BP 2014 12
BP 2014 12
BP 2014 12
12-14
The modular and flexible design of VARIS™ offers compelling benefits. The
desired power can be easily achieved via parallel connection of the modules.
You are also free to choose your preferred cooling type. And the use of
standard components makes VARIS™ both cost-efficient and sustainable.
Talk to the House of Competence, because VARIS™ fears no comparison.
Read online and search for key subjects from all articles in Bodo’s
Power Systems by going to Powerguru: www.powerguru.org
Market .............................................................................................. 18
Electronics Industry Digest
By Aubrey Dunford, Europartners
www.bodospower.com
www.tdk.eu · www.epcos.com
The Gallery
se
Get a free test licen
carabinbackhaus.com
al
w w w.plexim.com/tri
VIEWPOINT
CONTENT
Running Again
Fax: +49 4343 42 17 89
editor@bodospower.com
www.bodospower.com
Publishing Editor
Bodo Arlt, Dipl.-Ing.
editor@bodospower.com
Senior Editor It is Christmas time again, and I have fond
Donald E. Burke, BSEE, Dr. Sc(hc) memories of my toy trains. Engineers often
don@bodospower.com played with technical toys when they were
Corresponding Editor children, and electric trains inspired a num-
Dipl.-Ing. Wolfgang Patelay
Wpatelay@t-online.de ber of them, including me. And the fascina-
UK Support tion is still there. Our grandson, soon to be
June Hulme five, looks at my vintage trains with respect.
Phone: +44(0) 1270 872315 He knows they have been running almost as
junehulme@geminimarketing.co.uk long as I have. My parents, mostly my father,
Creative Direction & Production also got me interested when I was a five.
Repro Studio Peschke Mine are HO, the most popular size, mean-
Repro.Peschke@t-online.de ing “Half Zero”, a size started by Marklin
Free Subscription to qualified readers over 100 years ago, while Lionel in the US
Bodo´s Power Systems continued with Size O.
is available for the following
I’ve just returned from the electronica in
subscription charges:
Annual charge (12 issues) Munich, the platform for the world to show
is 150 € world wide what is currently available and what will drive
Single issue is 18 € the next level of design. And sps ipc drives
subscription@bodospower.com in Nuremberg has shown the practical solu-
tions for drives and the control of industrial
circulation print run 24 000 applications.
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 4 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 • Immunity from external fields for your
from 0.2 to 0.6 % of IPN compact design
• Exceptional offset drift of 0.1 % of IPN • 100 % fully compatible vs LEM previous
• Fast response time less than 0.5 μs generation
• Higher measuring range • -40 to +85 °C operation
• 5 compact sizes in a variety of mounting
topologies (flat or vertical)
IGBT Snubbers
RF Mica Capacitors
DC Link Capacitors
-High Current, High Voltage Film
-High Capacitance Aluminum Electrolytic
www.cde.com
AC Output Harmonic Filter Capacitors
CONTENT
NEWS
Expert to Present at SMTA Space Coast and Tampa Bay Expo & Tech Forum
Indium Corpora- tation, Meeting Future Stencil Printing Community College, where he received the
tion’s Ed Briggs, Challenges with Ultra-Fine-Powder Solder Douglas J. Bauer Award for Excellence in
senior technical Pastes, discusses a statistical comparison Chemistry. In addition, he earned his Six
support engineer, of the transfer efficiency of different solder Sigma Green Belt from Dartmouth College.
will present at powder particle sizes, specifically types 3, Briggs joined Indium Corporation in 1990 and
the SMTA Space 4, 5, and 6; and reviews observations of has held a variety of positions in production
Coast and Tampa post-reflow results in both harsh and optimal and technical support.
Bay Expo & Tech conditions.
Forum on Dec. 4 in Briggs is an SMTA-certified process engi- www.smta.org
Kissimmee, Fl. neer. He earned his associate degree in
Briggs’ presen- chemical technology from Mohawk Valley www.indium.com
With the ongoing increase of power densities in power electronics the thermal interface
between power module and heatsink becomes a larger challenge. A thermal interface
material, especially developed for and pre-applied to Infineon’s modules outperforms the
general purpose materials available.
TIM does not only provide the lowest thermal resistance, it also fulfills the highest quality
standards given for power modules to achieve the longest lifetime and highest system
reliability.
Main Features
Best in class thermal resistance 110
80
70
Benefits
60
Reduced process time in manufacturing
Simplified mounting 50
1 2 3 4
Increased system reliability 5
Time in HTS* [Weeks] n MOD-3
Increased system lifetime *HTS: High Temperature Storing, n MOD-2
Stresstest 1000 h, 125 °C
Optimized thermal management n MOD-1
n IFX-Solution
Improved handling in case of maintenance
www.infineon.com/tim
CONTENT
NEWS
• ECPE Tutorial ‘Power Semiconductor Devices & Technologies’ On the ECPE website you always find the up-to-date information on
29 - 30 April 2015, TBD the ECPE Events:
Chairman: Prof. D. Silber (Univ. of Bremen) www.ecpe.org
Bodo’s Power is offering you the chance to win a PICDEM LCD 2 the voltage of the thermistor and continuously displays the tem-
Demonstration Board. The PICDEM LCD 2 Demonstration Board perature in both Celsius and Fahrenheit. The Timer/Clock function
(DM163030) shows the main features of Microchip’s 28, 40, 64 and showing hours:minutes:seconds with a moving second hand, using on
80-pin LCD Flash PIC microcontrollers including the LCD voltage board 32 kHz watch crystal. Also, when using the PIC18F87J90 family
booster and contrast controller. It is populated with the PIC18F87J90. of devices the LCD module can be configured as a charge pump and
A sample LCD glass display is included for custom prototyping. The software contrast control is activated.
glass features 7-segment displays, wipers, thermometers, star bursts,
and other common icons. For the chance to win a PICDEM LCD 2 Demonstration Board, please
visit http://www.microchip-comps.com/bodo-lcd2-dev and enter your
Some of the applications included are: a voltmeter, which measures details in the entry form.
the voltage of the on board potentiometer and displays a voltage
between 0.00V and 3.30V on the LCD, a thermometer that measures www.microchip-comps.com/bodo-lcd2-dev
Representatives
ALLREM SSG Semiconductor Systems GmbH Consystem S.r.l.
94616 Rungis Cedex, FRANCE D79856 Hinterzarten, GERMANY I-20144 Milano, ITALY
Tel: +33 (0) 1 56 70 03 80 Tel: +49 (0) 7652-91060 Tel: +39 02 4241471
E-mail: info@allrem.com Website: www.ssg-de.com Website: www.consystem.it
E-mail: mail@ssg-de.com E-mail: support@consystem.it
www.allegromicro.com/camp1183
GUEST
CONTENT
EDITORIAL
Computer Engineering is 60 years old. Power Systems and Telecom- Acronym puzzle
munications are even older. Forthcoming out of these disciplines The last four years I have been closely
the original electrical engineering societies were founded 130 years involved with conferences and it in-
ago, including the VDE in Germany, IEE in the United Kingdom and trigued me how challenging it can be to
the AIEE in the USA. Power Electronics is comparatively young. come up with an acronym for an event
Despite the fact that power electronics started with the development or organisation that is unique and does
of the mercury arc rectifier that was invented by Peter Cooper Hewitt not twist the tongue. Power electron-
in 1902, it only recently emerged as a discipline that can stand on its ics related acronyms generally make use of a set of seven letters,
own feet. This was triggered by the spurt in technology development “A”,”C”,”E”,”I”,”M”,”P” and “S” and the acronyms have a special ring
of power semiconductor devices that occurred in the early eighties. about them if you say it loudly. Have you once seen the puzzlement
Power MOSFETs, Schottky diodes, GTOs and IGBTs made it possible on the face of a friend or a colleague that is not an engineer when you
to build reliable, efficient and affordable converters for many ap- tell him to which conference you are going? You can see the question
plications and ever improving micro-computers made it possible do in their eyes; “Where do these nerdy names come from?”
