all of a motor drive’s functional blocks, apart from the input
filter, DC capacitor and microcontroller (see Fig. 1).
A lot of space is saved with highly integrated components and bare
power chips to achieve a much smaller footprint than that of discrete
Motor drive assembly is streamlined and simplified with fewer external
components and smart isolation techniques.
The ceramic sheet used in thick-film technology improves the module’s thermal performance by providing the best possible direct cooling for power components.
Vincotech’s power modules for embedded drives come into very
compact housings eventually equipped with Press-fit pins (see Fig.2).
Press-fit technology reduces PCB assembly time and effort considerably by eliminating the need for soldering. This cuts process time and
costs and boosts production output.
Vincotech’s power module portfolio for embedded drives features
600 V and 1200 V intelligent power modules (IPMs) as well as power
integrated modules integrating PFC circuit (PIM+PFC) and achieves
the highest level of integration of any power module available on
Figure 2: Press-fit pins for solder-less mounting
Figure 1: Cost analysis – flowIPM vs. discrete and competing IPM
Bodo´s Power Systems®
November 2016
The module’s creepage and clearance distances fulfill the applicable
industrial standards. There are no special requirements regarding the
shape of the heat sink.
The thermal interconnection between the power module and the heat
sink is vastly improved by pre-applying phase-change material (see
Fig. 3). Vincotech’s in-house screen-printing process deposits the
material with great precision, achieving the proper thickness. The
material can be optimized for maximum heat transfer capability.
Fig. 3: Phase-change-material
Vincotech’s product portfolio provides the functional integration
and power density that engineers need to design embedded drive
systems. The outstanding level of integration achieved by Vincotech’s
intelligent power modules enables system engineers to come up with
more compact designs and to take advantage of a proven combination of power components and gate drive circuits, which happen to
be the most critical elements in the inverter’s design. This mitigates
the risk associated with circuit design, speeds up development, and
dramatically reduces time to market.
Industry's Lowest On-Resistance
Ultra-Junction MOSFETs at 650V and 850V
Enabling Very High Power Density
Ultra low on-resistance RDS(on) and gate charge Qg
Fast body diode
Superior dv/dt ruggedness
Avalanche capability
Low package inductance
High-efficiency switched-mode and resonant-mode power supplies
Electric vehicle battery chargers
AC and DC motor drives
DC-DC converters
Robotics and servo control
Power Factor Correction (PFC) circuits
Renewable energy inverters
[email protected]
+49 (0) 6206-503-249
IXYS Power
[email protected]
+1 408-457-9042
November 2016
IXYS Taiwan/IXYS Korea
[email protected]
[email protected]
Bodo´s Power Systems®
Innovative Integrated Magnetics
for Hybrid and Electrical Vehicles
Onboard Battery Chargers
How to lower the volume of magnetics in ZVS or LLC converters by integrating the
insulating transformer and its resonant chokes together on the same core.
By Patrick Fouassier, PhD.-Eng., Inductive Components R&D Manager, PREMO Group
Nowadays, the market of hybrid and electrical vehicles is growing
quite fast. These are alternative solutions to common thermal engine
cars to reduce the global pollution, especially in terms of rejected CO²
or other NOx pollutants as well as toxic-for-health thin particles. Such
new models require more and more power electronics inside, not only
for the electrical motor supply with speed and torque control, but also
for high-voltage battery chargers and stable in-car continuous lowvoltage power supplies.
New SUV cars become
plug-in hybrids too
LLC topology
The first possible and most efficient used topology is the LLC
resonant bridge. It can be formed from 2 switching transistors (halfbridge structure, figure 1) to 4 transistors (full-bridge operation). The
transistors in series form legs connected to the DC-bus (input voltage
Uin). The transformer and its related resonant tank are connected
between these legs. At the secondary side we find a rectifier module
made of 2 (center-tap transformer) or 4 diodes (single secondary
winding transformer) and a filtering capacitor.
Trend in HEVs on 2013-2024
For 200-450V battery chargers, different power electronic topologies
like quasi- or resonant half- or full-bridge can be used. They enable
the development of high efficiency converters (> 90%) with a
switching frequency commonly in the 70-350kHz range. A power of
3.5kW enables a complete charge of the batteries during one night
(around 6-8 hours through a 10/16Arms domestic plug) whereas 7kW
reduces the duration to approx half-a-day charging (32Arms plug).
The power can be increased from 11-22kW (AC 3-phase network)
to 50kW (DC network) for ultra-fast charging in 30-60 minutes but it
requires dedicated charging stations to be installed through the cities.
For modularity concept, 3.5kW bricks can be connected together
with the selection of 1-phase 3.5 or 7kW or 3-phase 11kW charging
Thus 3-4kW HF transformers are required in the charging operation
both for voltage level adaptation from the rectified mains to the battery
voltage and for the by circuits physical separation between the AC
and DC. As the component belongs to a switch-mode power supply
Common battery charger SMPS schematics
it is often associated to inductors for benefit in soft-switching, better
control and EMI reduction.
Bodo´s Power Systems®
Figure 1: Half-bridge resonant LLC converter topology
The resonant tank associates to the transformer a parallel inductor
(Lp) and a serial additional inductor (Lr) as well as a serial capacitor
(Cr). These 3 passive components are usually discrete components
that require space in the application. They can show losses and
possible related heating at additional costs. The idea here is to
integrate the Lp and Lr components inside a single magnetic set
around the transformer. The set of values is defined to fix the
operating frequency range according to the Uout/Uin ratio to achieve.
The turn-ratio of the transformer is another parameter linked to the
resonant tank values. Example for a 3.5kW LLC full-bridge converter
from 220-420Vdc to 200-450Vdc in the 70-200kHz range : Ns/Np = 1,
Lp = 130µH, Lr = 22µH and Cr = 100nF (all at +/-5% tolerance).
This resonant topology is preferred in kW battery charging devices
since the control in power is only made by frequency adjustment.
According to the required output voltage-current needs (following
the batteries charging load-chart), the control-loop and associated
microprocessor calculate the switching frequency to apply to have
the corresponding Uout/Uin ratio. Even if the voltage that is applied
to the transformer is of a symmetrical bipolar wave-shape, the current
shows a quasi-sinusoidal waveform which is a great advantage for
higher efficiency by soft switching as well as for EMI reduction.
November 2016
Your SMPS design deserves a Plus
ST’s VIPerPlus offline high-voltage converters combine an 800 V
avalanche-rugged power section with a leading-edge PWM controller
• A plus in efficiency
High efficiency over a wide load range and best-in-class in standby with less than 4 mW at 230 Vac
• A plus in reliability
800 V avalanche-rugged power MOSFET, thermal shutdown, soft start, short-circuit protection and auto-restart
• A plus in integration
Direct feedback, jittered switching frequency for low EMC, embedded HV start-up, embedded SenseFET MOSFET, and
clamp-less design
• A plus in versatility
Power scalability from a few watts up to 15 W, with optional auxiliary winding, and available in through-hole or SMD
packages, supporting all SMPS topologies
Zero-power mode
VIPerPlus series 0P
Low voltage/minimal BoM
VIPerPlus series 1
Quasi resonant
VIPerPlus series 5
Minimal BoM
VIPerPlus series 6
VIPerPlus series 7
Peak power
VIPerPlus series 8
Fly-back converter
85-265 Vac
Buck converter
150 mA
12 W
15 W
200 mA
350 mA
For further information and full design support, visit us at
ZVS topology
The second possible topology is the ZVS phase-shift full-bridge
which is a quasi-resonant structure to also reduce the losses in the
semiconductors at switching instants (figure 2). In this, the control of
the transistor legs is shifted between both diagonal of MOSFET or
IGBT transistors and the applied duty-cycle reflects the calculated
compensation to get the required output voltage Vout at a given Vin
point. The ZVS inductor in series with the transformer is used to
create the zero-voltage-switching condition with the self capacitance
of each MOSFET transistor.
Thus the result is a kind of magnetic set (figure 4) where we find a
common transformer, showing or not an air-gap, and an additional
choke in series with the primary winding. The magnetic core of both
serial components is shared together at one side. The choke is the
serial inductor Lr or Lzvs depending on the topology (LLC or ZVS).
The transformer can include the Lp value if the center leg is cut to set
the magnetizing inductance value at Lp.
The magnetic set integrates both a transformer and a couple of resonant inductors
Figure 2: ZVS phase-shift full-bridge converter topology
Even if it is often requested to use the leakage inductance value of
the transformer, this practice is not without any effect in the self heating of the transformer. As a matter of fact, a value from 5 to 15µH is
normally necessary to create the ZVS condition with MOSFETs showing some 1-5nF of parasitic capacitance value. It cannot correspond
to the self leakage inductance of the transformer without a bad coupling (k < 0.995) which generally leads to additional high frequency
losses inside the windings which are crossed by the energy not stored
inside the magnetic core. That’s why a discrete additional inductor in
series with the primary of the transformer is normally found (figure 3).
Figure 3 : ZVS transformer
and its related serial resonant
In some cases it can be better to introduce also a thin gap in the
transformer core to make its behavior more linear in case of any possible dissymmetry that can appear secondary side due to the rectifier
bridge or any difference in the cabling length or contact resistance
between the two polarities of the secondary. Thus the magnetizing
inductance value of the transformer is of a slightly lower value but with
a more accurate tolerance on it. Example for a 3.5kW ZVS phaseshift full-bridge converter at 100kHz : Lmag = 800µH+/-20%, Lzvs =
The integrated magnetic set
The idea is to focus on a magnetic set that provides all these alternatives at once in a single component. The main advantages are the
size and weight reduction at a more competitive price. In terms of
cost, one has to consider not only the cost reduction because one or
two physical components less but also in terms of assembling process of the converter where less connections have to be made. Some
interconnections can be also directly replaced by special techniques
at the proposed magnetic set level.
Bodo´s Power Systems®
The use of low-losses high thermal stability ferrite cores as well
as thin stranded and twisted conductors in parallel for the winding
participate to the optimization of the losses versus temperature, input
voltage range and current consumption according to the charging
level of the battery.
The use of special isolated conductors is followed to fulfill the isolation and creepage distance as required by many electrical safety
standards (UL-2202, IEC-61558…). As a matter of fact, the isolation
requirements does not only include a high dielectric strength (typically
2.5-4kV) but also construction criteria like creepage distance, clearance and distance through the insulation to guarantee the reliability of
the insulating system provided.
The size reduction of magnetic components embedded in battery chargers for the increasing demand of HEVs goes through the
integration of the switching transformer and its associated resonant
chokes on the same core while providing a reliable insulation and an
efficient cooling capability. This 3-in-1 concept (1 transformer + up to
2 chokes) enables not only a volume and weight reduction but also a
big cost saving because less materials and also a reduced connecting
system (chokes and transformer partially linked internally) are used.
The work is not finished at that point and PREMO R&D has been
already developing an innovative 3DpowerTM concept which will soon
enable to design more compact solutions for the benefit of automotive
SMPS. However the new BCIM Series (Battery Chargers Integrated
Magnetics) is already presented as available standard products from
PREMO catalogue.
PREMO is a Spain-based company engaged in the development,
manufacture and sales of electronic components with a special focus
on the growing market of HEV, smart metering and market segments
including automotive, telecommunications and industrial electronics. Our product portfolio includes NFC and RFID antennas, power
transformers, inductors and chokes, current sensors, EMC filters,
PLC components and accessories. In addition to our broad range of
standard components, off-the-shelf products, PREMO designs custom
solutions to fit customer requirements.
November 2016
Drive Smart.
Generate Multi-Pulse Waveforms for
Driving Multiple Inkjet Nozzles using the
MP113 Digital Dual Channel Printer Driver
The Apex MP113 combines the simplicity of a digital interface with the high
voltage, high current and high speed of a dual channel power operational
amplifier. The onboard firmware supports user-defined, multi-pulse waveforms
required for industrial inkjet printing applications. The MP113 solves the
challenges of generating precisely timed, nozzle head data streams, and
high fidelity fire pulses when driving a changing number of print nozzles. This
comprehensive solution incorporates the digital control and system interface
with a dual power amplifier rated at >10A PEAK per channel, 180V supply
operation and provides output voltages up to 135V referenced to ground.