advanced real-time digital control of electrical machines. By the end
of the eighties power electronics emerged as a distinct discipline with The largest common denominator is “PE”, standing for power elec-
its own journals and conferences. Last year the IEEE Power Electron- tronics, and is present in almost all acronyms: EPE, ECPE, PEMC,
ics Society celebrated its 25th anniversary and this year the European PESC, SPEC, APEC, PEAC, PEDES, ICPE, IPEC and IPEMC. The
Power Electronics Association became 25 years old. PELS global conference and European acronyms have two “E”s;
ECCE, EPE, ECPE while PCIM is the outlier as the only conference
For the next two years it will be my privilege to serve as President of that does not have an “E” in its name. Asian conferences all start with
IEEE PELS, taking over the reins to steer the professional society an “I” which stands for international: ICPE, IPEC and IPEMC. “A”
during a period when many exiting things are happening in the devel- stands for application and the two conferences, one in the US and the
opment and application of power electronics. It is a great honour to other in China use the same letters in a different sequence: APEC,
become the second European in the history of the Power Electronic PEAC. The meaning of “S” is the most inconsistent because in PESC
Society to take up this position. I look forward to strengthen the col- it means specialist, in SPEC it indicates Southern and in PEDES it
laboration between PELS and European institutions such as the EPE means system.
Association and ECPE when my term that starts on 1 January 2015.
http://ewi.tudelft.nl/epp
Important Author Dates The Seventh Annual IEEE Energy Conversion Congress and
January 15th , 2015: Digest submitted via the website Exposition (ECCE 2015) will be held in Montreal, Canada, on
May 1st, 2015: Notification of acceptance or rejection September 20 - 24, 2015. ECCE 2015 is the pivotal internation-
July 1st, 2015: Final papers with IEEE copyright forms
al conference and exposition event on electrical and electrome-
Other Important Dates chanical energy conversion. To be held for the first time outside
February 16th, 2015 : Submission of Tutorial proposals USA, ECCE 2015, in Montreal, Canada, will feature both indus-
March 31, 2015: Submission of Special Session proposals try-driven and application-oriented technical sessions, as well as
industry expositions and seminars. ECCE 2015 will bring togeth-
: er practicing engineers, researchers and other professionals for
interactive and multidisciplinary discussions on the latest
advances in various areas related to energy conversion.
Please visit http://2015.ecceconferences.org for more informa-
tion or contact the ECCE 2015 Technical Program Chairs at
ecce2015tpc@gmail.com.
For exhibiting at ECCE 2015, please contact Exhibition Chair,
Steve Sprague at ssprague@protolam.com.
For more about Montreal and its surrounding areas, please visit
http://www.tourisme-montreal.org/.
w w w. d y n e x s e m i . c o m
T: +44 (0)1522 502 901 +44 (0) 1522 502 753
E: power_solutions@dynexsemi.com
CONTENT
MARKET
Electrical Engineer
as Marketing Content
Specialist (m/f)
We are an innovative company with 2,900 employees worldwide. With our power modules
and systems, we have been a key presence on the international market for more than 60 years,
and our qualified and motivated staff are essential to this success. At our headquarters in
Nuremberg, where 1,300 of our employees are based, we develop new products and technologies
for application areas such as renewable energies, energy-efficient motor drives, industrial
automation systems, power supplies and commercial vehicles.
VARIS™ –
In addition, technology developed by
Texas Instruments (TI) is also expected
they can also do without some of the bulky
cooling systems.
small compo-
nents with
to contribute to the reduction in standby
power. In March of 2014, Texas Instruments For example, the company hopes to reduce
introduced two flyback power solutions that the size of the charger bricks required by
big power
claim the highest energy efficiency and low- laptops today—or eliminate them altogether
est standby power consumption for 5-100W by incorporating the necessary electronics
ac-dc power supplies. The UCC28910 700V into the computer itself. The more compact
flyback switcher and UCC28630 high-power, design is also important for automobile ap-
green-mode controller with primary-side plications, where space is limited and weight
regulation (PSR) both achieve sub-30mW is important to fuel efficiency. Transphorm
standby power and expand TI’s portfolio of believes it can reduce wasted electricity by
flyback controllers covering a power range 90%.
of ac-dc adapters and power supplies used
in personal electronics, printers, white goods A startup called FINsix has developed laptop
and smart meters. power adapters that are 75% smaller than
their conventional counterparts. The technol-
Additional areas to watch are advances in ogy employed could also be used to improve
technology, components and materials, in the efficiency of a wide variety of devices and
particular advanced semiconductor materi- appliances, including washing machines and
als such as Gallium Nitride (GaN) which is air conditioners.
expected to have an impact on the power
supply market over the forecast period. The The FINSix adapter is GaN-based and is cur-
Adoption of GaN technology will have a rently expensive at $79.00 per unit (Retail for
major effect on the external power supply a PC) which is about 7 or 8 times as expen-
market over the next several years, as the sive as a common notebook power supply.
technology promises both smaller and more However, if they can they bring down the
efficient ac-dc power supplies and dc-dc cost into the same $/W as today’s designs it
converters. will have a transformative effect on the ac-dc
adapter industry. The 65W power adapter
According to GaN Systems, a Canadian operates at 30-100MHz and can be used to
manufacturer of power conversion semi- charge a variety of laptops or other devices
conductors for cleantech applications, GaN such as smartphone or tablets (it comes with
devices offer five key qualities that make a 2.1A USB connector)—it can even charge
it unique: high dielectric strength, high more than one device at a time.
operating temperature, high current density,
high-speed switching and low on-resistance. FINsix’s power adapter is called the Dart and
These traits are due to the properties of is an after-market charger that can work with
GaN, which when compared to silicon, a variety of laptops and other devices. This
offer ten times higher electrical breakdown GaN-based device shrinks power electronics
characteristics, three times the bandgap, and by increasing the frequency at which these
exceptional carrier mobility. devices operate. The higher the frequency,
the smaller the device can be. Ordinarily,
However, the current challenge faced by higher frequencies also reduce efficiency.
GaN technology remains cost, as an ex-
pensive and extensive process is required However, researchers at MIT and FINsix de-
to grow a GaN crystal or wafer on which veloped a way to recycle much of the energy ● Flexible combinations of
transistors and integrated circuits (ICs) can that’s normally lost inside a power adapter, the phase components
be fabricated. A number of companies are improving efficiency and making it practical ● Inverter power rating from
exploring the use of GaN technology in a to use frequencies 1,000 times higher than
290 kW to over 1.4 MW
variety of areas and once the process is those used in conventional power adapters.
implemented on a large enough scale, the Production of the North American Dart (US ● Air or water cooling
cost should eventually come down. prongs) will begin in the fall of 2014. ● Also as boost or buck converter
The adoption of GaN technology is expected Due to the sheer size of the overall external engineered by
to have an effect on the external power ac-dc power supply market it is an attractive
supply market over the next several years. opportunity for digital power supply manufac-
The California-based startup Transphorm turers and is projected to see considerable
has developed power electronics technology advancement over the next several years.
based on gallium nitride, a semiconductor
that wastes far less power than the silicon
Welcome to the House of Competence.
components used in conventional devices.
Not only are these electronics more efficient, www.darnell.com
+49 (0) 621/7 89 92-0®
www.bodospower.com December 2014 Bodo´s Phone
Power Systems
VARIS@gva-leistungselektronik.de
21
www.gva-leistungselektronik.de
TECHNOLOGY
CONTENTDRIVING MARKETS
eGaN® Technology:
Transforming the Future
By Alex Lidow Ph.D.; Efficient Power Conversation
KEY TAKE AWAYS Elon Musk, CEO of SpaceX, has set as his mission to reduce the cost
• Transformative technologies, like the eGaN® technology devel- of putting objects in space by a factor of 10. With eGaN technol-
oped by EPC, change entire industries. ogy applied to satellites we can reduce the size of the electronics,
• eGaN technology has resulted in devices 5-10 times better than eliminate the shielding required, and greatly improve the performance
silicon counterparts, enabling large new applications like envelope of the data communications. This eliminates solar panels, makes the
tracking and wireless power. And, this is just the beginning. eGaN entire system smaller and lighter weight, and extends the life of the
technology can be 1000 times better than silicon – changing the satellite.
game in several industries.
• eGaN technology will evolve from discrete transistors to single- A factor of two reduction in weight is within our reach with today’s
chip integrated circuits to complete systems-on-a-chip, significantly technology, whereas a factor of 10 is possible when eGaN technology
increasing performance and value-added to the market. is used to produce entire systems on a single chip. Multiply the impact
of SpaceX with eGaN technology and we will change the way we use
In Issues 4 and 5 of Technology Driving Markets, we discussed a few space and accelerate the exploration (and possible colonization?) of
applications enabled by eGaN technology: Envelope Tracking, Wire- our universe.
less Power, and LiDAR. Each of these will make a major difference to
the consumer in the next few years.
In this issue we will look out 5 to 10 years and see how a transfor-
mative change in semiconductor technology is transformative to our
world in almost every way.
Expected in 2015, wireless power systems using eGaN technology leveraged to simplify
will begin to unload this wired burden by providing energy wirelessly sensor systems, to
to charge cell phones and tablets. By integrating thin transmission power medical implants
coils in the floor tiles and the walls of buildings and homes, the need and reduce electrical
for wall sockets will be eliminated altogether! This same wireless wiring in day-to-day care
power technology can be used to charge electric vehicles when of the patients.”
parked over a transmitting coil embedded in the floor of a garage.
But it’s not just eGaN
technology in wireless
power that is transform-
ing medicine. Imag-
ing technology is also
improving by leaps and
bounds! The resolu-
tion of MRI machines
is being enhanced
through the develop-
ment of smaller and
more efficient sensing coils using eGaN FETs and ICs. Diagnostic
colonoscopies are about to become a thing of the past due to today’s
eGaN FETs. These types of non-invasive imaging breakthroughs
significantly reduce the cost of health care through early warning and
non-invasive diagnostics. As we integrate entire systems on a single
eGaN chip, miniaturization and image resolution improves the stan-
dard of care while medical costs come down.
A project is underway to embed wireless chargers at bus stops. Even- eGaN Technology — Transforming the Future
tually, in a one-minute stop, a bus can get enough charge to drive In this issue of Fast Just Got Faster, we talked about a few of the
a mile to the next bus stop. This could eliminate the need for most transformations that will be enabled as eGaN technology evolves.
of the heavy batteries and overhead electrical systems that burden EPC is taking the 5-10 times gap in performance between eGaN FETs
electric buses today. and MOSFETs and improving it to a 1000 times gap. This technology
will also apply to integrated circuits made in eGaN technology.
Transforming Medicine
We are all getting older every day, and, as we age, we develop more EPC is pursing parallel paths – discrete power semiconductors and
opportunities for frailties and chronic health problems. Today there are (soon) fully integrated circuits that initially will form building blocks for
major advances in fields such as implantable systems, imaging, and multiple applications, but will ultimately evolve into complete systems-
prosthetics that are enabled by eGaN technology. on-a-chip for very high performance, low cost, and high value-added
applications like the ones discussed above.
Wireless power is already having an impact on implantable systems
such as heart pumps. Beyond just artificial hearts, many other medi- The eGaN journey has just begun!
cal systems can also benefit. As Dr. Pramod Bonde of the University
of Pittsburg Medical Center speculated, “[wireless power] can be www.epc-co.com
Field Charge Extraction (FCE) diode. Then, the new device concepts 170
and their electrical performance will be demonstrated for the 3.3 kV
160
voltage class.
150
IGBT and diode future development trends
140 ET-IGBT
The main three IGBT development trends today are targeting higher
EP-IGBT
power densities with (a) Enhanced Trench ET-IGBTs, (b) higher op- 130
erating temperatures above the traditional 125°C mark and (c) IGBT/ 120
Diode integration solutions referred to as Reverse Conducting RC- 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0
IGBT or Bi-mode Insulated Gate Transistor (BIGT). In the BIGT case, On-state voltage drop (V)
the single chip approach provides improved performance especially
Figure 1: 3.3kV IGBT trade-off curve between on-state voltage drop
with respect to the limitations due to the restriction in available diode Fig. 1. 3.3kV IGBT trade-off curve between on-state voltage drop and turn-off losses.
and turn-off losses. Comparison between the enhanced trench (ET)
area depending on the given application requirements. Nevertheless, Comparison between the enhanced trench (ET) and enhanced planar (EP) structures.
and enhanced planar (EP) structures
the traditional IGBT/Diode two chip approach remains as an important
development path for many mainstream applications. Today, state-of- The ET-IGBT concept
the-art high voltage devices with a similar loss performance employ The main approach followed for the realization of the ET-IGBT con-
Enhanced Planar IGBT (EP-IGBT) or Trench IGBT MOS cell concepts cept with the targeted enhanced carrier concentration near the trench
on Soft Punch Through (SPT) structures. However, for lower voltage emitter for lower losses is based on the introduction of a striped active
devices rated below 2 kV, in addition to trench IGBTs, advanced Trench MOS Cell with an n-enhancement layer. In order to reduce the
ET-IGBTs are already an established technology. Furthermore, the effective input capacitance for improved switching controllability, the
ET-IGBT concept is also capable of providing the next step in loss focus is on the optimization of the regions between the active cells,
reduction for high voltage IGBTs. Figure 1 demonstrates the on-state which contribute strongly to the device effective input capacitance
Vce(sat) loss reduction of a 3300 V IGBT for the same turn-off losses value during switching. By eliminating gate regions between the
Eoff achieved with the new ET-IGBT MOS cell on the same bulk SPT active cells as shown in the cross section in figure 2, we allow for a
platform. However, it is important to point out that the trench based low effective gate emitter input capacitance compared to state-of-the-
IGBTs, especially for higher voltage ratings, exhibit an inherently high N-source
effective gate input capacitance when compared to planar based
devices which results in less controllability for optimum switching
P
performance during IGBT turn-on. Overcoming this negative aspect n-enhancement
gate
gate
combined with the lower losses of the ET-IGBT will provide an ideal layer
solution for the next generation high voltage IGBTs.
For the fast diode part, the losses and reverse recovery softness
remain as a critical performance target for matching the performance
Figure 2: ET-IGBT MOS Cell concept
of the new ET-IGBTs. The Field Charge Extraction (FCE) concept Fig. 2: ET-IGBT MOS Cell concept.
art trench IGBT designs while providing optimum reverse blocking 3.3 kV ET-IGBT module prototypes
capability. The lower on-state losses of the 3.3 kV ET-IGBT provides 3.3 kV ET-IGBT and FCE-diode chips were manufactured with an ac-
potentially a 20% increase in the rated current capability compared to tive area of approximately 1 cm2 per chip with a defined rating of 75 A
the EP-IGBT generation. for the IGBT and 150 A for the diode. The chips were employed in a
standard high voltage insulated module (140 x 70) mm2 having a dual
configuration as shown in the inset of figure 4. Each IGBT/diode part
in the dual package consists of a single substrate containing 4 x ET-
IGBTs and 2 x FCE-diodes. The resulting current rating of the module
is 300 A compared to today’s 250 A for an equivalent EP-IGBT.
Nominal
T = 25°C
T = 150°C
ET-IGBT
200
Current (A)
EP-IGBT
100
0
0 1 2 3 4
Voltage (V)
Figure 4: 300A/3.3kV ET-IGBT module output I-V characteristics at
25°C and 150°C.
Fig. 4: 300A/3.3kV ET-IGBT module output I-V characteristics at 25 C and 150 C.