Use this module to upgrade an existing drive system to grayscale printing,
or as the integrated solution for an entirely new design while expanding
the capabilities to print higher quality images by changing droplet sizes for
greater color intensity.
Print Head
& Driver
© 2016 Apex Microtechnology, Inc. All rights reserved. Product information is subject to change without notice.
The Apex Microtechnology logo is a trademark of Apex Microtechnology, Inc. BPS0112016
Open Frame Product Technology
Footprint 110.5mm x 104.5mm
Power up at apexmic.ro/bodosMP113
The Flyback Dc-Dc Converter
The best bet for most SMPS”
The Flyback converter combines lowest cost with best performance and is the optimum
choice for offline SMPS up to more than 250 W. The simplicity of the circuit diagram is
deceptive, it is highly complex. This article is a practical guide to optimum designs and
lists the pitfalls.
By Dr.-Ing. Artur Seibt, Vienna
... Continue from October 2016 page 59:
5.5 RMS currents, winding losses.
The losses in the windings and all other resistances depend on the
rms values; the rms value of a trapezoid is between a sawtooth and a
square wave.
Irms = Iav x 2 x √τ x √(k2 + k + 1)/[3(1 + k)2 ]
= ip,max x 0.577 x √τ x √(k2 + k + 1)
k = 0 = sawtooth: 0.577 x ipeak x √τ k = 1 = square wave: ipeak x √τ
For k = 0.5 it follows: Irms, CM = 0.76 x ip,max x √τ = 0.51 x ip,DCM x √τ
Irms, DCM = 0.577 x ip,peak, DCM x √τ.
CM rms current is thus only lower by 0.51/0.577 = 0.87 or 13 %.
On one side it is true that all losses proportional to the rms current
squared are lower, but the number of turns and the winding resistance
are 73 % higher. Assumed the wire size and τ remain the same, and
the DCM values are considered for reference 1:
PCu, DCM = 1
PCu, CM = 0.757 x 1.73 = 1.31
Hence the losses are higher by 31 %. Whether the assumption of
equal τ is justified, would have to be checked in each case, because
τ is much more modulated in DCM. The current density in the wires
would be reduced only by 13 %, this is hardly a justification for selecting a thinner wire which would cause higher losses. Nor does this
suffice to compensate for the increased winding space.
Note that the rms currents in FB's with several output voltages are
interdependent and can have varying waveforms. Therefore it is mandatory to measure all rms currents in all relevant load combinations,
otherwise windings, diodes and electrolytics may be overstressed!
5.6 Voltage stresses on the semiconductors.
An increase of the primary number of turns by 1.4 to 1.73 means that,
the number of secondary turns unchanged, N will be increased and
the reflected voltages by the same factor. Remember that the numbers of turns have no bearing to the voltages. Example: Vout = 24 V,
N = 10, VB = 360 V. With DCM, the reflected voltage on the primary
would be 240 V, with CM it would increase to 336 resp. 415 V. While
with DCM the stress would be 600 V, it would rise to 700 resp. 775 V
with CM and rule a 650 V Coolmos out. There are 800 V Coolmos,
but they are more expensive and the Rdson is higher. Also emi would
become worse. The factor 3 for the Lp increase would not be realizable because 775 V is much to close to 800 V. N could be reduced
Bodo´s Power Systems®
by increasing the number of secondary turns, but this would require
still more winding space and cause higher losses.
5.7 Losses, efficiency.
Efficiency eventually has become the catchword, norms require > 87
% even at low loads, very low standby power etc. Although these figures do not yet apply to all offline SMPS, this will follow in due course.
Quite independent of these regulations, the very compact supplies do
not allow high losses. Comparisons of total losses will, in most cases,
be in favor of DCM. The smaller and less expensive transformer also
counts. In particular:
Winding losses:
These were treated above. In addition to the resistive losses of CM
the larger winding adds higher capacitances and dielectric losses.
Losses in the active components.
Even if Schottky diodes are used in the secondaries, the higher turnon switch losses remain. With standard Si ultrafast diodes there will
be higher losses in the diodes and the switch. With DCM, there are
only losses by the discharge of capacitances, but switching-off losses
may be higher.
Losses in the magnetic circuit.
DCM suffers from higher AC drive and the higher peak current. With
CM, the drive is lower, but on a high ip,min pedestal, i.e. in an area of
higher core losses. The higher amplitude DCM sawtooth will not necessarily cause higher core losses, this must be checked in each case.
Summary of the comparisons.
DCM advantages: 1. The transformer is smaller und less expensive.
2. Simple regulation. 3. Fastest converter. 4. Uncritical, less expensive secondary diodes. 5. The switching transistor turns on at zero
current. 6. Lower peak voltages on the semiconductors, lower emi.
Disadvantages: 1. Higher peak currents, because the inductances
are lower, the average currents are identical. 2. AC core drive level
higher, not necessarily higher core losses. 3. Higher ripple currents
in all active and passive components, larger capacitors. 4. Without
a regulation loop or a defective one it is a constant-power resp.
constant-current generator, without a sufficient load the output voltage
will rise to destructive levels.
Advantages CM: 1. Constant-voltage generator. 2. Lower peak currents. 3. Lower AC currents, lower ripple current loads.
November 2016
Disadvantages: 1. Bigger and more expensive transformer. 2. Higher
switch and diode losses. 3. Higher winding losses. 4. Regulation loop
design more complex. 5. Slower.
Whether DCM or CM is best has to be studied in each design because of the very many parameters and interdependencies
6. Transformer design.
FB transformers are by far the most difficult ones in SMPS design,
especially for offline SMPS, because, apart from technical viewpoints,
safety and emi norms have to be observed in addition; transformer
manufacturers have to use one of the acknowledged insulation systems. These items can not be treated within the scope of this article,
the reader is referred to the pertinent literature. Also, the reader is
kindly requested to refer to the article "The Focus is on Passive Components ..." in Bodo's Power Oct. and Nov. 2015 which treats winding
and insulation materials, the skin and proximity effects and their influences on transformer design. The knowledge of the basic equations
for magnetic circuits and the catalogs of major ferrite and transformer
materials (like coil formers etc.) manufacturers like Ferroxcube or
Epcos are indispensible.
With special regard to young engineers, the practical procedure is
explained step by step. This is not a chore for a young engineer. "...
even a novice can ...", no, he can not! Nor is it "child's play".
New designs require iterations because so many parameters which
are partly interdependent, partly contradicting, have to be reconciled.
There is no direct route to the optimum transformer, the allegations
of diverse software programs are arrogant and misleading. It was
explained that even grossly wrongly designed FB transformers often
function. The designers of such software claim that their program contained "the experience of decades". How come that they did not learn
in decades, e.g., that such important parameters like diode blocking
voltages exist and must be observed. During the design iterations the
designer has to take repeated decisions before continuing which no
software can take. The design task is expressly to create an optimum
and not just any transformer, not only for cost reasons, this chapter
will show that the transformer has the overwhelming influence on the
operating mode, the stresses on components, the efficiency, so only
an optimum design will ensure best performance and reliability. Today
optimum efficiency is asked for, so also dielectric losses have to be
considered. If several outputs are needed it depends entirely on the
transformer design how well those will be stabilized under variable
load conditions; a poorly designed transformer will require postregulators which add cost, need space and generate losses; a professionally designed one can hold output variations within a few percent.
Fixed-frequency, hard-switched operation with CMC is assumed
which is most appropriate, reliable and gets away with the smallest
transformer. As long as there is yet no transformer, windings and their
resistances are unknown, losses are disregarded, they are introduced
in the last step. At the present state-of-the-art of power switches,
Coolmos and especially cascodes, high performance ferrites and
winding materials, FB's own the territory to beyond 250 W. With SiC
cascodes higher output powers are possible. However, for higher
powers than 250 W it is advisable to partition the load among two or
more FB's driven by a multiphase clock.
6.1 1st step: Core selection.
The first step is the selection of the type and size of a magnetically
satisfactory core, it is essential to realize that this means just that, it
does not ensure that a transformer with that core will provide enough
winding space! This is a separate condition to be fulfilled; if it turns
out in the course of the design that a bigger size will be required,
all calculations of the magnetic circuit will have to be redone which
will again lead to reduced numbers of turns and less winding space
needed. Right here it becomes obvious that the design consists of a
series of iterations.
In Europe, the US and other countries there are standardized core
types and sizes, listed in the catalogs resp. on the home pages of the
ferrite manufacturers. The European types are based on the metric
system and should be preferred; in the last years many US types
based on the inch system were added which mostly differ little and
offer no advantages. The selection criteria are:
1. Dimensions. 4. Performance. 2. Cost. 3. Second sources.
5. Other criteria like stray fields.
Dimensions come first because the transformer is the biggest component in a power supply, the designer may, e.g., be forced to select
one of the special low-profile core types. Planar transformers are
the lowest-profile types, but their design should be left to specialized
firms. In higher power SMPS, the electrolytics may be the tallest;
unless absolutely unavoidable, low-profile electrolytics should be
November 2016
Bodo´s Power Systems®
shunned, because their life expectancy is lowest. After a first selection
of those core types which come into consideration, the next step is to
select the one best for the application. Single-source components are
an absolute no-no, they must be considered as non-existent. Some
manufacturers created non-standard types and their own ferrites
in order to lure customers into a single-source trap; even if such a
nonstandard type would fit, it must be rejected. If a designer chooses
a single-source component, his purchasing department is not only
forced to nod to any price the supplier will ask for, but if he can not
deliver for some reason or other, production will stop. Most control ic's
these days are single-source, requiring a redesign or a new design
of the SMPS if not available. This is dangerous, because in the past
there were fairly few control ic's which were standards and multiplesourced; today, a single manufacturer offers dozens and hundreds of
such ic's, products will be cancelled very fast if the expected volume
or profit was not realized. As the transformer is always the most expensive component in a SMPS, it is wise to share the final decision of
core type and manufacturer(s) with the purchasing department which
has its own viewpoints on suppliers. The purpose is to create a winding prescription for the purchasing department which it can distribute
to several transformer manufacturers. If transformer design is farmed
out to a manufacturer, the design would be his, so competition would
be excluded.
There are magnetically superior types like the pot cores (RM, PM etc.)
which also do not radiate, there are types like the ETD cores which
are designed especially for low leakage. Traditionally the E- resp. EE
cores are the lowest cost types which were derived from 50 Hz iron
core transformers, but differ somewhat for ferrite cores! Cores and
coil formers are simple and available from a host of suppliers. However, their strong stray fields increase the leakage and can couple into
the circuit and cause severe malfunctions and false measurements.
Before triple-insulated wires were available, many types like pot cores
were not eligible for offline SMPS because the 8 mm clearance was
impossible. E cores and their derivatives were preferred, on both
sides of the layers 4 mm were left free; a poor solution, especially for
small transformers, because this not only costs winding space, but
increases leakage. Triple-insulated wires allowed down-sized and
magnetically improved transformers.
The decision to take: if the dimensions of EE cores are acceptable,
will they fulfill the other requirements or is a better type required. The
problem is that any influence of the stray fields will show up very late
in the SMPS design phase. By proper orientation and clever component placement as well as correct routing of critical conductors it can
be handled.
Starting point: Energy content.
The core type settled, which size is the next question; in the following
it is assumed that E cores are selected, they range from tiny to sizes
good enough for several hundred watts. Where to start? In general,
nearly quadratic core types are preferable, this is the reason for the
versions of one type with different core thicknesses. Hence, before
going up in size it is better to choose the version with the thicker core.
The solution is based on the fact that the transformer is a storage
choke, capable of storing the amount of charge necessary to supply
the load and all losses at the operating frequency. This amount of
charge will be transferred - NOT transformed - f times per second to
the output. Starting from the basic specifications Pmax, VB,min and
assuming a partition of 50 % - 50 % for charge and discharge as a
reasonable starting point, the following equations from chapters 3 to 5
apply in this case:
Bodo´s Power Systems®
Pout = 0.5 x Lp x ip, peak 2 x f
Lp,max <= VB,min2/8Pf
Ls,max =<= Vs2 /8Pf
They are strictly only valid for fixed frequency DCM, but yield satisfactory first approximations. The frequency is assumed 100 ... 140 KHz.
Below 100 KHz has been the state-of-the-art 30 years ago; 150 KHz
is a practical limit, because EMI norms begin to become strict, they
are strictest from 0.5 to 5 MHz, so the 3rd harmonic of the clock frequency should also stay < 500 KHz. If the increased expense for emi
suppression and the higher losses are accepted, up to 250 KHz are
possible; this depends mainly on the ferrite, see below. The control
ic must already be selected and its built-in duty cycle limit, e.g. 50 %
known, because the transformer must guarantee that under worst
circumstances the duty cycle stays below. Design for 40/60 % will be
a reasonable choice. The necessary storage capacity is given by
Energy content E = I2 L = 2P/f (Joule) by rearranging the basic equation.
The two factors can in principle be interchanged. The energy is stored
in the air gap, which reduces L which allows the current to reach
higher levels in a shorter time, but first decreases P. The equations
show there are limits to Lp and Ls if DCM shall be ensured. For the
determination of an approximately satisfactory core size, first only Lp
needs to be considered. Lp,max can be calculated with the formula
above. ip,peak = √ip,peak 2 . The current rises with a speed given by
di/dt = V/L; if it can not attain the necessary ip,peak within the 40
% assumed, the desired P can not be achieved. So L must then be
reduced and ip,peak2 raised until the product E can be realized If, e.g.,
L is reduced to ¼, the number of turns n to 1/2 , ip,peak has to rise by
a factor of 2. The Θ = I x n remains constant.
Magnetic relationships..
From Θ follows the field strength H = I x n/lcore . The crucial parameter
is however the flux density B = μ0 μrel x H = Φ/Acore , where Φ is the
magnetic flux. For the selection of a core and final checks these forms
of the equation are most handy:
B = 1.26 Gcm/A x μeff x I x n/lcore
[I in A, lcore in cm. Result in G. 1 G = 10-4 T = 10-8 Vs/cm2 ]
B = AL x n x ipeak /Acore
AL is the inductance factor of a coil or winding,
L = AL x n2 resp. n = √L/AL
μeff is a dimensionless factor for a specific core with a specific air gap
for a given ferrite, to be taken from the catalog data as well as AL .
lgap = δ is not needed for any calculation because it is implied in AL
resp. μeff . Some ferrite manufacturers specify their core series by AL,
some by δ. This is why the first equation above is most appropriate.
► Only cores those cores are adequate which allow an operating
Bpeak,100 at 100 C.
Ferrite considerations
The higher the operating Bpeak,100 , the more power can be moved
through a transformer; in a given winding, Bpeak is proportional to
ipeak , P ~ ipeak2 ~ Bpeak2 , a better ferrite allows a smaller transformer and lower losses, but it is more expensive; on the other hand
the smaller core and the smaller coil former cost less. A smaller
transformer offers more benefits: 1. The smaller the transformer, the
November 2016
Can you engineer Power Electronic
solutions of tomorrow?
Danfoss Silicon Power is growing!
We are looking for bright individuals to join us on our exciting journey.
Are you our new colleague?
Are you ready to engineer tomorrow´s power electronics for automotive,
industrial and renewables applications?
Interested? Discover more at www.danfoss.com/career
larger the specific surface for heat radiation. 2. Winding resistances
and losses are lower. 3. Also the ferrite losses depend on the volume.
4. Capacitances will mostly be lower. As the only "penalty" is the
price of the ferrite, it will turn out eventually that the higher price for
a premium ferrite pays off; it is justified because extremely pure iron
oxides have to be used, and the final heat treatment which belongs
to the trade secrets, is more complicated. There is much black art in
ferrite production, so there are only a few manufacturers in the world
which can deliver these qualities. Ferrites for SMPS use are made
of MnZn, useful to about 2 MHz. But high B's can only be used up to
about 250 KHz, above the B's have to be drastically reduced as the
losses rise steeply.
permissible, this value is not always given in the data; in these cases,
it is safe to derate the material's spec by 20 to 25 %.
f x Bmax is called the Performance Factor. Fig. 6 depicts f x Bmax of
some Ferroxcube ferrites ; it shows how much the modern ferrites like
3C96 surpass the older materials like 3C90 or 94. It also proves that
the optimum frequency extends to about 250 KHz. This picture does
not yet include the flat-temperature materials 3C95 nor the extended
saturation materials like 3C92 or the materials for higher operating
If there is already some experience, this is the fast route. Example:
the core estimated to be satisfactory is an E 42/15; the data sheet
delivers the following table:
Size selection method 1: With some experience.
The route to core size selection is now clear: after picking a size
which might suffice one picks a gap length which might be right,
reads the 2 parameters from the data sheet and, inserts them in first
equation for B above; if the B calculated is below the Bpeak, 100 of
the selected core, heureka. If not, a larger gap is tried, if that does not
lead to a solution, the next bigger core is picked and so forth. There
are several routes to core selection, manufacturers have their preferences and differ in the data they offer for that purpose.
AL μeff air gap/mm
The AL values are standardized.
Fig. 6. The performance factor f x Bmax of some Ferroxcube ferrites.
All power ferrites except these new materials show loss curves vs.
temperature which fall towards a minimum around 100 C; the "95"
type materials have a flat curve. There are special types with the
loss minimum at 60 C and such for up to 140 C. Beyond the respective minimum, the curves rise steeply towards the Curie temperature
around 220 C.
The other major manufacturers have equivalent materials so second
sourcing is assured, but the designations all differ. The topic "ferrites"
far exceeds the scope of this article and will be deferred to a future
one. Important in this context are these items:
1. All 25 C specs are for the birds and have no practical meaning
whatsoever! No power supply remains at 25 C in operation. The internal air temperature is rather between 60 to 85 C, and the transformer
must heat up more in order to get rid of its heat. And the parts of the
core which are inside the coil heat up most. Therefore only the 100 C
values are of practical use.
2. Do not mix up the B values in the socalled "materials specifications" with practically usable ones, they are always higher and illusory.
Many core types suffer from deficiencies like unequal areas and
flux concentrations, consequently only a lower value of Bmax,100 is
Bodo´s Power Systems®
Air gaps of 2 mm are rather uncommon with such a small core, because δ/lcore must remain within reasonable limits, from experience
it is known that AL 's mostly range between 250 to 400. So 1.5 mm is
first chosen. The number of turns for the Lp,max calculated in the beginning follows from n = √Lp,max /AL . Now n, lcore , ipeak from above
and μeff for 1.5 mm are entered into the formula for B and checked
whether this value is below a reasonable operating Bmax,100 for the
ferrite selected. 3C96 assumed, in the material's spec, the saturation
flux density at 100 C is 0.44 T. From the hysteresis curve it follows
that one should stay clear of 0.35 T. The diagram for the losses as a
function of B with the frequency as the parameter and the spec of 300
KW/m3 at 0.2 T indicate that one should rather stop at 0.3 T. This is
still based on the materials spec. In the core's specs, there is none
for a Bpeak,100, only a spec for the losses at 100 KHz, 0.2 T for the
similar 3C92 as a guidance. Bpeak,100 should hence be limited to
0.25T. The waveform has a strong influence, all the catalog data are
taken with sine waves while the sawtooth or trapezoids cause different losses. It is therefore possible to let the peaks of the waveform go
up to 0.28 .. 0.3 T. There is no other way as to pick such a reasonabe
value and to proceed with transformer design. In any case, it will be
necessary to measure the inner core temperature in the actual worst
case operation, it must not extend into the steep part of the losses'
curve above 100 C.
If this first assumption resulted in too high a B, AL = 160 can be tried.
If in the first test B was much below 0.23 T, AL = 400 can be tried.
Nonstandard air gaps are possible if ungapped cores are bought,
there are companies specializing in core gapping. The formulas follow.
Bear in mind that this calculation only proved that this core size is
magnetically ok, whether the windings will fit is indeterminate at this
point. If it should turn out later that they will not fit, a bigger size has
to be selected, the determination of the optimum air gap would then
have to be repeated and yield a higher AL and thus a lower number
of turns.
November 2016
The special high Bmax grades like 3C92 deserve attention; they
are destined for all applications where the core is driven far towards
saturation. This applies in the first place to all such chokes where
a fairly low AC ripple is ontop a high DC bias. It is the material of
prime choice for PFC chokes. In transformer applications, losses
are somewhat higher than with standard premium power ferrites like
3C96, Experience shows that one should always perform the calculations also for 3C92 and build test samples; in many cases 3C92 will
turn out to be superior, because the higher Bmax allows higher power
resp. a smaller and less expensive transformer!
Fig. 7. Inductance per turns squared of a material. These curves look
alike for all materials. The designation "DC bias" on the abscissa
is misleading, because only the peak Θ = ipeak x n counts. If it is
known, the necessary air gap can be read from the graph. However,
if the temperature is unknown for which the curve is valid, it can give
only a rough indication.
Size selection method 2: via the curves AL vs. ipeak with the air
gap δ as parameter.
Some manufacturers provide such curves like Fig. 7.
Size selection method 3: Hanna curves.
The Hanna curves were published 1927 and show the energy content E = L x ipeak2 per volume in cm3 of a given core vs. the field
strength H in Oe with the relative air gap δ/lcore as the parameter.
They yield the air gap and the number of turns for the desired L.
Fig. 8. An example of Hanna curves.
4:1 DC-DC Converters – Standard DIP-24 package
The PXC-M Series are packaged in the industry
standard DIP-24 footprint. With 5,000Vac input
to output isolation, these efficient converters
are suitable for IGBT gate drivers, medical
applications and battery powered equipment.
• Suitable for IGBT gate drivers
5kVAC Isolation input to output
• 3, 6 & 10W Output
• 4:1 Input: 24V, 48V
• 3.3, 5, 12, 15, 24, ±5, ±12 and ±15V Outputs
Visit us
08. – 11.11.2016 ∙ Munich
Hall A2 ∙ Booth 205
November 2016
Bodo´s Power Systems®
TDK-Lambda Germany GmbH · www.de.tdk-lambda.com · +49 (0) 7841 666-0
There is a curve for each core. Their use requires some caution;
lower points may already be close to saturation! Procedure:
1. First calculate Ldesired x ipeak 2 / V.
2. Now the curve is entered, preferably in the center region, also
there, some points may already be close to saturation, and one
does not know whether the saturation level on which the curve is
based is the 25 C or 100 C. Ferroxcube specifies for 25 C. Hence
Hanna curves should be used with care and the values obtained
checked by calculations. But they give at least an approximation.
3. Determination of the air gap: from the value calculated above on
the vertical scale move to the right and read δ/lcore (in the picture
called "Δd/de "): δ = δ/l x l.
4. Determination of the number of turns: Read the corresponding
value of H (Oe) on the abscissa. 1 Oe = 0.8 A/cm. H = 0.4 x n x
ipeak /lcore [Oe]. n = H (Oe) x lcore /0.4 ipeak .
5. Final check of B whether it is below the permissible Bpeak,100 .
Size selection method 4: I2 L curves vs. δ
Most manufacturers offer curves which only require to calculate I2
L, the volume is already taken into account by showing all sizes of
a core type as the parameter. This is the most practical and popular
method. Fig. 10 shows these curves for E cores.
To the right of the I2 L value on
the ordinate all core sizes and the
appropriate air gaps are shown
which would fit. As these curves are
provided for all core types, it is easy
to see which core sizes would fit in
the various types. Ferroxcube states
that all specs pertain to 25 C. It is
therefore wise to directly select one
size bigger than indicated and remain rather in the middle, i.e. small
air gaps and very wide ones should
be disregarded.
There is a more practical version of the Hanna curves (Philips) : P x f/
fref on the abscissa, AL resp. √AL on the Y axis and a set of straight
lines from top left to right below for the various standard sizes of a
core type, completed by a set of straight lines from left below to top
right with the ampere x turns. This diagram is based on the formulas:
0.5 LI2 = k1 x n x I + k2
P = 0.5 LI2 f
AL = L/n2
The first formula should provide an approximate linear relationship
for the non-saturating portions of the Hanna curves. The procedure
for using this diagram is this: determine P x f/fref , draw a vertical
line from this value to the abscissa, this will intersect the curves for
the core sizes at those values of AL which correspond to saturation
(probably 25 C). As AL is only available in those standard values, one
has to choose whether to pick the value at the point or a higher value
of a bigger core.