The modules were tested electrically under static and dynamic condi-
tions. Figure 4 shows the on-state characteristics for the ET-IGBT at
Figure 3: The combination of FCE and FSA concepts (a) and (b) 25°C and 150°C and compared to the EP-IGBT. The ET-IGBT module
cross sections, (c) doping profile exhibits much lower static losses compared to the EP-IGBT together
combination of FCE and FSA concepts (a) and (b) cross sections, with(c) doping
strong profile.
positive temperature coefficient for safe paralleling of
The FCE diode concept chips. At the rated current of 300 A, the ET-IGBT design has a Vce(sat)
For the new diode, a combination of the Field Charge Extraction of 2.75 V compared to 3.55 V for the EP-IGBT at 150°C.
(FCE) concept and the well-established Field Shielded Anode (FSA)
design is utilized as shown in figure 3 when compared to a conven- Figure 5 show the nominal turn-off and turn-on switching waveforms
tional design. The thickness of the n-base plays a key-role for the for both ET-IGBT and the reference EP-IGBT, respectively. The test
overall loss generation where low-loss diodes require a thin n-base conditions were kept the same to better evaluate the device perfor-
design. However, further reductions of the thickness of the n-base mance. The devices were switched against an applied DC-link voltage
region have been typically restricted by the snappy reverse recovery of 1800 V and a rated current of 300 A at 150°C with a gate emitter
behavior of the resulting diodes. By introducing small p-doped areas capacitance of 47 nF. The stray inductance was 600 nH and the turn-
at the cathode side of the diodes as shown in figure 3, a field-induced off gate resistance was 9Ω while the turn-on gate resistance varied
carrier injection process is enabled during the recovery phase, which per design as indicated. The turn-off losses Eoff of the ET-IGBT were
generates inherently soft diodes. Therefore, the n-base of a 3.3 kV at around 650mJ compared to 600 mJ for the EP-IGBT. However,
rated diode can be thinned by 10% while the blocking capability is larger variations were obtained for the turn-on losses Eon where the
maintained by increasing the
3000 3000 600
resistivity without compromis- 75
EP-IGBT
75 EP-IGBT (R =10) G
Current
ing soft reverse recovery. The 300
ET-IGBT
2500
ET-IGBT (R =18) G
60 60 2500 500
benefit of this approach is
a 20% improvement on the 45 Voltage 2000 45 2000 400
Voltage
Gate voltage (V)
Gate voltage (V)
Current
technology curve. Moreover,
Voltage (V)
Current (A)
30 200 1500 30
Current (A)
Voltage (V)
Figure 5: 300A/3.3kV module Turn-off (left) and Turn-on (right) waveforms (1800V, 300A, 150°C).
Fig. 5: 300A/3.3kV module Turn-off (left) and Turn-on (right) waveforms (1800V, 300A, 150 C).
ET-IGBT was at 860 mJ albeit with a different gate resistor compared ET-IGBT Turn-off and Short Circuit SOA Performance
to 910 mJ for the EP-IGBT. The total switching losses for all tested The turn-off (RBSOA) behavior was tested for two paralleled chips
devices were approximately at the same level just below 1.5 J. The under high current and voltage conditions. For the RBSOA, the ET-
FCE-diode reverse recovery performance can also be seen in the IGBT was tested against a high DC-link voltage of 2500 V and the
IGBT turn-on waveforms. The controllability of the ET-IGBT is illus- maximum achieved switchable current is approximately 5x and 4x the
trated in figure 6 when plotting the turn-on parameters (Icmax, Eon and nominal current at 25°C and 125°C, respectively as shown in figure 8.
di/dt) against the variation in the gate resistance. The device enters and withstands both stress conditions known as dy-
namic avalanche and Switching Self Clamping Mode SSCM at 25°C.
600
At a temperature of 125°C, the device experiences stronger dynamic
550 avalanche as expected due to the higher levels of carrier concentra-
tions which results in a lower but still sufficient turn-off capability.
IC max (A)
500
2040 Standard diode
450 FCE diode
Voltage (V)
1360
Average di/dt (A/s) Turn-on losses (mJ)
1500
EP-IGBT 680
1200 ET-IGBT
0
20
900 Current (A) 0
-20
600
-40
-60
600
-80
25°C
900 3600
150°C 600
2400
750 3000
voltage 450
Voltage (V)
Current (A)
Voltage (V)
4×Inom
300
450 1800
current
1200
300 1200 150
0 0
0 -150
0 1 2 3 4 5 6 0 2 4 6 8 10 12 14 16 18
During turn-off and turn-on events, the body capacitor of each FET
is charged and discharged, respectively. For circuits where the FETs
(not the reverse diode) experience hard-switching during turn-on, the
charge stored in the body capacitor is dissipated through the FET,
adding to the switching losses. These losses can be captured in
Figure 1: LLC resonant converter simulation (without adding a capacitor parallel to the FET) by simply
including them in the look-up tables for the losses. When measuring such as saturation. In the LLC resonant converter presented here, the
the switching energy losses of FETs for certain operation points to high-frequency transformer model is composed of two E-shape core
generate the loss look-up table, it is important to include the charge components with an air-gap on the middle leg, and three windings, as
stored in the FET body capacitor in the loss measurements. A major shown in Figure 3. The geometric parameters of the iron core in this
advantage of ZVS for FETs, in addition to eliminating turn-on losses, example are per the “E70/33/32” device specification from EPCOS,
is that the energy stored in the body capacitance is recycled into the while material parameters regarding the saturation effect come from
circuit [3], assuming there is enough blanking time between turn-off of the B-H curve in the ferrite N87 datasheet. The winding turns values
Q1/Q4 and turn-on of Q2/Q3. Therefore, in soft-switching topologies are chosen to achieve the expected magnetization inductance.
operated with ZVS, turn-off losses can be overestimated by 10 to 20
% [4], if the effects of the capacitive elements are not reflected in the Simulation results
loss look-up tables. A 200 VDC input is connected to the DC-side of the full-bridge. A slew
rate is applied to the converter to limit the rate of change in output
Variable-frequency modulation voltage during a change in the reference voltage. Additionally, a soft-
Frequently, to generate PWM signals in simulation, a duty cycle start algorithm is employed to limit the rate of change in commanded
value is compared against a fixed-frequency triangular waveform. voltage during startup. This is reflected on the startup transient as the
Such simplified models have inherent limitations in comparison to the output voltage incrementally ramps up. After the soft-start, a voltage
functionality provided by real MCU peripheral modules. As a result, controller is used to control the switching frequency of the full-bridge.
the system model fidelity is substantially reduced and effects critical The system reaches the desired 300 VDC output after 20 ms. The
to the power controls may be lost or inaccurately simulated. In [1] we thermal system has a smaller time constant, reaching steady state
discussed the efficient modeling of high-fidelity peripheral modules. A after 40 ms. Figure 4 shows the transient simulation of the LLC con-
characteristic of these models is the ability to operate the high-fidelity verter from startup.
PWM peripheral module with variable-frequency. The frequency and
duty cycle are provided as inputs to this peripheral module. These
inputs are sampled to update the PWM signal for the next switching
period. The user can dynamically update the frequency and duty cycle
in simulation, and therefore efficiently implement a variable-frequency
modulator, as is needed for an LLC converter.
Figure 5 shows the operating points of the core on the B-H curve
during a transient simulation to steady state. Engineers can size their
Figure 3: High-frequency transformer model using saturable magnetic magnetic component using the saturable core component to operate
elements in PLECS the system at the desired point of the B-H curve.
Conclusion
The modeling and simulation of a resonant LLC converter has been
presented in this article. PLECS permits analysis of transient effects
from multiple physical domains in a single system model without
excessive simulation times. This provides an effective and accurate
means to investigate and address issues related to compact power
supply design. Further, high-fidelity peripheral models allow design-
ers to simulate the LLC under variable-frequency operation. Such
comprehensive models provide power electronic designers with more
insight into the system before hardware is built, leading to time and
cost savings.
www.plexim.com
Reference
[1] Prausse, F.; Ahmed, M., “Efficient Microcontroller Peripheral Modeling with
PLECS”, Bodo’s Power Systems, July 2014 Issue, pp. 34-37
[2] Munk-Nielsen, S.; Tutelea, L.N.; Jaeger, U., “Simulation with Ideal Switch
Models Combined with Measured Loss Data Provides a Good Estimate STANDARD OR SPECIAL – YOUR CHOICE
of Power Loss”, IEEE Industrial Application Conference, vol. 5, 2000, pp.