Size selection method 4: via curves AL vs. δ with the core sizes
as the parameter.
Some manufacturers provide curves like Fig. 9 which shows an
example for pot cores. Here, AL is shown vs. δ with the sizes as the
parameter. The curves show that often several sizes are eligible; it is
preferable to select the bigger one and directly check Bmax with the
above formula.
Fig. 9. Core selection by AL vs. δ
Bodo´s Power Systems®
► Remember that, irrespective of
the method used, Bmax,100 eventually has to be checked, using the
data sheet values and the formula
given above.
Fig. 10. I2 L vs. δ curves for E
cores (Ferroxcube).
This completes this step, the result
is a core size only commensurate
with the requirements from the
viewpoint of the magnetics. Whether
the windings will fit, will be a different story.
6.2: 2nd step: Primary winding and transformer size determination.
Simple case: choke.
The procedure for determining the primary winding is in principle identical to that for designing storage chokes, e.g. PFC chokes. In such
cases there is a fairly small AC ripple ontop a high DC bias (in case of
PFC chokes those are the peaks of the 100 Hz half-sine). Therefore
it is admissible to use a higher Bpeak,100 ; the true rms value causes
winding losses, the AC part core losses. Note that the current consists
of the 100 Hz half-sine and the high frequency sawtooth ripple which
is typically 20 %; this applies to the high-quality PFC, the inferior PFC
types have a high frequency ripple of twice the average line current,
so their chokes are far bigger, lossier and more expensive. The popularity of the inferior PFCs is due to the misconception that a cheap
control ic also means a cheap PFC. Comparing complete designs
resp. BOM's counts.
Transformer size selection.
In transformer winding design only the total height counts, calculations based on the winding volume are too inaccurate; this is especially true in case of offline transformers which require a lot of insulation.
In most cases it is standard to split the primary in two halves, one at
the bottom, one on top, so there are two insulation barriers. Also, as
will be seen in the following chapters, the secondary winding construction can be very complicated.
At this point it is important to note that the designer may well have
decided to operate in CM and with a 20/80% partition, but no final calculations can be performed because there is still no transformer, only
a core! The assumptions taken in step 1, i.e. a 50/50 % partitioning
and DCM are reasonable starting values; if eventually CM is desired,
it is clear from the outset that the transformer will become bigger.
November 2016
With the formulas from step 1 resp. chapters 2 and 3 Lp can be
Pout = 0.5 x Lp x ip, peak 2 x fLp,max <= VB,min2/8Pf
Ls,max =<= Vs2 /8Pf
n = √L/AL yields the number of turns; the rms current follows from
the general equation:
Irms, DCM = 0.577 x ip,peak, DCM x √τ. for 50/50 %: = 0.4 x ipeak
For each transformer type and size there exists a maximum current density which depends on the size (smaller sizes have a better
surface-to-volume ratio and allow more). Please refer to the article
"The Focus is on Passive Components" for an explanation of wires
and insulation materials. PFC chokes as well as primary windings use
hf litz. For a start, one can take current density values from the same
size 50 Hz transformers' norms' data sheets for magnet wires. The
copper area follows from: ACu = Irms /S (current density), the outer
diameter is taken from a table of hf litz. There are insulation materials
like Teflon which allow much higher temperatures than the enamel
on copper wires. But the current density figures given for the specific
transformer sizes are based on definite copper losses which heat the
transformer to a definite temperature. Even if high temperature materials are used, one can not allow the winding to get too hot, because
the standard power ferrites show sharply increasing losses above 100
C and would run away! In order for the ferrite to get rid of its heat and
stay below 100 C, a temperature difference to its surroundings must
be maintained. A thermal run-away towards the Curie point would
result in loss of permeability and destruction of the switching transistor with ensuing burning of the transformer. If this should happen, for
verification clip a DC/AC current probe around a primary conductor
of a new sample with temperature sensors within the winding close
to the core and on the core inside the coil former and watch. The
onset of thermal runaway resp. saturation is very fast. This is why the
finished transformer must be thoroughly tested, for temperatures and
The hf litz type which provides the necessary copper area selected,
the winding width is divided by the litz' maximum outer diameter which
yields the number of turns per layer. Dividing the total number of turns
by this number gives the number of layers. This figure multiplied by
the litz' maximum outer diameter gives the height of the winding. In
case of the primary the 4 KVrms insulation must be added for each
half. A 10 % reserve should be provided.
This check of the winding height for the primary will show how much
is left for the secondary, if less than half is left, choose the next bigger size core und iterate, until half the height is available. Each time
a bigger size is selected, AL will be increased, np decreased, due
to greater winding width the number of layers and the height will be
reduced further; but the wire size may have to be increased, because
the current density permissible will be lower. The interdepence of
all parameters is obvious. The necessary barrier for 4 KVrms test
voltage is achieved, if either P or S are wound with triple-insulated
approved litz. However, Furukawa litz is only available with 7 strands,
so, for higher power transformers, several wires have to be wound in
parallel which wastes winding space. Rubadue can provide hf litz with
hundreds of strands, also Teflon-insulated for lowest dielectric losses.
In most cases it is the primary which uses the triple-insulated litz,
especially if there are several secondaries.
The transformer size thus defined, with a half empty coil former, will
be close to the final one if DCM or "mild" CM operation is used. For
"deeper" CM operation the primary winding should already here be
Bodo´s Power Systems®
increased; repeating from the foregoing chapters it is not advisable to
exceed Lp,CM = 2 x Lp, DCM which requires an increase of np by
a factor of 1.4. Remember that CM is not only defined by Lp but also
by Ls which is still unknown. This will in most cases require a bigger
Although the primary of an offline transformer will practically always
have a much higher number of turns than the secondary, also here
rules the strict requirement that only full layers are allowed! So one
must play with n and the wire size to realize full layers. It this should
not be possible, the last layer must be evenly distributed over the
whole winding width. In order to minimize the voltage between layers,
Z winding should be preferred which also helps to alleviate the dielectric stress. The beginning of this half, on the bottom of the coil former,
is the hot end and is connected to the drain of the switching mosfet
which has to be placed so that this connection is as short as possible.
The end of the winding is thus almost cold, and here the low voltage
auxiliary winding for feeding the primary side control circuitry is placed
and distributed
Another problem arises if the core has an overly large air gap: the
stray field will be strong and induce additional losses in the windings
directly above; in such cases there should be no turns above the air
gap at least in the two bottom layers.
Last, by far not least: the insulation barriers between the primaries
and the secondaries must conform to the safety norms. However, in
the author's opinion, the usual materials are not good enough. Most
triple-insulated wires and litz use low-cost thermoplastic materials
like polyester, nylon etc. which melt beyond their maximum operating
temperatures, so does polyester as a barrier. The insulation properties of these thermoplastic materials deteriorrate drastically with rising
temperature! A burning winding can reach much higher temperatures.
What about the primary fuse? The typical failure is a destroyed
switching transistor, this places the primary directly across the line,
i.e. up to 254 Vrms ,maybe only 90 V. The fuse must be rated for the
highest current at low line, i.e. 90 V, and it must withstand the current
surge at turn-on at 360 Vpeak without fatigue. Fuses are frequently
rated for a much higher current than in normal operation. The author
has seen 4 A fuses in a 75 W SMPS made by a big concern. A fuse
must withstand 1.5 rated current for at least one hour, in that case
254 V x 6 A = 1.5 KW can flow into this SMPS for one hour. To make
a long story short: it is highly recommended to use barriers of twice
Kapton; this material takes 400 C and disintegrates at 800 C without
obnoxious gases. Kapton is expensive, but in those small transformers very little is needed.
6.3. 3rd step: Turns ratio(s)
Lp and the number of turns for the core selected sofar are known, all
on the assumption of a 50/50 % or 40/60 % partition of a period. The
paragraph above described how to arrive at the winding height of the
primary. Keeping to the step-by-step procedure, the turns ratio has to
► Remember that the numbers of turns and the turns ratios of a FB
transformer have nothing to do with the primary and secondary voltages.
► Contrary to most texts there is not one turns ratio, but a range.
The turns ratios N are determined first of all by the permissible peak
voltages on the switching mosfet and the secondary diode! Regarding
the reflected voltages, the transformer indeed acts as one. Whether
November 2016
a PFC precedes the FB converter or not: the maximum input voltage
VB,max = 360 V. This pertains to a well designed PFC, others need
output voltages up to 420 V. The switch will be a Coolmos or a cascode
with an upper mosfet. Please refer to the former articles in Bodo's
Power for details. The standard Coolmos takes 650 V, there are 800 V
types, but they are unnecessary for offline SMPS. Beyond this voltage
the part will enter a nondestructive avalanche breakdown; the energy
in this mode will be dissipated, the part will survive as long as it is not
overheated. Principally, the part could break down in each period,
but this is not advisable, because avalanching is a stochastic process which generates output ripple and emi. Note that only individual
mosfet's have a rating, the high voltage transistors on one-chip ic's will
be destroyed by any overvoltage! Samsung invented the two-chip components which comprise a standard mosfet chip and a control ic on the
same leadframe, these are available from several manufacturers, the
best combine a Coolmos and a CMC control chip, e.g. from Infineon.
Schottkies can not absorb overvoltages resp. very little. Spikes have
to be damped by RC's.
In order to stay below the maximum voltage across the mosfet
Vmax <= VB,max + Vrefl,S-P . or
Vrefl.S-P = Vmax - VB,max = Vout x N.
It follows that
Nmax < (Vmosfet,max - VB,max )/ Vout = 250/5.6 = 45. Note the 5.6 V.
A high N is good for the diode, but if it is too small, the diode will be
endangered. VD,max = Vout + Vrefl,P-S ; Vrefl,P-S = VB,max /N.
A lower limit for N follows:
Nmin > VB,max /(VD,max - Vout ) = 360/27 = 13
A low N is good for the mosfet.
Behind the reflected voltages there are "hard", i.e. low impedance
voltage sources: the primary and secondary capacitors, the transformer is low impedance and low leakage; it is therefore of vital
importance to ensure that the sums of the reflected and the primary
resp. secondary voltages do not exceed the blocking voltages of the
switch resp. the diode(s), because this would destroy them. With a
650V Coolmos and a VB = 360 V, the maximum permissible reflected
voltage would be < 290 V, so < 250 V would establish a reasonable
design margin. Spikes of some ten ns can be allowed.
In the following example it is assumed that there is one output voltage
Vout = 5 V and that the blocking voltage of the diode is 45 V, typical of
a low voltage Si Schottky. A 30 % margin would allow 32 V. Note that
► Permissible N range: Nmax > N > Nmin
In the course of all iterations N must be kept within this range. For a
start, it is reasonable to choose the average value
Nav = (Nmax + Nmin )/2 = 58/2 = 29
Of course, the final value may differ, but it is mandatory to stay within
this range which incorporates safety margins. Note that these days
there are no reserves in the specs, i.e. if a 100 V mosfet measures
101 V in final test, it will be delivered.
Magna-Power’s high frequency IGBT-based programmable DC power supply line spans 1.5 kW to 2000 kW+
with hundreds of models to target a variety of different applications.
Using a Magna-Power supply is as simple or sophisticated as the application demands with front panel control, 37-pin
isolated analog-digital I/O and a computer interface. Remote programming is supported through a variety of development
environments by a provided National Instruments LabVIEW™ driver, IVI driver and SCPI command set over RS232,
TCP/IP Ethernet, IEEE-488 GPIB and USB.
Designed and manufactured in the USA. Available worldwide.
XR Series
TS Series
MS Series
MT Series
2 kW, 4 kW, 6 kW, 8 kW, 10 kW
5 kW to 45 kW
30 kW, 45 kW, 60 kW, 75 kW
100 kW to 2,000 kW+
1U Rack-mount
2U Rack-mount
3U to 9U Rack-mount
Floor Standing
Floor Standing
No. of Models
Voltage Range
0-5 Vdc to 0-1,000 Vdc
0-5 Vdc to 0-10,000 Vdc
0-5 Vdc to 0-4,000 Vdc
0-5 Vdc to 0-4,000 Vdc
0-16 Vdc to 0-4,000 Vdc
SL Series
Power Levels 1.5 kW, 2.6 kW, 4 kW, 6 kW
Range 0-1.5 Adc to 0-250 Adc
0-0.2 Adc to 0-600 Adc
0-1.2 Adc
to 0-2,700 Adc
Adc to 0-4,500Power
Adc to 0-24,000®
6.4. 4th step: One secondary winding resp. output.
The design of the secondary winding(s) is the most difficult task in
FB transformer design. Note in the first place, that all secondaries
must have the opposite polarity of the primary, if those are connected wrongly, the SMPS will "kind of" function, namely as a forward
converter which may or may not be noticed, mostly high losses and
overheating will point to this. The regulation loop always tries to bring
the output up to the programmed voltage.
In chapter 6.2 the - still preliminary - size of the transformer and its
AL were defined. Because N will have to be limited to the range
Nmax to Nmin it is now necessary to know np which follows from np
= √Lp,max /AL . In order to set up an example, it is assumed AL = 250,
P = 100 W, it is further assumed that this FB is fed by a preceding
PFC with 360 V, so this is VB,min = 360 V.
Lp,max = VB,min2 /8Pf = 3602 V2 /(8 x 100 W x 100 KHz) =
1,6 x 10 -3 V2 /VA x s-1 = 1.6 x 10-3 Vs/A = 1.6 mH.
np = √L/AL = √1.6 mH/250 nH = 125
ns may thus only vary between 125/45 = 3 and 125/13 = 10. Remember that these N limits stem from the specifications of the semiconductors and are rock-solid resp. any change would require other
semiconductors. If, e.g., the 45 V of the Schottky diode is considered
too low, an ultrafast Si diode would offer 200 V before becoming too
slow, but would have somewhat higher losses. As mentioned 600 V
SiC diodes outperform Si ultrafast down to appr. 24 V output, so this
would offer considerable leeway.
With Nav = 29 ns = 125/29 = 4
Disregarding the other secondary parameters for a while, but remembering that low leakage = tight coupling is absolutely essential for a
FB transformer, it is evident that 4 secondary turns spell miserable
coupling, this is totally unacceptable! Nor is 10 turns any better. The
transformer E 42/15 which can easily handle 150 W, the example in a
preceding chapter, e.g., has a winding width of 26 mm, spreading the
4 turns over the 26 mm would improve the coupling but waste winding
space. Also, the 20 Aav would cause an Irms = 1.64 x Iav = 33 A. A
single litz good for carrying 33 A would be quite thick which would hurt
the leakage again and waste still more space. This is no way to go.
A copper foil must not be thicker than twice the depth of penetration
at 100 KHz, i.e. 0.25 mm, not good enough for 33 A. Using two such
foils in parallel would be disastrous, why is explained in the article
"The Focus is ...". The current would only flow in the lower one. No
way either.
The solution is multifilar winding: The copper area necessary for 33
A at a current density acceptable for an E 42/15 has to be calculated,
this area distributed to as many thin litz as are necessary to fill one or
more layers fully, with the stress on "fully". ns may vary from 3 to 10,
this is helpful in fulfilling the condition. Ten parallel wires resp, litz are
not uncommon.
But: it is by far not that simple: each time ns is changed, so does Ls
which influences t2 resp. the t1/t2 %. E.g., if a higher ns is used to
fulfill the foregoing condition, the increased Ls may shift the converter
operation from DCM into CM. Consequently, each time any of these
parameters is changed, the consequences must be calculated! If CM
operation is desired, Ls resp. ns may have to be increased, so ns may
run against the stop given by Nmax or Nmin !
Bodo´s Power Systems®
6.5. 5th step: Several secondary windings resp. outputs. Cross
Matters become considerably more complicated as soon as several
outputs are required. The outputs are coupled by true transformer
action, that means that their relations are equal to their respective n's,
including the diode forward voltages, this imposes additional restrictions. Example:
The Vout's desired may be: 3.3 - 5 - 12 - 24 V. It will be quickly evident
that higher numbers of turns are necessary in order to realize the
relations: 3.9 : 5.6 : 12.6 : 24.6 V. As the currents remain unchanged,
higher numbers of turns may require a bigger transformer, apart from
higher losses and the possible undesired transition from DCM into
CM or from an intended mild into a deep CM. Example: Assumed 1
turn for 3.8 V would mean 3.8 V/turn. 5.5 V would require 1.45 turns,
impossible. 2 turns for 3.8 V equals 1.9 V/turn, now 2.9 turns would
be required for 5.5 V. The error if 3 turns are used would be 3.6 %;
this might be tolerable, also because the diode forward voltages are
temperature-dependent. For 12 V it should be 6.3 turns, the error
is already 11 %. For 24 V it should be 12.6 turns, the error is 3.2 %.
These errors can only be reduced by again increasing all windings
with the consequences.
The main output, as the rule that with the highest power, is the one
with the pick-off of the regulation loop, its output voltage (at the pickoff point, the first capacitor, downstream the resistances of filter coils
will cause voltage drops) is typically extremely stable and recovers
from large load steps within a few ten microseconds. However, the
main output may be 3.3 V, although FB's are a poor choice for low
voltages at high currents. In such a case one would take another output as the reference and stick a linear FET postregulator to the 3.3 V
output which requires only a few ten millivolts for proper operation. As
mentioned earlier, the best solution for low voltages at high currents
with a FB is to design it for 24 V and add a simple buck.
The temperature drift in the other outputs can, at least partly, be
compensated if the same type of diode is used in all. The drift in the
diode of the reference output wil, however, be multiplied by the ratios
of the voltages.
Cross regulation.
The term Cross Regulation denotes the interdependence of several
outputs under varying load conditions. This depends extremely on
the winding structure. Winding the secondaries as was used with 50
Hz transformers, i.e. one after the other, would result in pretty poor
performance. This is the way to go: all secondaries are combined
in one multifilar structure, that means that first the lengths of litz for
each secondary will be cut to size, then the beginnings of all will be
soldered to the respective pins of the coil former. Then the bundle of
wires will be wound by carefully keeping them side by side, i.e. they
should not overlap. The wire belonging to the regulated output in the
middle. Still better performance can be realized if the winding of the
regulated output is split into several wires such that the wires as they
lie side by side are ordered like this: unreg. output 1 - reg. output unreg. output 2 - reg. output and so forth. This must be done very
orderly. Of course it holds what was said about one output: only full
layers are acceptable. Therefore one must play with number of turns,
wire sizes, numbers of parallel wires until this is fulfilled. A professional design like this will allow the unreg. outputs to stay within a
few percent. Cross regulation must be tested by changing the load
on one in turn and looking at the others. The most critical situations
are: maximum load on the regulated output, no loads on the others
and minimum load on the regulated output and maximum loads on all
others. Note: a minimum load must be assured on all unreg. outputs,
November 2016
because the diodes will rectify spikes, so that unloaded outputs will
rise in voltage which will endanger the electrolytics and also the load.
If that was not enough complexity: the secondary rms currents, necessary for the determination of wire sizes, diodes and capacitors, are
interdependent, i.e. they change their waveforms any time a load on
one or more outputs changes and so also their rms values. Therefore
they must be determined by accurate measurements.
6.6 Final adjustments
After all windings are fixed so that the desired operating mode is
ensured, now also the winding resistances are known, so their losses
can be calculated with the respective rms currents. The core losses
can be taken from the respective curves in the ferrite manufacturers'
data sheets, bearing in mind, that the curves are based mostly on
sine waves. The calculations should then be iterated and the losses
considered. A modern offline SMPS by itself, i.e. without a PFC, attains efficiencies of 90 to 95 % at 230 V. A modern wide-range PFC
has an efficiency of 97 to 98 % at 230 V, the combination achieves
close to 90 % with a transformer loaded to its limits; if the transformer
is allowed a bigger size, this can be improved. The losses in the
electrolytics are considerable, due to the dual high rms load by the
transformer and load currents, the latter is often overlooked; even
if the load is straight DC, in phase 2, the discharge phase, it will be
supplied solely out of the capacitor. The stress on the capacitor is
far higher than e.g. in a converter where there is a fairly continuous
stream of current, e.g. in a buck in CM; here, the capacitor sees only
the hf ripple. The placement of the transformer and the electrolytics is
most critical, the author has seen many SMPS where all electrolytics
were placed around the transformer, a worse case is inconceivable.
This is a matter of FB circuit design which may be covered in a future
6.7. Test And fine adjustments.
Electrical stresses on the semiconductors, the semiconductor, transformer winding and ferrite temperatures and especially the temperatures of the electrolytics under worst case conditions are a minimum
of final design tests. The emi tests are so special that they require a
separate description.
7. Selection of the switching transistor.
There are quite some requirements on a FB switch; it is not sufficient
if the voltage, current and wattage ratings are met for the intended
load. A SMPS does not only know normal operation, but there are
also overload situations like turn-on. At turn-on all capacitors on the
secondary are still empty and present almost a short. Regulation is
mostly from the secondary, at turn-on there is no secondary voltage
so the loop does not yet function. In order for the supply to reliably
start under full load, the primary must be designed for a much higher
maximum load than during regular operation; the supply thus starts
with this maximum power and runs the secondary voltage up until the
regulation circuit is powered up and senses that the desired voltage
is reached and, e.g. via an optocoupler, causes the primary control
circuit to reduce the charge per period to the actual level required by
the load. This must function with maximum load, at the highest ambient temperature and after all components reached their maximum
temperatures. Si power mosfet's profit from several decades of
improvement, most notably they can withstand enormous overloads
November 2016
Bodo´s Power Systems®
by going into a nondestructive avalanche mode; the overload is converted into heat, and the part will survive as long as its maximum Tj is
not exceeded.
For some time to come, Coolmos (Infineon designation, other firms
use the term "superjunction") switching mosfet's remain the best
choice. Coolmos mosfet's combine lowest on resistance with low
capacities, and they switch extremely fast, especially in cascode, in
offline SMPS down to 5 ns. Too fast rise and fall times resp. dv/dt's
cause excessive emi and dielectric stresses and are undesirable. In
other words: Si Coolmos is already faster than needed.
The latest Coolmos families also have their drawbacks which are,
understandably, not pointed out by the firms. A most important parameter is the Rthj-c . The low capacities and fast switching came about by
drastically shrinking the chips; this has a deleterious effect on the heat
transfer which can go so far that these chips get much hotter inspite
of lower losses and come closer to destruction under overload! This is
a stern warning: in general the power transistors which were designed
in can not simply be replaced by a newer family without prior checking
everything thoroughly! Some firms shrunk the chips but left the data
sheet unchanged... Comparable SiC and GaN chips are much smaller
than Si chips, so their Rth is much worse!
This is why the older families of Coolmos like C2 or 3 are still being
made, because their chips are larger and more robust. In cascode
they are as fast as the newer ones and to be preferred if at all possible. Note that in a cascode, it is the lower transistor which dictates the
performance, this is why older and sturdier chips are not outmoded.
For offline SMPS up to some ten watts, the best solution are two-chip
components as they are available from several manufacturers, invented some years ago by Samsung; the best contain a Coolmos and
a bipolar control chip with CMC. Such parts which switch operating
modes back and forth are only useful for known, constant loads.
Do not fall for the shrilly siren tunes from GaN and SiC manufacturers! Extensive tests have shown that in FB SMPS neither SiC nor
GaN outperform Si Coolmos, see the article "Performance comparisons ... " in Bodo's Power.There is no reason to go out on a limb and
replace proven Si Coolmos by any of the new components which
are no better in the first place, unproven, more expensive, especially
the composite parts, without second sources. The boisterous, mostly
ridiculous claims of the manufacturers are directed mainly to their
investors who are no engineers and must be convinced that Si will be
replaced by their products so fast that gigantic profits are to be expected. In particular, since many years over and over, they tout that Si
will disappear within 2 years. Serious drawbacks are not mentioned,
e.g. only one firm openly admitted that their GaN transistors can not
take any overload but will be destroyed on the spot, this applies to all.
Also SiC transistors have no avalanche spec sofar.
8. Selection of the rectifier diodes.
In general, ultrafast Si diodes are the types of choice. Hyperfast
diodes trade speed for increased forward voltage such that they
can only be used for high output voltages e.g. in PFC's. Si Schottky
diodes make only sense for low voltage outputs; there are types up
to 200 V available, but they offer no advantage vs. ultrafast types.