2915-2922
[3] Erickson, R.; Maksimovic, D., “Fundamentals of Power Electronics”, 2nd
Edition, Springer, 2001, Chapter 19
[4] “A More Realistic Characterization of Power MOSFET Output Capacitance
Coss”, International Rectifier Application Note, AN-1001
[5] Allmeling, J.; Hammer, W.; Schönberger, J., “Transient simulation of ELECTRONICON Kondensatoren GmbH · Keplerstrasse 2 · Germany - 07549 Gera
magnetic circuits using the permeance-capacitance analogy”, IEEE 13th Fon: +49 365 7346 100 · email: sales@electronicon.com · web: www.electronicon.com
Workshop on Control and Modeling for Power Electronics (COMPEL),
2012, pp.1-6
1.5 TO-247
0.67 mJ
=
TO-247 4pin
1
100% Eon
C
VDRV
0.92 mJ Collector
103%
0.5 = 98%
100% Eoff 83%
30% PWM RG G Gate
20% 18%
0 Emitter
TRENCHSTOP™ / EC HighSpeed 3 / EC TRENCHSTOP™ 5 / TRENCHSTOP™ 5
Turn-On Losses
Diode Diode Rapid 1 Diode /Rapid 1 Diode
IG
Turn-Off Losses 2006 2010 2012 2014
On the other hand, turn-on energy has practically not changed. One Figure 2: IGBT connection in Kelvin emitter configuration
of the main reasons for this is the fact that the turn-on behavior of the
IKZ50N65H5 - Etot
IGBT is strongly depending on the counterpart diode and its amount 10000
of reverse recovery charge. Actually, the amount of recovery charge Comp. F
8000
tends to increase when the diode is combined with a faster IGBT, thus 3pin
Etot [µJ]
6000 4pin
increasing the switch’s turn-on losses. 3pin
Configuration 4000
2000
To achieve a considerable reduction of turn-on losses, the TO-247
0
4pin package is now being introduced for devices of the TRENCH- E1
E2
0 25 50 75 100
STOP 5 family. This package contains an extra emitter pin to be IC(sw) [A]
connected exclusively to the control loop and has already been Figure 3: Total switching energy of an IKZ50N65EH5 in both, 3pin and
successfully used in combination with super junction MOSFETs from 4pin configurations
nect the gate and emitter pads to the gate The total switching losses measured in the
resistor and the gate driver. During transitory three tests is presented in Figure 3. The
conditions like the turn-on or turn-off of the advantage of the Kelvin emitter configuration
IGBT, the parasitic inductances will induce is more prominent for the highest currents.
Analyze Your
voltage drops that will make the effective These are the conditions where highest cur- Passive Components!
voltage over the device pins different from rent change rates are expected. Therefore,
the effective voltage over the die pads. the lead inductance will attenuate the gate
voltage the most in the 3pin configuration.
The voltage attenuation given by the emitter
lead inductance Le causes a reduction of the Figure 4 presents the difference in switching
effective gate to emitter voltage. It can be energy between the 3pin and 4pin configura-
quantified as the value of Le multiplied by the tions as a function of the switching current.
change rate di/dt of the emitter current. Due The benefit of the Kelvin emitter therefore
to this attenuation, the commutation time is can be quantified in absolute values so as in
extended, leading to higher switching losses. percentage to the standard 3pin configura-
tion.
The newly introduced TO-247 4 package has From Figure 4, it is possible to deduce that in
an extra connection to the IGBT’s emitter, applications where the current is higher than
labeled E2 in Figure 2. This point shall be the rated current of the IGBT, the switching
connected to the gate driver. Also known as loss reduction can be higher than 20%. This
Kelvin emitter terminal, this pin is not subject often is the case in Uninterruptable Power
to the attenuation coming from the power Supplies (UPS) and welding machines. For
loop. The current coming from the IGBT’s applications where the current is typically
collector is solely conducted by the power around half the IGBT’s current rating, like
emitter lead E1. for example photovoltaic inverters or switch
mode power supplies (SMPS), the benefit is
Another difference from the TO-247 4 pack- slightly lower, i.e up to 15% lower switching Easily measure impedance using
age is the pin-out, which is different from the losses. the Vector Network Analyzer
standard TO-247 3. Bode 100 in combination with
IC(sw) [A] Absolut
This is done to keep Relative impedance test fixtures for SMD
the creepage distance 5A 10A 20A 30A 50A 70A 100A and THT components.
between the high 0 0%
TR1:°Mag(Impedance)
dard TO-247 package. To quantify the benefit refer to the dedicated application note [2]. 100
used as the device under test (DUT). It is a [1] Infineon Technologies AG: AN 2013-05, 0
50A rated IGBT from the TRENCHSTOP™ 5 “CoolMOS™ C7 650 V Switch in a Kelvin -50
Distributed power architectures now dominate in the design of power The larger the differential between the input and output voltage(s) of
systems for high performance datacom and telecom networks, and a DC-DC converter, the less efficient the conversion process. It soon
data center equipment. The change from centralized power - a simple became clear that for maximum system efficiency it is desirable to
AC-DC power supply, perhaps with battery backup, feeding the cards handle the last conversion very close to the load i.e. the processor,
in a system rack - came about by necessity. Semiconductor operating FPGA or other device. Power supplies used in this final conversion
voltages decreased as ever-smaller process nodes were developed are point of load (POL) converters. In addition to boosting efficiency,
to boost the processing power of ICs. The ICs became more powerful placing power converters as close as possible to their loads prevents
and their lower operating voltages meant that higher current was de- instability due to stray impedances in long PCB tracks or system
manded. Long printed circuit board tracks would mean unacceptable wiring. Datel, which was acquired by Murata as part of the Power
I2R losses, reducing power system efficiency, so distributed power Electronics Division of C&D Technologies in 2007, was an early
architecture became the norm. This trend is continuing. Since the pioneer of isolated DC-DC converters and POL modules during the
1980s the power demands of data centers have risen from 300W to 1980s and 1990s.
1200W per board and some forecast that it will reach 5kW by 2015 as
network IP traffic grows dramatically over the next few years. Distributed power architectures can be implemented in a number
of different ways, using regulated or unregulated bus voltages. As
The economics of creating distributed power systems were trans- systems became more complex, demanding a number of different
formed back in 1984 when a group within Sweden’s part of Ericsson voltages, perhaps 12, 5, 3.3, 2.5 and 1.2V, power system design-
AB, Ericsson Components – RIFA Power (later to become Ericsson ers began to adopt intermediate bus architectures (IBA) around 15
Power Modules) launched the PKA series of DC-DC converters. The years ago. Here, the AC-DC power supply feeds an IBA converter at
introduction of the PKA, described by Ericsson as ‘the world’s first perhaps 24V or -48V. The isolated IBA converter outputs 5 to 14V DC
high frequency DC-DC switching power supply’, meant that engineers and feeds the required number of POL converters.
no longer had to design relatively expensive and complex circuits
from discrete components. They could now use a compact board-
mounted module on each card, adding just a few external filtering and
decoupling components, to create much more efficient and effective
power systems. The use of power modules also delivered marked
improvements in system reliability. This was particularly important
for those designing communications networks, where the target for
operating life was sometimes 25 years or more.