Too often their very high capacitances are overlooked. For PFC's SiC
Schottky diodes became standard, because their losses are lower
than those of UF diodes, and due to their fast recovery the losses
in the switching transistor are drastically reduced. Funny enough,
their manufacturers never realized that these 600 V diodes are also
superior to Si UF diodes for all secondary rectification purposes down
to about 24 V. The property of SiC comes to bear that the losses do
not increase with temperature. Modern SiC diodes differ from the first
ones by the integration of a junction diode in parallel which takes over
at higher currents, alleviating the sensitivity to overload. Nevertheless
SiC diodes have to conservatively derated!
27th International Exhibition
for Electric Automation
Systems and Components
Nuremberg, Germany, 22 – 24 November 2016
entry tick
Your free
Answers for automation
Gain first-hand experience on electric automation
than 1,650 exhibitors
82 • moreBodo´s
Power Systems®
• products and solutions
• Industrie 4.0 Area
November 2016
9. The art of correct measurements within
Correct measurements inside SMPS require
solid knowledge and appropriate measuring
instruments. 1. A high quality oscilloscope
with a bandwidth of at least 200 MHz. 2.
Passive probes 100:1 and 10:1. 3. A DC/AC
current probe with a bandwidth of at least 50
MHz and a sensitivity range of 1 mA/cm to at
least 1 A/cm. 4. A power analyzer.
Scopes: Lucky is the owner of a high
performance analog scope like a Tektronix 7000 series model, but there are only
Digital Storage Osciloscopes (DSO's) on the
market. While analog scopes show the signal
itself and are absolutely reliable, hence need
no explanations, this is not so with DSO's.
They do not show the signal, but only a more
or less correct reconstruction which can be
grossly distorted and bear no resemblance to
the original! Even a superficial listing of their
problems would require some 30 pages, here
just a few important hints. The bandwidth of
DSO's is not constant, but depends solely on
the memory length and the time scale selected, totally irrespective of the maximum values. All manufacturers lie about the fact that
the bandwidth is dependent upon the actual
sampling rate and advertise e.g.: "Max. sampling rate 5 GS/s, bandwidth 500 MHz". This
is factually wrong, because it should run:
"Max. bandwidth". The bulk of DSO's in use
are low and medium-priced instruments with
memories between 1 and 10 KB because
they employ cheap, noisy and rather poor
socalled MOS CCD's. Such scopes have no
place in a SMPS lab! Example: the current
in the choke of a PFC shall be measured; in
order to see a full cycle of 50 Hz, the sweep
rate must be set to e.g. 10 ms/cm. With a
memory length of 1 to 10 KB the maximum
sampling rate can not be upheld because
of memory overflow; the scope must hence
reduce the actual sampling rate from e.g. 5
GS/s to 50 KS/s. The actual sampling rate
is not displayed, but concealed somewhere
deep in the menus. The bandwidth is 1/10 of
the sampling rate, i.e. 5 KHz, not 500 MHz!
The PFC converter's 100 KHz sawtooth
riding on top of the 100 Hz half-sine will be
suppressed resp. all kinds of aliases will be
shown if anything. In SMPS labs DSO's with
at least 1 MB are required. Another example:
DSO manufacturers advertise an "improved
resolution" function. Yes, 3 more bits of
vertical resolution can be obtained, what
they do not say: at a price, and this price is
a most drastic bandwidth reduction because
this function does nothing else but averaging which is the same as low-pass filtering!
If a lab still has an analog scope it should be
kept and always used as the reference when
DSO displays are doubtful.
Probes: In SMPS the preferred probes
should be 100 : 1. 10:1 probes seldomly can
take more than 400 Vp, also the loading is
much less and ground connection problems
are not as critical. Probes contain an enormous knowhow, it is highly recommended
to use only probes from the few renowned
manufacturers, others may show grossly distorted high amplitude signals. In order to precisely measure the peak voltages in SMPS,
the probes must be carefully adjusted to the
specific scope input which includes the adjustment elements in the compensation box,
a pulse generator with a rise time of < 0.5 ns
and a perfect top is required, the calibrator
output of the scope is not good enough. For
these adjustments a special 50 Ohm connector with a probe soocket is necessary like the
Tektronix type 017-0088-00.
Even with 100: 1 probes the use of the
probe's ground strip should be avoided. Most
SMPS use E cores and similar types for cost
reasons. These have intense stray fields
which couple into the loop of the ground strip
and cause grossly erroneous measurements.
Especially output ripple measurements are
thus corrupted. Reliable results can only be
obtained by installing the special Tektronix
probe connectors in the circuit und inserting
the probe with the ground strip removed!
However, these don't fit the 100:1 probes.
RMS measurements: The importance of
rms current measurements was already
mentioned. Because all components are soldered to the board, it will be necessary to unsolder the part to be tested. For this purpose
take a 20 cm length of Teflon stranded wire,
fold and twist it, leaving a small loop for the
current probe. This is then soldered between
one wire of the part and the board. This will
even work between the drain of the switching
transistor and the transformer. Any DSO can
calculate the rms value from the waveform.
The current probe amplifier output can also
be connected to a true rms instrument.
Warning: only DC/AC current probes may be
used, no AC probes, they would saturate and
display nonsense.
Very High
Power Density
DC/DC converters for
industry and electromobility
Wide range of PCB mounted
DC/DC converters for industrial,
battery-powered, and railway
9-36VDC and 18-78VDC,
power from 8W to 240W
CE & EN50155 certified
for railway applications
CE & EN60950 certified
Visit u
in Hall s
Booth 2,
High efficiency up to 93%
Temp. range -45°C to +85°C
30W 1″x1″, 60W 2″x1″,
180W 1/4 brick, 240W 1/2 brick
Dr. Arthur Seibt
Lagergasse 2/6; A 1030 Wien, Austria
Tel. +43-1-505.8186
email: [email protected]
November 2016
w w w. r e c o m - p o w e r. c o m
1200W - 1500W Modular Power Supply Series has Lowest Acoustic Noise
TDK Corporation announces the introduction
of the QM series of AC-DC power supplies;
the first 1200W to 1500W rated modular
series to have full MoPPs isolation, with the
lowest acoustic noise available on the market
at that power level. This product is the latest in a 37-year heritage of modular power
supplies, beginning with the invention of the
ML series - a world first in 1979. Having
both medical and industrial safety certifications, the QM is suitable for a wide range
of applications, including BF rated medical
equipment, test and measurement, broadcast, communications and renewable energy
Product definition is made extremely easy
with an online configurator, giving an
optimised module selection with a choice
of signals, leakage current and standby voltages. Once the desired output voltages and
currents have been entered, the configurator
will automatically produce a unique 8-digit
code for easy order placement.
With a wide range 90-264Vac, 47-440Hz
input the QM can deliver 1200W, and 1500W
with a 150-264Vac input. Up to 16 outputs
can be provided, with voltages ranging from
2.8V to 52.8V, with additional higher voltages
later. Module power ratings start with a low
power 35W, with single or dual outputs, all
the way to a powerful 1200W single output.
The units will operate in ambient temperatures of -20 to +70°C, derating output
power and output current by 2.5% per °C
above 50°C. Overall case dimensions are a
compact 176 x 63.3 x 270mm (W x H x D),
and the weight is between 2.6 and 3.1kg,
depending on the module configuration. The
QM is covered by an industry-leading sevenyear warranty.
External Ac-Dc Power Supplies Now Meet EU’s CoC Tier 2
Efficiency Standards
CUI announced that the majority of its line of external ac-dc power
supplies now meet the European Union’s (EU) Code of Conduct
(CoC) Tier 1 and CoC Tier 2 efficiency standards.
The European Union’s CoC Tier 1 effectively harmonizes the EU with
US DoE Level VI and became effective as a voluntary requirement
from January 2014, two years ahead of Level VI. Its adoption as an
EU Ecodesign rule is currently under review to become law with an
implementation date of January 2017. The more stringent CoC Tier 2
requirement became effective on a voluntary basis from January 2016
and is similarly under review to become law as an Ecodesign rule
from January 2018.
The key difference between the CoC requirements and Level VI is the
new 10% load measure, which imposes efficiency requirements under
a low-load condition where historically most types of power supplies have been notoriously inefficient. While CoC Tier 1 includes the
new 10% load measure, its no-load and active mode limits are less
stringent than DoE Level VI. CoC Tier 2 further tightens the no-load
and active mode power consumption limits for key classes of power
adapters enacted by Level VI and covers both standard voltage and
low voltage adapters.
Hermetically Sealed Aluminum Electrolytic
Capacitor Line Expands to Higher Voltages
Cornell Dubilier Electronics, Inc. has announced significantly higher
voltage values for its MLSH Slimpack series of hermetically sealed
aluminum electrolytic capacitors. These unique components feature
a true glass-to-metal seal that prevents dry-out of the capacitor electrolyte. As a result, they provide extraordinarily long life (5,000 hours)
to more than meet the most demanding applications for military and
aerospace. With the line expansion just announced, they are now
available in nine values, from 120 µF to 3,200 µF, with ratings up to
250 Vdc.
The hermetic Slimpack is an offshoot of the well-established nonhermetic Flatpack series. The new MLSH Slimpack can be used as
a replacement for parallel and series banks of wet tantalum capacitors for both new and existing designs. This change is especially
beneficial where bulk storage is paramount. MLSH Slimpack devices
measure 1.0” x 1.5” x 0.5”, weigh less and have more capacitance
than a parallel bank of three or more wet tantalum capacitors. High
rated capacitance is maintained at temperatures as low as -55 °C, a
Bodo´s Power Systems®
key requirement for power supplies used in
military and aerospace applications. Topend temperature rating is 125 °C.
All devices in the MLSH Slimpack series
feature rugged stainless steel cases with a
vibration rating of 80g. In most applications,
there are significant cost and space savings
vs. a comparably rated bank of wet tantalum capacitors. The new
high-voltage MLSH Slimpacks eliminate the need for series banking.
Since its founding in 1909, CDE has been dedicated to advancing
capacitor technology for new applications. The company combines innovative products with engineering expertise to provide reliable component solutions for inverters, wind and solar power, electric vehicles,
power supplies, motor drives, HVAC, motors, welding, aerospace,
telecom, medical equipment and UPS systems.
November 2016
200 Degree Celsius High Voltage Capacitor Products
KEMET Corporation announced the addition
of EIA 2824, 3040, 3640, and 4540 case
sizes to the HV-HT surface mount multilayer ceramic capacitor product line. These
added case sizes allow for up to 150 nF
of temperature-stable capacitance in 200
degree Celsius environments.
voltage, frequency, and temperature in Wide
Band Gap semiconductors (SiC, GaN) technology, KEMET’s HV-HT capacitors are ideal
for realizing design goals such as maximum
power density and efficiency.
Typical applications include critical timing,
tuning, circuits requiring low loss, circuits
with pulse, high current, switch mode power
supplies, high voltage coupling, DC blocking,
and voltage multiplier circuits in extreme
environments such as down-hole exploration, aerospace engine compartments, and
geophysical probes.
HV-HT capacitors were developed under KEMET’s patented 200°C high reliability base
metal electrode (BME) dielectric platform
which is 100% Pb-free, RoHS and REACH
compliant without exemptions. In addition to
a low profile and aspect ratio, these devices
exhibit low inductance and ESR with respect
to application frequency, allowing for very
high ripple current capability.
HV-HT capacitors are specifically designed
to withstand the demands of harsh industrial
environments such as oil exploration and
automotive/avionics engine compartment
circuitry. In addition, with the
growing trend of increased
WT1800E Precision Power Analyser
Yokogawa Europe announced the launch of
the WT1800E power analyser. This instrument introduces new standards of accuracy
and flexibility in power measurement across
a broad range of applications.
Today’s energy conscious world is driving
efficiently. Renewable energy markets like
photovoltaic and wind power are growing and
we are seeing accelerated development of
environment-friendly vehicles and energyefficient home and industrial appliances.