Typically, the front-end AC-DC unit would have a -48V output and a
DC-DC power module on each card would convert this to 12V or 5V,
or a combination of both, to provide the correct operating voltage for
semiconductors in the system. 3.3V semiconductors then started to Figure 2: The proliferation of supply rails at the board level has re-
appear and today we’re down to 0.9V for many with some processors sulted in an intermediate bus architecture (IBA) that requires multiple
drawing up to 90A current at full load. POL converters on the system board
Distributed goes digital More advanced than DVS, adaptive voltage scaling (AVS), adopts
The demand to provide sophisticated power management functions, a closed loop, real-time approach that adapts the supply voltage
including the sequencing of power supplies with controlled ramp rates precisely to the minimum required by the processor, depending on its
as required by large FPGAs, along with a desire to reduce board clock speed and workload. This technique also compensates auto-
space and the number of external components are the drivers that matically for process and temperature variations in the processor.
have led power system designers to increasingly turn to digital power
over the past decade. Key to this are the power management ICs Most switching power supplies use a closed control loop with negative
from companies like Texas Instruments, who introduced the industry’s feedback from output to input. A compensation network is needed
first digital signal processing (DSP) development kit specifically for to adjust the frequency response of the loop to achieve the optimum
power supply applications back in December 2002. transient response without compromising stability. The design of the
compensation network can be a time-consuming task involving con-
However it has been the development of off-the-shelf digital power siderable trial and error. Even then, the performance of the compo-
converter modules that has accelerated the adoption of digital power nents in the network can change with variations in temperature or due
in the last five years. The first of these came to the market in 2008 to ageing. In 2010, CUI Inc, a North American company, was the first
in the form of Ericsson’s BMR453 intermediate bus converter. Digital manufacturer to develop and launch non-isolated POL converter mod-
converters have much in common with their analog counterparts, ules featuring auto-compensation, a digital function that completely
including similar power switches and output filters. However, the inner eliminates this problem.
control loop provides digital flexibility for tailoring power delivery to
the application and enabling the power systems to dynamically adapt Using digital power modules also simplifies or enables many other
to changes in operating conditions in real time. Communications, aspects of power system design including active current sharing, volt-
monitoring and control are implemented over the industry-standard age sequencing and tracking, soft start and stop, and synchronization.
PMBus.
The drive for standardization
Digital control is particularly important in improving the efficiency of As DC-DC power modules have become more widely adopted, there
data network power systems. The power drawn by networking equip- has been a drive towards some level of standardization of products
ment increases with data throughput. At times of low data traffic, the from different manufacturers. Customers, concerned about supply
network is operating well below capacity, power supplies are operat- chain reliability, demanded second sources for products, leading to
ing well below their maximum load and processors can run at lower trade associations being formed by power supply and component
clock speeds. vendors to address this issue.
At low loads the power supplies are relatively inefficient, resulting Typically however, the alliances formed in this space succeeded in
in excessive energy consumption and waste heat generation, with little more than agreeing to standard footprints and pin-outs for certain
undesirable technical, financial and environmental consequences. By categories of power converter, such as non-isolated and isolated
implementing a digital control loop encompassing both intermediate DC-DC converter modules. While this has enabled a degree of inter-
bus and POL converters, the intermediate bus voltage can be varied changeability between products from different manufacturers, there
dynamically in response to varying loads. The input voltage to the has not been full consensus with respect to how all of the electrical
POL converters is reduced under low load conditions, which increas- functions of converters are implemented, making it less than straight-
es conversion efficiency at low loads. forward to swap out one product for another. This is particularly true
for digital power, which adds another layer of complexity to the chal-
lenge of ensuring compatibility between solutions.
Not all has been ‘plain sailing’ and despite good intentions and a num-
ber of positive moves, there were issues that stalled developments
by other manufacturers. Most notable was the patent infringement
lawsuit issued by Power One in 2005 to protect the Z-Bus technol-
ogy used in its POL regulator ICS for monitoring and controlling
Figure 3: Digital power is particularly effective for improving efficiency power supplies. This held back the widespread adoption of PMBus
under low-load conditions SOURCE - Ericsson (EGG October 2014) for around four years, until the use of licensing royalty agreements
became routine for such technology.
Some digital power devices offer a dynamic voltage scaling (DVS)
function to save energy. If the demand for computing power is low, Most recently power supply companies have been coming together
both the clock frequency of the processor and its supply voltage can once again to address these issues, particularly the more challenging
be reduced. DVS is usually implemented as an open-loop function requirements of digital power. In July 2011, CUI announced a coop-
with a look-up table holding pre-determined combinations of frequen- erative agreement with Ericsson Power Modules and in September
cy and supply voltage. that year demonstrated a new family of POL modules that are pin and
function compatible with Ericsson’s BMR46X series of converters. A To this end, CUI, Ericsson Power Modules and Murata, founded the
year later, CUI entered into a license agreement with Ericsson for the Architects of Modern Power (AMP) Consortium in October 2014. The
latter’s 3E Advanced Bus Converter family, allowing it to offer custom- goals of the Group go far beyond the ambitions or achievements of
ers an intelligent intermediate bus converter to pair with its portfolio established trade associations in the power industry. AMP Group will
of POL products. The companies agreed at that time to cooperate be characterized by deep collaboration between the firms in develop-
on a common standard for digital intermediate bus converters going ing leading-edge digital power technology, in terms of both functional-
forward. In July 2014, Murata and Ericsson announced a technical ity and efficiency. Common standards will encompass mechanical,
collaboration agreement with the goal of accelerating the adoption of electrical, communications, monitoring and control specifications.
digital power products by offering customers fully compatible products Members will focus on developing products that deliver high efficiency
from each company. power conversion under all operating conditions and provide supply
chain security to customers through true plug-and-play compatibility
Meeting future needs between their products.
According to the Ericsson Mobility Report, annual IP traffic will reach
7.7 zettabytes by the end of 2017, up from 2.6 zettabytes in 2012.
Video communications, cloud-based services and the interconnection www.ampgroup.com
of physical objects, dubbed the Internet of Things (IoT), are the pri-
mary drivers of this unstoppable growth. This will place even greater www.cui.com
demands on data network power system designers and fully exploit-
ing the functional and efficiency benefits provided by digital power www.ericsson.com
devices is the only way that they will be able to rise to the challenge.
Furthermore, the technology that started in datacoms and telecoms www.power.murata.com
is proliferating into other industries and applications as advanced
processors and FPGAs become commonplace, such as in medical,
industrial, and test/measurement equipment. This is creating the need
for a simple, intuitive, multisource solution across the board. The chal-
lenge is to achieve “perfect power conversion, under all conditions, all
of the time”. Hence there is pressure on power supply manufacturers
to accelerate their rate of innovation to the extent that much deeper
collaboration will be needed.
PROGRAMMABLE DC POWER
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.
www.magna-power.com
No. of Models 54 70 80 80 65
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
Current 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 0-7.2 Adc to 0-4,500 Adc 0-24 Adc to 0-24,000 Adc
6.7mm2. The TPS82740 is according to TI the industry’s first step for controlled start up of the power stage which minimizes the stress
down converter module featuring typically 360nA quiescent current on the power-stage components.
consumption. This new DCS-Control based device extends the light
load efficiency range below 10µA load currents. It supports output The circuit integrates a 700-V MOSFET, start-up current source and
currents up to 200mA. The device operates from rechargeable Li-Ion internal current sensing in a tiny 7-pin SO package size of 29.4mm2.
batteries, Li-primary battery chemistries such as Li-SOCl2, Li-MnO2 Additionally, no external compensation is needed which further
and two or three cell alkaline batteries. The input voltage range up reduces component count and board space. High-voltage pins are
to 5.5V also allows operation from an USB port and thin-film solar segregated to one side of the package, which maximizes separation
modules. The output voltage is user selectable by three voltage select between low and high voltage pins. Along with current limit func-
pins (VSEL), within a range from 1.8V to 2.5V (TPS82740A) and tion, the controller inductor current runaway protection helps protect
2.6V to 3.3V (TPS82740B) in 100mV steps. The TPS82740 features against under load short-circuit conditions to ensure design reliability.
low output voltage ripple and low noise. Once the battery voltage The device provides over-temperature protection with hysteretic
comes close to the output voltage (close to 100% duty cycle), the re-start for safe operation. The high voltage switcher reduces system
device enters no ripple 100% mode operation preventing an increase cost and minimizes the overall size of the power supply, while main-
of output voltage ripple. In this case the device stops switching and taining high efficiency and system performance.
the output is connected to the input voltage. The integrated slew rate
controlled load switch with a typical ON-resistance of 0.6Ù distributes An offline AC/DC reference design PMP8550 based on the
the selected output voltage to a temporarily used sub-system. UCC28880 enables designers to quickly design a low-cost, low-power
non-isolated high-side buck converter with a total solution size of
Energy savings to “always-on” smart meters 38mm by 32mm by 22mm that supports13V, while generating up to
Hundreds of millions of smart meters around the world continu- 100mA of output current. The reference design features a buck con-
ously draw power from the grid to measure energy use and provide verter with integrated switch and can be used in numerous industrial
feedback to the utility. Even though meters consume relatively applications. The design’s schematic, CAD files and test results are
small amounts of power that is not charged to the customer, when available for download. The reference design uses the UCC28880 low
you multiply the amount of energy needed to power the devices, it power, low cost, off-line buck converter with integrated 700V MOSFET
makes sense to minimize the energy consumed. Therefore Texas to generate a non-isolated 13V/120mA output from a universal AC line
Instruments is expanding its portfolio of high-voltage power solu- input. Thanks to its hysteretic mode, no loop compensation is needed
tions for offline AC/DC designs and introduced a 700V switcher with reducing BOM count to 23 components producing a small single
the industry’s lowest quiescent current of less than 100µA – half the sided, double layer board which can be used in numerous industrial
power consumption of existing solutions. The UCC28880 controller applications such as smart meters, home automation equipment and
integrates a 700V power MOSFET and high-voltage current source, white goods.
increasing overall energy efficiency of “always-on” non-isolated power
systems with output currents up to 100mA, such as smart meters, New family of 36-V and 60-V converters
home automation equipment and white goods. The controller inte- Also introduced were seven Simple Switcher regulators that simplify
grates the controller and a 700-V power MOSFET into one monolithic wide VIN synchronous power supply design and help engineers
device. The device also integrates a high-voltage current source, en- create energy-efficient, electromagnetic interference (EMI) compliant
abling start up and operation directly from the rectified mains voltage. products. The easy-to-use LM43600/1/2/3 and LM46000/1/2 DC/DC
The low quiescent current of the device enables excellent efficiency. converters feature an input voltage range of up to 60 V for high reli-
With the UCC28880 the most common converter topologies, such as ability in rugged systems, and 27mA of standby current that minimizes
buck, buck-boost and flyback can be built using a minimum number of power consumption at light load. The unique synchronous power
external components. The UCC28880 incorporates a soft-start feature stage architecture reduces radiated emissions for EMC compliance
Figure 2: The UCC28880 high-voltage switcher delivers energy sav- Figure 3: Simple Switcher DC/DC regulators enable quickly design
ings to “always-on” smart meters and home automation designs energy-efficient, EMI-compliant systems
InDUR
v
by STS
InDUR
helping designers generate, optimize and simulate a wide VIN design,
and then export that design to a CAD program.
www.ti.com
� How?
Just scan this code!
www.sts-trafo.de
Assistance with your own designs in any design phase. Design approvals, failure analyses,
Redesigns to weed out problems or to reduce cost.
Seminars, Articles and Books. Translations of technical and other critical texts
German - English, English - German, French - German, French - English.
Former manager of R & D / managing director in D, USA, NL, A. DR.-ING. ARTUR SEIBT
Consultant and owner of an electronics design lab since 23 yrs. Lagergasse 2/6 Tel.: +43-1-5058186
140 publications resp. patent applications, inventor of A1030 Wien (Vienna) Fax: 5037084
Austria Mobile: +43-699-11835174
the current-mode control in SMPS (US Patent 3,742,371).
email: dr.seibt@aon.at
Names and business affairs of clients are kept strictly confidential.
http:// members.aon.at/aseibt
Active and passive vibration control techniques system behaviour was measured again. The correlation between
Within the first part of this project a proof of concept demonstrator has calculated and measured characteristics is very good. The predicted
been built. This includes the evaluation of different vibration damping “total” compensation was validated with a maximum reduction of the
techniques – a piezoelectric shunt damping technique and an active vertical vibrations of more than two decades (s. Figure 3).
open-loop control. To investigate and set up these techniques a suit-
able model is essential. Therefore the ultrasonic transducer has been For the passive vibration damping technique, a passive inductance-
described by a finite element model, which can be used to calculate resistance (LR)-circuit is connected to the electrodes of the piezo-
the dynamic vibration behaviour of the transducer. The FE-model has ceramics, in simulations as well as on the prototype. This passive
been validated by frequency response measurements. Additionally a damping technique does not require external power and is inherently
modal reduction of the finite element model to the two most relevant stable. For proper operation the network parameters must be tuned
eigenmodes is made. The system is then described by a model with precisely. Measurements of the free vibrations and of the frequency
two degrees of freedom (2-DOF model) which can describe longitudi- response showed a strong damping effect and a decrease of the verti-
nal and bending motions. This model is essential to calculate the op- cal vibrations by a factor of more than 10. However, a total suppres-
timal parameters for both the active vibration control and the passive sion is not possible with this technique.
damping technique. These investigations and the model equations
are reported in detail in [1] and [2]. To study the active vibration control technique, initial experiments
were conducted without wire. In this case, other than in the bonding
process, there is a steady friction contact at the tool tip.
Figure 4 shows that the active control can reduce the basic harmonic
completely in this case. But the investigations also revealed that
because of the nonlinear friction force characteristic higher harmonics
of quite low amplitude remain.
VARIS™ XT –
advantages of the passive shunt damping technique are its simplicity and inherent stability. No
additional hard- or software is necessary. But its lower damping performance and the sensitivity
compact
and
towards parameter deviations are limiting its application.
The very profound advantage of the active feed-forward vibration control strategy is its possibility
powerful
to realize an almost complete suppression of orthogonal vibrations at any desired operating fre-
Figure 6: Full scale bonding tests of the active vibration control method, compensation of vertical
tool vibrations at bonding start (opt. phase 80°, left) and at bonding end (opt. phase 10°, right)
quency. But for this considerable more control effort and a high voltage supply for the control ac-
tuators are necessary. It was found that because of the nonlinear time-varying bonding process
the optimal control settings are not fixed, but a function of the process and thus time-varying.
As depicted in Figure 6, especially the phase angle has to be adapted to gain full compensation
during the entire bonding process. A vertical vibration sensor, e.g. integrated into the control ac-
tuators, can enable the set-up of a closed loop control in order to suppress the vertical vibrations
in any case.
An additional promising observation during the test was a reduction in the standard deviation
of the measured signals. With active vibration control the natural fluctuations of the process
seemed to be decreased. Further studies and tests have to confirm this.
Conclusions
The passive vibration damping system can be incorporated with low effort into a serial machine,
leading to a reduction of vertical vibration of about 50%. With an active vibration damping
system an almost complete compensation of vertical vibration is possible. For this an additional
sensor has to be incorporated and a closed loop control is necessary. For applications with ex-
tended demands this extra effort could be reasonable and such a system could be implemented
in future machines to reach optimal performance.
Literature References
[1] Brökelmann, M.; Neubauer, M.; Schwarzendahl, S. M.; Hesse, H.-J., Wallaschek, J., “Active
and Passive Vibration Control of Ultrasonic Bonding Transducers”, Proceedings of ACTUA- ● Six freely interconnected
TOR 2012, 13th International Conference on New Actuators, Bremen, Germany, 18-20 June IGBT half bridge modules
2012. ● Approx. 2 MW inverter output
[2] Neubauer, M.; Brökelmann, M.; Schwarzendahl, S. M.; Hesse, H.-J., Wallaschek, J., “Optimi-
power
zation of bond transducer vibrations using active and semi-active control”, Proceedings of
SPIE Vol. 8341, Active and Passive Smart Structures and Integrated Systems, 2012. ● Interface board for control,
feedback and sensors
http://www.hesse-mechatronics.com/ engineered by
What he does not want to see is a bright red feature, where the color
red indicates that >99.99% of the ultrasonic pulse has been reflected
as an echo. In a monochromatic acoustic image, these highest-ampli-
tude echoes are bright white; they are often pseudocolored bright red
Figure 1: With doors open, the two IGBT modules are seen.
The scanning transducers are out of sight beneath the modules.
(as here) for greater visibility. This type of echo is characteristic of the the solder lies directly between it and the transducer, but a module
interface between a solid production material and a gas such as air, that has such a problem is very likely already a reject.
or a vacuum. The vast difference in material properties of air and, say,
copper causes the essentially total reflection of ultrasound.
The data that is collected during scanning is analyzed immediately. Figure 3: Made during the same scan as Figure 2, this image shows
The system’s Digital Image Analysis™ (DIA) software will measure small red die attach voids at the largest of the three die.
the total area of the voids in the solder above each ceramic raft. The
accept/reject criteria, defined by the user, may call for a module to be The acoustic image in Figure 3 covers the same area as Figure 2, but
rejected if the total area of voids is, for example, >2% of the area of was gated on the fifth of six depths within the IGBT module. The red
the raft. Criteria may also take into account the x-y location of a defect defects seen in Figure 2 are here seen as black features because the
- a void that is not directly under a die is not as serious as one that is. ultrasound returning from the fifth depth is blocked by these voids.
At the end of the scanning process, accept modules are stacked in Thus the three-pronged void in the center of the left portion of Figure
front of the system, while rejects are stacked at the back. The acous- 2 is black in Figure 3. The depth that Figure 3 is gated on includes
tic data and images may be stored locally, in the system’s computer, the interface between the ceramic raft and the die attach material.
or in the user’s Factory Information System. The small red features within the area of each die are voids in the die
attach.
The inspection process is thus carried out rapidly and with minimal
human intervention. A technician will later examine the images of the One benefit of gating, and especially of setting rather narrow gates, is
reject modules to see which ones are candidates for rework, but no that the tilting of ceramic rafts becomes more visible. If one region of
one needs to examine the images of the good modules. a raft is actually displaying part of the solder layer, part of the die at-
tach, then the raft is tilted and is a potential contributor to overheating.
The simplest inspection process, described above, is not the only
alternative. In some instances, the inspection time can actually be What this new automated system provides is the ability to nondestruc-
reduced by using a recipe to scan only the significant regions of the tively inspect the internal features of IGBT modules at a high through-
module, and to skip the regions that do not have solder, die attach, or put rate according to the user’s definitions of accept and reject. Aside
a die. from startup, it operates without a technician to give the user higher
module quality.
This inspection process described to this point is suitable for inspect-
ing a single layer of interest, such as the solder. But the user of the Contact information:
system may be concerned with defects at more than one level within Sonoscan, Inc.
the module. Typically the critical defects are voids in the solder layer 2149 E. Pratt Blvd.
and in the die attach material. Both can block heat and thus pose the Elk Grove Village IL USA 60007
risk of module failure. Phone: 847 437-6400
Fax: 847 437-1550
Both the solder layer and the die attach layer can be imaged at the E-mail: info@sonoscan.com
same time during scanning, without increasing the scan time. A scan-
ning recipe is used that sets a gate for the solder layer - meaning that
it accepts for imaging of the solder layer only those echoes whose www.sonoscan.com
arrival time indicates that they were reflected by a material interface
between the heat sink and the raft. A second gate is set for the die
attach material, which will use for imaging only those echoes that
originated between the top of the raft and the bottom of the die.
Data Link presented a new lamp designed with their new patented
planar and AC LED technology. The AC LED technology has enabled
that the LCL230R-AC lamp has a minimum of components and no
electronic parts that are susceptible to aging. Additionally this lamp
can endure millions of power cycles and has a very long life. Flush
Mount Ceiling Lamp is low energy alternative to incandescent and
Compact Fluorescent bulb. LCL230R-AC provides a large luminous
flux of at least 1400 lumens, while its consumption is only 16W and
has a real lifespan of over 60,000 hours. The lamp has a very pleas-
ant white light suitable for domestic dwellings and residential market,
with colour temperatures available from warm white 2700K to neutral
white 4000K depending on the application. This lamp has thermal
management due to the large surface cooler that allows the low
LED junction temperature of maximum 65°C. The lamp is thermally
protected, which means that at higher ambient temperatures reduce
the intensity of light and prevent overheating of the LED and shorten
lifespan.
www.murata.com
www.toshiba-components.com
www.richardsonrfpd.com
52 Bodo´s Power Systems® December 2014 www.bodospower.com
NEW
CONTENT
PRODUCTS
✓
electronics applications requiring higher bus
voltages. These devices are targeted for use
in a wide variety of applications including DC
microgrids, Vehicle Fast chargers, server,
telecom and networking power supplies, un-
interruptable power supplies, solar inverters,
Wind power systems, and industrial motor
control systems.
SiC Junction Transistors (SJT) offered by
www.cui.com/PowerExpert
www.recom-lighting.com
www.murata.com
www.plexim.com
in co-operation with
www.husumwind.com
NEW
CONTENT
PRODUCTS
www.irf.com
Strategical Support
Corporate/Product Positioning, Market/Competitive Analysis, PR Programs, Roadmaps,
Media Training, Business Development, Partnerships, Channel Marketing, Online Marketing
Tactical PR
Writing: Press Releases, Feature Articles, Commentaries, Case Studies, White Papers
Organizing: Media Briefings, Road Shows, Product Placements in Reviews and Market Overviews,
Exhibitions, Press Conferences
Monitoring and Research: Speaking Opportunities, Editorial Calendars, Feature Placement,
Media Coverage, Competitive Analysis
Translations: Releases, By-Lined Articles, Websites, etc.
www.information-travels.com
www.microchip.com/PIC16_LF161X-Family-Product-Page-111114a
Expanded
Family of 650V
SiC Diodes
YOU CAN’T COPY
Richardson RFPD, Inc. announced the avail-
ability and full design support capabilities for
five new 650V silicon carbide (SiC) Schottky
EXPERIENCE
diodes from Cree, Inc.
www.richardsonrfpd.com
Innovation by Tradition
www.parker.com/chomerics
www.epcos.com
Further information:
web: e - emc.com
phone: + 49 711 61946 63
email: emv@mesago.com
www.bodospower.com December 2014 Bodo´s Power Systems® 63
NEW
CONTENT
PRODUCTS
Advertising Index
ABB Semiconductor C3 EMV 62 Magna 39
Allegro 15 esar 31 Omicron 35
APEC 37 Fuji 25 PCIM Asia 45
CDE 7 GvA C2+21+45 PCIM Europe 41
CUI 53 Hitachi 11 Plexim 3
Danfoss 57 Husum Wind 57 Semikron 19
Dr. Seibt 43 Infineon 9 STS 43
Dynex 17 IR C4 Texas Instruments 13
ECCE 63 Isabellenhütte 61 VMI 51
electronicon 33 ITPR 58 Würth 23
embedded 29 Lem 5 ZES Zimmer 64
Flicker from steel smelters is compensated with multi-level voltage source SVC
(Static VAR compensation) taking advantage of ABB’s innovative HiPak IGBT
modules, eg the new 3,300 volt 500 ampere dual module.
Demanding high-power applications such as traction inverters, medium voltage
drives, wind turbines, HVDC and FACTS benefit from the high reliability of ABB’s
HiPak modules.
ABB’s family of HiPak modules are available from 1,700 to 6,500 volt as single IGBT,
dual / phase-leg IGBT, chopper and dual diodes. All modules feature low losses
combined with soft-switching performance and record-breaking Safe Operating Area
(SOA).
For more information please contact us or visit our website:
www.abb.com/semiconductors