Developing these technologies requires
engineers towards an ever-increasing focus
accurate measurements of electric paramon energy conservation, energy efficiency
eters like voltage, current and power to
and renewable energy. Engineers seeking to
validate even the smallest changes in energy
improve their product through better energy
consumption. In the WT1800E high performanagement therefore need more accurate
mance power analyser, engineers have the
power measurement.
ideal tool to accurately measure power, its
To curb global warming, greater efforts are
quality and efficiency.
being made to generhttp://tmi.yokogawa.com/products/digital-power-analyzers/
ate and use power more
November 2016
Bodo´s Power Systems®
Connecting Global Competence
Planet e:
Where the
future begins.
Electronics of tomorrow.
Tickets & Registration:
World’s Leading Trade Fair for Electronic Components,
Systems and Applications
Bodo´s Power Systems®
Messe München I November 8–11, 2016 I electronica.de
November 2016
Bussmann™ Series Fuses Meet Higher
Power Requirements
Power management company Eaton announced the Bussmann™ series 1025HC
high current fast-acting surface mount
(SMD) fuses designed for the power
requirements of new generation telecommunications and computing systems.
The compact footprint of the new series
allows developers to utilize less circuit
board space with a higher current rating,
higher voltage rating and improved performance. The increased current squared
time (I2t) ratings in the 1025HC are ideal
for power systems that are used in next
generation server and desktop computers, gaming consoles, industrial and
consumer electronics, storage systems,
base stations, test equipment and LED
“Our goal is to bring new technology that
adds value to our customers’ designs,”
said Matt Joiner, global product manager,
Eaton’s Electronics Division. “This line of
fuses offers high-current-carrying capability
in a very compact, surface-mount package
saving customers space in their advanced
high-power designs.”
The 1025HC offers one of the industryleading I2t ratings from 25 to 600 amperes
squared seconds (A2s). It provides a superior solution to power system designers who
require performance and technical advance-
Ensure optimum
system performance!
ment in demanding operating environments. The 1025HC fuse delivers excellent
protection during in-rush currents on start-up
helping to eliminate nuisance fuse openings.
The compact size of the 1025HC, 10 mm in
length by 3.15 mm in height by 1.7 mm in
width, helps designers utilize board space
Digital LED Driver ICs with Flicker Free
Control and Low Stand-by Power
Modern LED technology offers many advanced possibilities for lighting applications.
With the XDPL8220, Infineon Technologies
AG extends its product portfolio of digital and
configurable LED driver ICs. The IC enables
the lighting industry to realize essential
features for smart lighting and increases the
benefits for both, end user and manufacturers. The primary side control of this device
saves extra components, reducing cost on
the one hand and increasing reliability on the other.
The XDPL8220 offers a modern
two stage architecture, significantly
easing the implementation of up
and coming flicker standards. This
was made possible by eliminating
the low frequency variation from
the mains supply and guaranteeing a stable output. Aiming at high
energy efficiency, the low stand-by
power facilitates permanent operation of the Electronic Control Gear
(ECG). With a stand-by power of
less than 70 mW, the XDPL8220
reduces power consumption in
the non-active mode while still reacting to
external events or user requests.
Engineers in more than 50
countries trust in the
Vector Network Analyzer
Bode 100 when great
usability, high accuracy and
best price-performance
ratio are needed.
Measure from 1 Hz to 40 MHz:
• Loop gain / stability
• Input & output impedance
• Characteristics of EMI filters
• Component impedance
November 2016
Smart Measurement Solutions®
Bodo´s Power Systems®
2016-07-28_optimum performance_third_electronica.indd 12016-07-28 07:47:14
3-5 Cell Li-Ion/Li-Polymer Battery Protection Analog Front-End IC
Ricoh Europe (Netherlands) B.V. Semiconductor Support Centre
has launched a Analog Front-End (AFE) Integrated Circuit (IC), the
R5601. The new addition to Ricoh’s battery management ICs portfolio
is designed for Lithium-ion or Lithium-polymer battery packs of three
up to five cells in series. Applications include cordless power tools,
robot vacuum cleaners, lawn mowers, and electrical bikes.
Traditional stand-alone protection ICs have internally fixed hardware
settings for various monitoring thresholds that decide to interrupt the
charge or discharge process and control the external dual MOSFET.
The advantage of an AFE IC versus a stand-alone protection IC is
the flexible concept enabling short time to the market. An AFE IC is
a measurement device of which the results are forwarded to a MCU
(microcontroller unit). The fact that all threshold settings are saved
within the software of the MCU and software settings are easily adaptable means a significant advantage. Therefore, an AFE IC is more
convenient to use as a standard measurement IC for various battery
pack designs.
LED Drivers for Integrated Lighting
RECOM DC current (RACD04) and DC voltage (RACV04) LED
drivers have been designed for cost-sensitive applications. Safety
extra low voltage or SELV terminals are ideal for supplying power to
integrated LED lighting.
RECOM LED drivers can be built into furniture, hidden under shelving, or integrated into applications with extremely space available
due to their their compact design, including cable connections. The
RACD04 series is aimed at LED spotlights, reading lamps and accent
lighting, whereas the RACV04 series is ideal for LED strips, background and wall lighting, and furniture and cabinet illumination. Each
4W AC/DC LED power supply in IEC protection class II packed into a
compact casing measuring 38 x 27 x 21 mm at a weight of just 40g.
The RACD04 series has 350, 500 and 700 mA DC current terminals,
whereas the RACV04 has 12 and 24V DC current terminals. The isolation voltage in either case is 3.5 kV AC/1 minute. These LED drivers
are classified to IP65, and are therefore suitable for damp areas such
as kitchens and bathrooms, or covered garden or outdoor facilities.
The drivers provide complete protection against short circuit, overvoltage, overload and overheating, and are certified according to CE,
DB, CSA and UL8750 standards and comply with the RoHS2.1 and
ErP directives. Both series have a three-year guarantee. RECOM
RACD04 and RACV04 are now available from the RECOM distribution network.
TVS Diodes Suited to High-Speed Interfaces
The continuous growth of data traffic - driven by smartphones,
wearables and applications such as virtual reality and the Internet of
Things (IoT) - leads to increasing numbers of high-speed interfaces
that typically require protection against ESD events. To address this
demand, Toshiba Electronics Europe has released ESD protection diodes based on its 4th generation ESD diode array process (EAP-IV),
which uses Toshiba’s proprietary snapback technology.
The DF2B5M4SL, DF2B6M4SL, DF10G5M4N and DF10G6M4N offer
protection for high-speed interfaces including USB 3.1 applications. A
choice of operating voltages (3.6V and 5.5V) and packages (SOD962
and DFN10) provides flexible options for realizing ESD protection in a
variety of designs.
Thanks to Toshiba’s new process, the four devices simultaneously
deliver low capacitance, low dynamic resistance and high ESD endurance. Minimum signal distortion of high-speed data signals is guaranteed by the ultra-low capacitance of 0.2pF, while a typical dynamic
resistance of RDYN=0.5 Ω ensures low clamping voltages. High ESD
protection levels are supported as electrostatic discharge voltages of
at least ±20 kV according to IEC61000-4-2 are guaranteed.
Bodo´s Power Systems®
November 2016
16bit ‘Tough’ MCUs Optimized for Rechargeable NiMH Applications
LAPIS Semiconductor, a member of the ROHM group, has announced the development of the ML620130 family of 16bit low power
MCUs, optimized for compact industrial equipment requiring battery
drive in noisy environments, featuring superior processing capability
with low power consumption.
In addition to clearing the ±30kV noise measurement limit, operating
voltage has been successfully reduced to 1.6V. Optimizing the operat-
ing voltage to an integral multiple of standard nickel metal hydride
batteries (NiMH, 0.8V× 2=1.6V) ensures efficient use without wasting
battery charge. This contributes to decreased battery consumption,
prolonging battery life in portables and battery-equipped industrial
equipment. The ML620130 family consists of 9 models offered in a
variety of memory capacities, pin counts, and other characteristics
that make it possible for users to select the ideal solution to fit set
To provide increased miniaturization and lower costs,
many applications are opting to eliminate noise and/
or thermal countermeasures. However, this makes
it difficult to balance the conflicting demands for
increased environmental resistance while maintaining safety utilizing fewer external parts. In addition,
reducing the size of the battery mounted in the
module while also increasing battery life requires
that the power consumption (voltage and current) be
optimized for each application.
To meet these disparate needs, LAPIS Semiconductor has expanded its lineup of market-proven 16bit
low power microcontrollers to include the ML620130
family of ‘tough‘ MCUs that incorporate multiple functions optimized for battery drive operation.
assembled Cores Done Right
for unique
Visit us at
BOOtH# B6.512
+1 412 696 1333
[email protected]
+852 3102 9337
[email protected]
November 2016
Bodo´s Power Systems®
Solid State Relay with a 170mA Load
Current Rating
IXYS Integrated Circuits Division, Inc., a wholly owned subsidiary
of IXYS Corporation, announced the immediate availability of the
CPC1010N, 250V, single-pole, normally open (1-Form-A) Solid State
Relay (SSR) with a 170mA load current rating. The CPC1010N is
specially designed to provide the best combination of performance,
size and price.
Packaged in a 4-pin SOP, the CPC1010N employs optically coupled
MOSFET technology to provide 1500Vrms of input to output isolation.
The efficient MOSFET switches and isolated driver IC use IXYS ICD’s
patented OptoMOS architecture and technology.
IXYS ICD’s advanced molded vertical construction makes the
CPC1010N one of the world’s smallest relays and offers board
space savings over the competitor’s larger 4-pin SOP relay. Additional features include low drive power requirements, high reliability,
arc-free with no snubbing circuits, and no EMI/RFI generation. The
CPC1010N is designed to
replace, and offers superior
reliability over, electromechanical relays.
The 250 volt load voltage
rating, 170mA load current
rating and 11.5 ohms of on-resistance makes this device suitable for
industrial applications, instrumentation, multiplexers, data acquisition,
and electronic switching.
Approvals include: UL Recognized Component: File E76270, CSA
Certified Component: Certificate 1172007, EN/IEC 60950-1 Certified
Component: TUV Certificate B 13 12 82667 003.
High Isolation DC-DC Converter for High Power
Power electronics specialist, Amantys Power Electronics Limited, has
announced the launch of a high isolation DC-DC converter to power
high voltage gate drives for 4500V and 6500V IGBT modules. The
DC-DC converter is designed to comply with international standards
including EN 50155 for railway applications and IEC 61800-5-1 for
variable speed motor drives.
The DC-DC converter can accept a 15V or 24V input voltage and has
an output power of up to 15W making it suitable for driving several
gate drives in parallel. It features low quiescent current of 50mA and a
maximum current output of 1100mA. The DC-DC converter provides a
minimum of 12kVrms working voltage isolation, making it suitable for
two level and three level converter topologies.
The robust design can operate from -40ºC to +85ºC and has protection features for: input reverse polarity; input overvoltage; output short
circuit; and output overload. The protection features are complemented by status LEDs that indicate DC-DC operation and fault conditions.
The isolated DC-DC converter will be used in applications such as
railway traction, HVDC infrastructure and medium voltage motor
drives. The DC-DC converter is available for sample orders in Q4
November 2016
Bodo´s Power Systems®
High Performance, Compact, Medium and High Voltage
Transducers for Traction
LEM introduces the DVM series for insulated nominal voltage measurements in traction applications. This family of devices spans the
range from 600-4200 VRMS and incorporates LEM’s proven and
patented insulating technology. Despite achieving very high levels of
isolation with a safety insulation voltage of 12 kV, the DVM transducer
is compact, measuring only 138 x 63.4 x 69 mm.
It is fully compatible and out-performs the previous generation of
transducers in terms of functions and performance with new improved
levels of accuracy and temperature stability, thus greatly simplifying
DVM is also suitable for base mounting, but with slightly different outline dimensions to take into consideration for primary and secondary
connections locations for example. The DVM is 30% smaller in height,
occupies 25% less volume and is 56% lighter! The reduction in size
does not compromise the DVM’s high immunity against the external
surrounding perturbations or against the high voltage variations. The
new size is also an advantage when confronted with size constraints
in modern railway propulsion converters.
LEM developed the DVM to be fully compliant with the International
Railway Industry Standards (IRIS), providing engineers in the railway
industry, who are working with both rolling stock and
sub-stations, with a versatile transducer that is equally applicable to
measuring network voltages, or the main converter DC link on-board
Miniaturization with High-Temperature
Surface-Mount Silicon Controlled Rectifiers
STMicroelectronics has introduced the industry’s first 800V surfacemount Silicon Controlled Rectifiers (SCR, or thyristor) specified for
operation at temperatures up to 150°C without derating, giving freedom to miniaturize power modules for applications that demand high
reliability in harsh conditions.
With its 80A current rating, the new TM8050H-8 SCR, housed in the
High-Voltage D3PAK (TO-268-HV), enables mid-power applications
in the 1-10kW range to leverage surface-mount assembly efficiencies
and reduce PCB and heatsink sizes, lowering system cost. The package has very low junction-to-case thermal resistance of 0.25°C/W,
ensuring efficient heat dissipation, and a large pin-to-tab creepage
distance of 5.6mm that gives a large safety margin in the presence of
high applied voltages. A TO-247 package option is also available.
The TM8050H-8 is the latest addition to ST’s family of SCRs that all
bring state-of-the-art device and package technologies to automotive
and industrial power control. Spanning current ratings from 12A to
80A, the devices enable designers to create extremely compact and
reliable car or motorcycle voltage regulators, induction motor starters,
soft starters, industrial heater or cooker controls, Solid-State Relays
(SSRs), uninterruptible power supplies (UPS), and AC-line conditioners.
With low dynamic resistance (RD) and on-state voltage (VTO) of 5.5
mΩ (TJ = 150°C) and 0.85V respectively, and leakage current of 20µA
max (at 800V, Tj = 25°C), the TM8050H-8 ensures extremely high
energy efficiency under all operating conditions.
Optical PHY leadership with New Generation of Gearbox Products
Broadcom Limited announced a generation of gearbox PHY IC devices, the BCM82332, BCM82793 and BCM82864, designed for data
center and enterprise networking applications.
With the general availability of switch chips that have flexible multispeed port interfaces supporting 10G, 40G, 100G and emerging Ethernet speeds, new gearbox PHY solutions are needed to ensure robust signal integrity and optimize port adaption. The latest Broadcom
gearbox PHY devices extend the capabilities of switches by providing
Bodo´s Power Systems®
flexible port interface andnmaximizing the electrical link distance to
optical interface while consuming minimal power. The BCM82793 and
BCM82332 are application-optimized, cost-effective 100G gearbox
devices that support IEEE802.3bj Clause 73 Auto-negotiation, Clause
91 FEC and Clause 92 & 93 transmit training.
November 2016
Record-Breaking Low Profile Product Series
Power, leading designer, manufacturer and seller of high efficiency
and small size Power Supplies, will present its new Ultra Low Profile
(ULP) series of Medical and Commercial, High Efficiency, High Power
Density AC/DC power supplies at the Electronica 2016 in Munich.
The new (M)ULP 180 & 250 Watt Series have power densities which
reach 30 watts per cubic Inch with maximum height profiles of 0.75
Inches (19.05mm).
The (M)ULP40, 180 Watt Series offer existing and new
customers the possibility of using standard 2 x 4 and 3
x 5 inch open frame medical and
commercial grade power solutions
in height restricted environments
without having to change mechanical footprints while increasing efficiencies. The 0.75 Inch (19.05mm)
height profile is a market first in this
AC/DC open frame power range
which expands even further the
options available to EOS Power
customers and their applications.
The MWLP225 and MWLP350
series of medical open frame products released last year by EOS
Power have been a global success
due to their power densities and
market leading specifications.
To further enhance the appeal
of these medical products to our
medical customers EOS Power is
now offering extended warranty
option on both the 225 and 350
MWLP product series. See EOS at
electronica, Hall A2/475.
Reference Design adds Wi-Fi Capability to
Electric Vehicle Charging Stations
Texas Instruments introduced the first reference
design that adds Wi-Fi connectivity to an electric
vehicle (EV) charging station. EV owners will now be
able to remotely monitor and control the charging of
their vehicles from just about anywhere with Wi-Fi,
presenting dozens of potential use cases from home
automation to checking the availability of nearby
public charge points. Download the new EV Charging Station with Wi-Fi reference design.
Battery technology advancements and government
regulations have resulted in a growing number of
new electric vehicles around the world. But EV
makers still need more charging stations to make it
easier for drivers to charge their vehicles. The new
reference design uses TI’s SimpleLink™ Wi-Fi wireless microcontroller (MCU) technology that allows
design engineers to create stations that intelligently
charge at non-peak times or detect and communicate when a charging station is available.
One barrier to widespread EV charging station
adoption is the amount of time it takes to charge
a vehicle. The reference design supports Level
1 charging, which is compatible with household
outlets, as well as Level 2 EV charging, which helps
vehicle owners tap into higher current (15A to 30A
and higher) connections available in commercial
office buildings. Level 2 chargers typically take up to
eight hours to fully charge the vehicle if the owner
wants to plug it in while at work.
Later this year TI plans to introduce a Level 3 EV
direct-current charger reference design scalable
to 600V and 400A that cuts charging time down to
only 20-30 minutes – enough time to stop at a Wi-Fi
enabled restaurant that has a charging station and
charge the vehicle during lunch. Read the blog post,
Electric vehicle charging stations are getting smarter
and charging faster.
November 2016
5 x 3 x 0.75 INCHES
UP TO 93%
12 V/0.5 A FAN
Hall A2, Stand 475
High-Speed, Small Footprint RS-485 Receivers for
Industrial Applications
Exar Corporation announced a family of high-speed RS-485/RS-422
receivers that accommodate the complex demands of next generation
high-speed serial communications designs. The XR33180, XR33181,
XR33183, and XR33184 occupy a 3mm x 3mm footprint and feature
52Mbps data rates, fast 15ns propagation delay and 2ns maximum
receiver skew. This family is best suited for high performance industrial applications such as multi-drop clock distribution, telecom networking, robotic control, process automation and local area network
The XR33180/81/83/84 (XR3318x) operate from either a 3.3V or 5V
supply and are available in tiny 5-pin and 6-pin TSOT23 packages,
ideal for high-speed point-to-point RS-485 applications where space
is a concern. The XR33184 includes a low voltage logic pin that eliminates the need for a level shifter when interfacing to MCUs or FPGAs
that run off of lower supply voltages.
All devices feature ±15kV ESD protection on the receiver inputs and
exceed the highest ESD rating of IEC 61000-4-2 and operate over an
extended temperature range of -40°C to +125°C.
Monolithic Power Modules for DC/DC Applications
Mouser Electronics, Inc. is now stocking the Monolithic Power Modules (MPM) from Monolithic Power Systems (MPS). Providing
superior performance and inherent reliability by eliminating unnecessary assembly steps and minimizing external components, the MPM
family comprises integrated DC/DC solutions that consist of a monolithic regulator, passive components, and a mold compound all in a
single chip. Based on MPS’ innovative single-step assembly process,
these step-down DC/DC converters offer circuit simplicity for ease of
design; are highly efficient; yield superior noise, ripple, and transient
performance; and are available in a compact module form factor that
is up to 50 percent smaller than competing solutions.
The MPS Monolithic Power Modules, available from Mouser Electronics, are high-frequency, synchronous, rectified, step-down, switchmode converters with built-in power MOSFETs and integrated inductors. Each of the modules can regulate a continuous output current to
input voltage with excellent load and line regulation results. The modules operate on a switching frequency of up to 2.4 MHz to achieve a
fast load transient response, and feature a wide operating input range
of up to 36V. The devices also include a variety of protective features
including over-current protection and thermal shut down.
Bodo´s Power Systems®
November 2016
The H-Bridge to up Customers
Application’s Speed and
Vincotech, a supplier of module-based solutions for power electronics
announced the launch of a fastPACK 0 HC improving UPS, SMPS,
solar and welding applications in speed, efficiency and full-current
capability for bidirectional usage. The H-bridge module is available in
the low-inductive flow 0 housing.
The fastPACK 0 HC module featuring high frequency 650 V IGBT H5
technology in combination with a fast diode enables outstanding efficiency, reliability and perfectly balances cost and performance. This
utmost efficient H-bridge
module delivers 30+ kHz
fsw and is suitable for soft
switching to reduce switching losses.
The fastPACK 0 HC
enables bidirectional operation for charger and SMPS
applications with full current FWD.
It is also equipped with integrated
capacitors reducing electromagnetic interference. Packaged in the low-inductive flow 0 12 mm
housing the module can also be provided with phasechange material on special request.
November 2016
Bodo´s Power Systems®
Power Electronics Capacitors
DC link capacitors
AC filter capacitors
Snubber capacitors
Energy storage capacitors
ZEZ SILKO, s.r.o., Pod Černým lesem 683, 564 01 Žamberk, Czech Republic
tel.: +420 465 673 111, fax.: +420 465 612 319, e-mail: [email protected], www.zez-silko.cz
ZEZSILKO inzerat 210x99 AJ 2015.indd 1
29.01.15 11:46
UFHV Series - Fast Axial Lead High Voltage Diodes
Dean Technology, Inc., announced the reintroduction of the UFHV series axial lead high voltage diodes. This series of diodes are available
in 2,000, 3,000 and 4,000 volt ratings and all have a 75 nanosecond
response time. Recent improvements in manufacturing processes
have made this series of diodes a viable offering again. The work
we’ve done to keep the UFHV series cost effective is a perfect
example of our commitment. We want to be sure that we offer our
customers products that will meet their needs for a very long time.”
With current ratings ranging from 350 to 550 milliamps, the UFHV
series is very flexible and ideal for many uses. These diodes offer exceptional performance for the price, making them a very good value.
Full product details can be found on the Dean Technology website,
and the diodes are in stock and immediately available from Dean
Technology, or through any approved sales partner or distributor.
Next-Generation Output Transistor Arrays
Toshiba Electronics Europe has announced a series of next-generation output transistor arrays that feature a DMOS FET type sink
output. The TBD62183AFNG and TBD62183AFWG are ideal for level
shift applications and for directly controlling photocouplers, LEDs, and
relays that require high-voltage input signals.
The TBD62183AFNG
and TBD62183AFWG
deliver high-voltage
drive capabilities with
an input rating of 30V
and output rating of
50V. An 8-channel sink
type output is incorporated into the small
ABB Semiconductor C3+29
APEX Microtechnology
electronic concepts
Kiel University
Magna Power
Bodo´s Power Systems®
SMD packages, enabling reduced component count for the control of
multiple circuits.
In addition, the TBD62083A-series has an IOUT of 500mA/ch while
the TBD2183A-series is designed for low power applications with a
maximum IOUT of 50mA/ch.
By adopting a DMOS FET type output, the two new transistor arrays
eliminate the need for a base current for the input pin. The devices
achieve operation with a low maximum input current of 0.1mA @
VIN=3V, while delivering very low power consumption. They also have
output characteristics similar to the Vce (sat) properties of a Darlington Bipolar transistor, making them suitable for replacing in-line
bipolar transistors in Toshiba’s TD62083A series.
Advertising Index
PCIM 2017
Power Integration
November 2016
SMT Machines
sps ipc drivers
TDK Lambda
Teledyne LeCroy
Texas Instruments
ZEZ Silco
62Pak. Quality for demanding applications.
Coming from high-power semiconductors, ABB is regarded as one of the world’s
leading suppliers setting standards in quality and performance. ABB’s unique knowledge in high-power semiconductors now expands to industry standard mediumpower IGBT modules. ABB launched in a first wave its 62Pak, 1,700 V voltage class,
150 A, 200 A and 300 A current ratings, phase leg IGBT modules in standard 62 mm
packages. The 62Paks are designed for very low losses and highest operating
temperatures in demanding medium-power applications such as variable speed
drives, power supplies and renewables.
For more information please contact us or visit our website:
ABB Switzerland Ltd. / ABB s.r.o.
[email protected]
Tel.: +41 58 586 1419
Infineon at SPS IPC Drives 2016
Paving the way to smart factories
Our powerful, secure and smart semiconductor solutions are
key enablers for Industry 4.0.
› End-to-end security solutions
› Previously unattainable efficiency and power density levels
› Robust real-time communication capabilities
› Easy to set up evaluation solution for motor drives
› High-power thyristors & diodes for energy-efficient industrial drives
Get inspired by our live demos and expert talks. Hall 1, booth 550.
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF