MARCH 2022 - ENABLING GREATER POWER DENSITY IN CHARGER AND ADAPTER DESIGNS WITH HIGHLY EFFICIENT GAN-BASED CONVERTERS - POWER ELECTRONICS NEWS
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
MARCH
2022
Enabling Greater Power Density in
Charger and Adapter Designs with
Highly Efficient GaN-Based Converters
DRIVE HARD.
DRIVE SAFE.
WE-AGDT Gate Drive transformer
WE are here for you!
Join our free webinars on:
www.we-online.com/webinars
WE-AGDT • Optimized for SiC gate driver supply
The WE-AGDT series from Würth Elektronik allows implementing discrete SiC • Interwinding capacitance down to 6.8 pF
gate driver designs easier than ever before. These standard parts are compact • CMTI over 100 kV / µs
SMT transformers optimized for silicon carbide applications. With extremely • IEC62368-1 / IEC61558-2-16
low interwinding capacitance, the WE-AGDT helps to achieve higher Common • Up to 6 W power
Mode Transient Immunity (CMTI). The series is compliant with safety standards • Unipolar & bipolar output
according to IEC62368-1 / IEC61558-2-16 in addition to AEC-Q200 • Compact & lightweight
qualification. Reference designs are available for each WE-AGDT transformer.
The complete solution is compact and capable of fully automated assembly.
Products available from stock. Samples free of charge.
For further information, please visit: www.we-online.com/agdt
#ReadyForTheFuture
VIEWPOINT
Wide-Bandgap
Semiconductors
Power electronics involves a whole range of critical applications, from electrification to smart grids.
It is a fundamental pillar for the entire industry to meet climate-change demands and involves
increasing energy efficiency, reducing our carbon footprint through new materials, and adopting
new circuit topologies. Physical limitations prevent current silicon technology from achieving the
higher power density, miniaturization, and energy-conversion efficiency that the market needs from
power products to meet growing environmental concerns. Wide-bandgap (WBG) silicon carbide and
gallium nitride materials enable significant efficiency improvements in applications such as traction
inverters for SiC and adapters/chargers for GaN. SiC and GaN technologies have grown enormously
over the past few years, proving to be commercially available energy-saving technologies. These
two WBG semiconductors complement each other to address a wide variety of applications in
which energy efficiency is vital. However, silicon products are expected to co-exist with WBG
products due to their cost-effectiveness in some low-power applications.
In this issue, Infineon Technologies’ Zhong Fang Wang, principal applications engineer, and Matt
Yang, senior staff applications engineer, explain how Infineon’s CoolGaN Integrated Power Stage
technology can be applied in active-clamp flyback, hybrid flyback, and LLC converter topologies.
This way, designing charger and adapter solutions is made quicker and easier, leading to smaller
and lighter products or products that provide more power from a device of the same size to charge
devices faster or charge multiple devices from one adapter.
Other topics are drain-current characteristics of enhancement-mode GaN HEMTs, thermal
management, electric vehicles, battery technology, and wireless charging. A combination of electrical
failure analysis (EFA) and physical failure analysis (PFA) can lead to a deeper understanding of fault
mechanisms and, ultimately, improved performance, reliability, and manufacturing yield. We will
analyze advanced EFA and PFA analytical tools to localize and characterize subtle electrical issues
faster and with greater accuracy in WBG materials.
Yours Sincerely,
Maurizio Di Paolo Emilio
Editor-in-Chief, Power Electronics News
MARCH 2022 | www.powerelectronicsnews.com 3
Contents
VIEWPOINT Keeping Cool and Calm in Tight
Wide-Bandgap Semiconductors 3 Environments — The GaN Way 55
COVER STORY — SEMICONDUCTORS TEST & MEASUREMENT
Enabling Greater Power Density in Test Solutions for EV Inverters 26
Charger and Adapter Designs with
Wide-Bandgap Materials Address EV
Highly Efficient GaN-Based Converters
Power and Efficiency Requirements
6 31
SEMICONDUCTORS ENERGY STORAGE
PowerUP Expo 2021: Wide Bandgap, Battery Technology for Automotive: An
Motor Control, and Energy Trends — Interview with Vicor’s Nicolas Richard
Conference Proceedings 13 36
Increasing Fault-Analysis Accuracy in ELECTRIC VEHICLES
Wide-Bandgap Power Devices 51 Wireless Charging for the Future
DESIGN 42
Drain-Current Characteristics of EV Wireless Dynamic Charging Will
Enhancement-Mode GaN HEMTs Cooperate with the Grid 47
14
An Approach to Thermal Management
of High-Power PCBs 21
4 MARCH 2022 | www.powerelectronicsnews.com
NEWS VIDEO & PODCAST
Testing a Fuel Cell Powered by Liquid Vertical GaN Technology (Podcast)
Hydrogen 60 61
German Electric Vehicles Market How to Solve the Climate Change
Growth Goal 60 Problem (Podcast) 61
Investments in Semiconductors SiC’s Cost-Competitiveness (Podcast)
Frontend Fab 60 61
Power Supplies for Horticultural and From Silicon to Silicon Carbide (SiC)
Commercial Lighting 60 (Podcast) 61
MARCH 2022 | www.powerelectronicsnews.com 5
COVER STORY — SEMICONDUCTORS
Enabling Greater Power
Density in Charger and
Adapter Designs with
Highly Efficient GaN-Based
Converters
By Zhong Fang Wang, principal applications engineer, and Matt Yang, senior
staff applications engineer, both at Infineon Technologies
The most popular power converter topology used in today’s charger and adapter applications
is quasi-resonant (QR) flyback topology, thanks to its simple structure and control, low
bill-of-materials cost, and high efficiency due to valley-switching operation. However, the
frequency-dependent switching loss of the switches and the leakage energy loss of the transformer
limit the maximum switching frequency of the QR flyback converter and thus limit the power
density.
6 MARCH 2022 | www.powerelectronicsnews.com
Cover Story — Semiconductors
The adoption of GaN HEMTs and planar transformers in QR flyback converters helps push the
switching frequency and the power density higher. However, to achieve even higher power density
for ultra-slim charger and adapter designs, soft switching of the switches and recycling of the
transformer leakage energy become indispensable. This inevitably leads to selecting converter
topologies with intrinsic higher efficiency.
This article explains how Infineon’s CoolGaN Integrated Power Stage (IPS) technology can be
applied in active-clamp flyback (ACF), hybrid flyback (HFB), and LLC converter topologies. This way,
designing charger and adapter solutions is made quicker and easier, leading to smaller and lighter
products or products that provide more power from a device of the same size to charge faster or
charge multiple devices from one adapter.
CONVERTER TOPOLOGIES QUALIFIED TO ACHIEVE AN EVEN
HIGHER POWER DENSITY
Some half-bridge topologies such as ACF, HFB, and LLC converters have been proven to be able to
achieve high efficiency, even at a very high switching frequency, due to zero-voltage switching (ZVS)
and zero snubber loss.
ACTIVE-CLAMP FLYBACK (ACF)
Figure 1 shows a typical application example of CoolGaN IPS operated in an ACF converter. In the
ACF topology, the clamp switch provides a path to recover the energy stored in the transformer’s
leakage inductance (Llk) when the main switch turns off and the clamp switch turns on. Cclamp
and Llk resonate together through the clamp switch and the transformer, resulting in energy
transfer to the load. This energy recovery increases the system efficiency compared with the
passive-clamp flyback, in which the energy stored in Llk damps in the traditional RCD clamp circuit.
A well-designed ACF topology operates in soft-switching ZVS condition; therefore, it can run
with a much higher switching frequency
than a QR flyback, which operates in
hard-switching conditions. This helps to
reduce the size of magnetic components,
including the transformer and EMI filters.
The ACF converter consists of a high-side
and a low-side switch, the transformer, a
clamp capacitor (Cclamp), and the output
stage of the rectifier and capacitors.
Figure 2 shows the typical operating
waveforms that briefly explain the ACF
converter’s operation principle.
Figure 1: Application circuit of the ACF converter
MARCH 2022 | www.powerelectronicsnews.com 7
Cover Story — Semiconductors
Figure 2: Operation of the ACF converter
The ACF converter stores energy in the primary-side inductor and the leakage inductor when the
low-side switch is turned on. Afterwards, when the low-side switch is turned off, the energy is transferred
to the output. During the off state of the low-side switch, the energy stored in the leakage inductor
is transferred to output
when the high-side switch
is turned on. In addition,
the ZVS operation of the
switches offers further
efficiency improvement.
This operation ensures the
high-efficiency performance
of the ACF converter.
HYBRID FLYBACK
(HFB)
Figure 3 shows a typical
application example with
CoolGaN IPS operated
in a hybrid flyback (HFB)
topology. Figure 3: Application circuit of an HFB converter
8 MARCH 2022 | www.powerelectronicsnews.com
Cover Story — Semiconductors
The HFB converter consists of a high-side and a low-side switch, the transformer, the resonant tank
(Llk and Cr), and the output stage ofthe rectifier and capacitors. It is anothertopologythat benefits from the
soft-switching operation of the power switches and can achieve high power density and
efficiency. In this topology, the transformer leakage and magnetizing inductance resonate
with the capacitor under the same concept of the LLC converter. The implemented advanced
control scheme with a non-complimentary switching pattern provides a solution that
supports a wide range of AC input and DC output voltage, which is necessary for universal
USB-C PD operation.
Figure 4: Operation of the HFB converter
HFB can achieve full ZVS operation on the primary side and full zero-current switching (ZCS) operation
on the secondary side. Subsequently, the leakage energy is recycled, thereby achieving high efficiency.
The HFB can easily achieve a wide output range with a changing duty cycle. This overcomes the
limitation of the LLC topology in wide-output–range applications. For more information about HFB
converters, see [1]. Figure 4 shows the typical operating waveforms that briefly explain the
working principle of the HFB converter. When the high-side switch is turned on, the HFB
converter stores energy in the primary-side inductor. When the low-side switch is turned on,
this energy is transferred to the output. With proper timing control during the switch transition
of both MOSFETs, HFB runs under ZVS for both devices, ensuring high system efficiency
without additional components. Both benefits, coming from ZVS and the additional efficiency
improvement from ZCS operation in the secondary side, make HFB a cost-competitive solution for
ultra-high-power–density converters, such as USB-PD fast chargers.
MARCH 2022 | www.powerelectronicsnews.com 9Cover Story — Semiconductors
LLC CONVERTER
Figure 5 shows a typical
application example with
CoolGaN IPS operated in half-
bridge LLC topology. The LLC
converter is part of the resonant
converter family, which means
that the regulation is not
achieved through conventional
pulse width modulation
pulse-width–modulation (PWM)
schemes. The LLC converter
achieves regulation through
frequency modulation by running
Figure 5: Application circuit of half-bridge LLC converter
at a 50% duty cycle and a fixed
180˚ phase shift. The half-bridge
LLC converter consists of a
Figure 6: Operation of half-bridge LLC converter
high-side and a low-side switch, the transformer, the resonant tank (Lr and Cr ), and the output stage
of the rectifier and capacitors. Figure 6 shows the typical operating waveforms to briefly explain the
half-bridge LLC converter’s working principle. When the high-side switch is turned on, the half-bridge
10 MARCH 2022 | www.powerelectronicsnews.comCover Story — Semiconductors
LLC converter operates in power delivery (PD) mode. In this switching cycle, the resonant tank is
excited with a positive voltage, so the current resonates in the positive direction. When the low-side
switch is turned on, the resonant tank is excited with negative voltage, so the current resonates in the
negative direction. During the PD operation mode, the difference between the resonant current and
the magnetizing current passes through the transformer and the rectifier to the secondary side, and
power is delivered to the load.
Furthermore, all primary-side MOSFETs turn on resonantly with ZVS, resulting in a full recycling of the
energy contained in the MOSFETs’ parasitic output capacitance. In the meantime, all secondary-side
switches turn off resonantly with ZCS to minimize switching losses that are normally associated with
hard switching. Resonant operation of all switching devices in the LLC converter results in minimized
dynamic loss and thus increased overall efficiency, especially at higher operating frequencies in the
hundreds of kilohertz to megahertz range.
To achieve ZVS of the high-voltage switches, all three topologies utilize a circulating current in the
transformer to discharge the Qoss of the switches. Apparently, higher Qoss needs a higher circulating
current and longer discharge time. The circulating current causes additional transformer loss (both
core and winding losses), while the discharge time significantly contributes to the dead time. The dead
time reduces the effective duty cycle and causes higher RMS current in the circuit, which increases
conduction loss. Therefore, for very high-switching-frequency operation, minimizing dead time is
critical. With the superior figure of merit (FOM) of RDS(on) × Qoss, a GaN HEMT helps reduce the dead time
as well as the circulating current in the circuit. This benefit, combined with low driving loss and zero
reverse recovery, makes GaN HEMTs a perfect match for ACF, HFB, and half-bridge LLC converters.
COOLGaN IPS AND
65-W ACF CONVERTER
EVALUATION BOARD
To further improve system size, Infineon
recently launched CoolGaN IPS, which
combines a 600-V enhancement-mode
CoolGaN switch with dedicated gate
drivers in a thermally enhanced small
QFN package.
To demonstrate the performance of
CoolGaN IPS, a 65-W ACF converter
(Figure 7) designed with CoolGaN IPS
IGI60F1414A1L was developed.
Figure 7: Top view of the 65-W ACF evaluation board featuring
CoolGaN IPS half-bridge
MARCH 2022 | www.powerelectronicsnews.com 11Cover Story — Semiconductors
Measured efficiency curves (Figure 8)
show that it meets the CoC Tier 2 and
DoE Level VI efficiency requirements both
for four-point average efficiency and 10%
load condition efficiency.
SUMMARY
GaN HEMTs have become popular in
today’s high-power–density charger and
adapter applications, as they can offer
high-frequency switching due to their
much-improved FOMs compared with
Figure 8: ACF evaluation board efficiency curve under different
input voltage and load conditions silicon MOSFETs. With its high efficiency
and integrated gate driver in a compact
package, CoolGaN IPS technology is well-positioned to enable charger and adapter designs with
even higher power density, thanks to its perfect application in ACF, HFB, and LLC converters.
To find out more about Infineon’s CoolGaN IPS product portfolio and comprehensive solutions,
make sure to visit our website. Also learn more about our high-frequency CoolGaN IPS
half-bridge 600-V evaluation board featuring IGI60F1414A1L (click on this link to check out the
board page: EVAL_HB_GANIPS_G1)2.
For More Information
▶ 1
Infineon Technologies. “Hybrid-flyback converter design with XDP™ digital
power XDPS2201.” Application note, March 2021.
▶ 2
Vartanian, R. "CoolGaN™ IPS half-bridge evaluation board with
IGI60F1414A1L," Application Note, Infineon Technologies, April 2021
▶ 3
Bainan, S. “Quick-reference guide to driving CoolGaN™ GIT HEMTs 600 V.”
Application Note, Infineon Technologies, December 2021.
12 MARCH 2022 | www.powerelectronicsnews.comSEMICONDUCTORS
CONFERENCE
PROCEEDINGS
FEBRUARY
2022
PowerUP Expo 2021: Wide
Bandgap, Motor Control, and
Energy Trends — Conference
Proceedings
WBG semiconductors, smart and renewable
energy, and motion control were significant
topics covered during the PowerUP Virtual
Conference.
DOWNLOAD >
MARCH 2022 | www.powerelectronicsnews.com 13DESIGN
Drain-Current
Characteristics of
Enhancement-Mode GaN
HEMTs
By Maurizio Di Paolo Emilio, editor-in-chief of Power Electronics News
Two distinct structures have been developed for the enhancement mode of GaN-based
high-electron–mobility transistors (HEMTs). These two modes are the metal-insulator–
semiconductor (MIS) structure, which has a low gate leakage current driven by voltage, and
2
the gate-injection transistor (GIT),3 which has a ridge structure and a high threshold voltage.
Both have some shortcomings as well. MIS has less reliability for gate interference and a low
threshold voltage, whereas GIT has less gate switching speed and a higher gate leakage current.
The original article is available here.
Figure 1 shows the setup used to test these two structures. A single model can be used for both MIS
and GIT structures. GIT is used for developing equivalent circuits using current models, whereas
MIS is used for core drain-current modeling.4 Following that, S-parameter measurements are used
to evaluate circuits for each of these devices.
14 MARCH 2022 | www.powerelectronicsnews.comDesign
DEVICE STRUCTURES
Figure 1a shows the basic structure of an MIS transistor along with an embedded source field
plate (ESFP). Metal-organic chemical-vapor deposition is used to develop a silicon nitride (SiN)
passivation layer on it. A sheet of two-dimensional electron gas (2DEG) with a carrier density of
1.4 × 1,013 cm2, mobility of 1,203 cm2V/s, and sheet resistance of 382 W/square is used. A gate
electrode made of a 500-nm–thick silicon dioxide (SiO2) film on top of MO film, an intermetallic
dielectric, is extended over the passivation
film, decreasing Cgs. The ESFP divides
gate-to-drain electric field in two peaks.
This decreases electron density due to
negative biases and increases density
under gate insulator films in active
biases.
Figure 1b shows the basic structure of a
ridge GIT transistor along with a source
field plate (SFP). Its structure has a
10-nm aluminum gallium nitride (AlGaN)
layer acting as a barrier and a 60-nm
p-GaN layer. The gate surface is protected
with a SiN film 100 nm thick after the
surface is etched using ICP etcher. Drain Figure 1: Simplified test structure
and source electrodes are also formed
from etched SiN films, producing ohmic
electrodes. The source electrode is extended over the gate to the drain side to create SFP. The SFP
divides the gate-to-drain electric field into two peaks, decreasing the strength of the electric field
under the gate edge. The measurement used for these experiments are: Lmask = 0.8 mm, Wmask = 100 mm,
source-to-gate distance = 0.9 mm, gate-
to-drain distances = 3.5 mm, and gate
capacitance (Cox), which can be calculated
using gate oxide film thickness (Tox) and
SiO2’s dielectric constant (εox). As shown
in Figure 2b, it is difficult to calculate
gate channel capacitance (Cch) accurately
using electron density accumulated in the
hole injection of the p-n diode. Hence,
Cch is measured before any parameter
extraction procedure begins.
Figure 2: Enlarged picture of the gate
MARCH 2022 | www.powerelectronicsnews.com 15Design
DRAIN CURRENT EQUATIONS
Gate oxide capacitance of MIS-HEMTs
(1)
Gate-to-channel capacitance of ridge HEMTs
The Schottky contact and p-n junctions for ridge GIT HEMTs are shown in Figure 2b. The channel
area consists of the p-type gate to the 2DEG region with a hole injection from the channel.
Drain-current derivations have an incorporated gate channel capacitance per area (Cch).
Threshold voltage
(2)
(3)
Electron mobility
(4) 5)
(6)
(7)
(8) (9)
Drain-source resistance
(10) (11)
(12)
16 MARCH 2022 | www.powerelectronicsnews.comDesign
EQUIVALENT CIRCUITS
Figure 3 shows that both our models,
the MIS and the ridge HEMT, have the
same macro circuits. The main HEMT
transistor operates as a FET to reduce
the drain electric field, while the sub-
transistor acts as an SFP. Figure 4 shows
AC equivalent circuits with intrinsic small
Figure 3: Circuit for MIS and ridge HEMT model
signal for MIS and ridge HEMT types.
Metal interconnected inductances are
labeled Lg, Ld, and Ls, whereas gate
capacitances are labeled Cgs and Cgd,
which are divided into constant (Cgs0 and
Cgd0) and bias-dependent capacitances
using empirical functions.10 Capacitance
for the drain-to-source is labeled as
Cds. Dispersion resistance is Rdis_T, while
capacitances are labeled as Cdis_T and
Cgdis. The gate-to-source internal
resistance is ri. Gate, drain, and source
resistances are represented by Rg, Rd_T,
and Rs_T, respectively. Gate-to-drain
resistance is represented by Rgd. The
scalable gate capacitances, Cgs_sfp and
Cgd_sfp, are in parallel with Cds, as the ESFP
must be connected between drain and
ground.
The diffusion capacitance (C_diffusion) and
the junction capacitance (C_ junction) for
ridge HEMTs in the gate-injection p-n
diode are shown in Figure 4b.11 Terminals
used here are between the source and Figure 4: Equivalent circuits
the gate. The C_diffusion may even work as
Cdis_T.
EXPERIMENTS AND DISCUSSIONS
A curve tracer with a pulse-measurement mode is used to measure DC for both the transistor
structures adopted. The biases provided for this have a pulse width of 100 ms and a duty cycle of 50%.
MARCH 2022 | www.powerelectronicsnews.com 17Design
Measurements with a multi-gate length and width device were made prior to the experiment in
order to obtain linear and saturation drain current, model parameters for threshold voltages, and
length and width dependencies in the gate channel.
This model has high accuracy and can be used in both MIS and ridge HEMT devices for stimulating
static drain currents in the linear and saturation regions. This is clearly represented in Figures 5 and
6. S-parameter measurements along with small-signal AC characterizations can be used effectively
to evaluate equivalent circuits.1
Figure 5: Measured and simulated Ids-Vgs
Figure 6: Measured and simulated Ids-Vds
18 MARCH 2022 | www.powerelectronicsnews.comDesign
CONCLUSION
This article summarizes the two models of HEMTs: drain-current model MIS and ridge GIT.
Small-signal equivalent circuit models for AC and transient simulations are also made with
measurements and S-parameters. For ridge HEMTs, the gate leakage current model as well as
excess drain current are discussed in detail. Other modified equations for the drain-current model
can be created using the MIS-HEMT model. HSPICE was used in conjunction with the Verilog-A
language to create this model. Our test setup worked well with this model and its parameters, and
it can be applied to power-supply design. Transient and noise-equivalent circuits, as well as model
equations, can be designed to switch power supplies more rapidly.
For More Information
▶ 1
H. Aoki, H. Sakairi, N. Kuroda, A. Yamaguchi, and K. Nakahara. “Drain Current
Characteristics of Enhancement Mode GaN HEMTs,” Graduate School of
Environmental Information, Teikyo Heisei University, Nakano-Ku, Tokyo.
▶ 2
K. Chikamatsu, M. Akutsu, T. Tanaka, S. Takado, K. Sakamoto, N. Ito, and K.
Nakahara. “Embedded Source Field-Plate for Reduced Parasitic Capacitance
of AlN/GaN MIS-HEMTs on Si Substrate,” SSDM2015 Conf. Dig., pp. 122–123,
Sept. 2015.
▶ 3
Y. Uemoto, M. Hikita, H. Ueno, H. Matsuo, H. Ishida, M. Yanagihara, T. Ueda, T.
Tanaka, and D. Ueda. “GIT-A Normally Off AlGaN/GaN Power Transistor Using
Conductivity Modulation,” IEEE Trans. Electron Devices, Vol. 54, pp. 3,393–
3,399, Dec. 2007.
▶ 4
H. Aoki, H. Sakairi, N. Kuroda, Y. Nakamura, K. Chikamatsu, and K. Nakahara.
“A Scalable Drain Current Model of AlN/GaN MIS-HEMTs with Embedded Source
Field-Plate Structures,” IEEE APEC 2018, Dig. pp. 2,842–2,847, March 2018.
MARCH 2022 | www.powerelectronicsnews.com 19Design
▶ 5
D. Schroeder. “Modelling of Interface Carrier Transport for Device
Simulation,” 1st ed., Springer, 1994.
▶ 6
BSIM: www-device.eecs.berkeley.edu/bsim.
▶ 7
U. Radhakrishna, L. Wei, D. S. Lee, T. Palacios, and D. Antoniadis. “Physics-
based GaN HEMT Transport and Charge Model: Experimental Verification and
Performance Projection,” IEEE IEDM, Dig., pp. 13.6.1–4, Dec. 2012.
▶ 8
R. Rodriguez, B. Gonzalez, J. García, and G. Toulon, F. Morancho, and A.
Nunez. “DC Gate Leakage Current Model Accounting for Trapping Effects in
AlGaN/GaN HEMTs,” Electronics, 7, 210, 2018.
▶ 9
A. M. Cowley and S. M. Sze. “Surface States and Barrier Height of Metal-
Semiconductor Systems,” Journal of Applied Physics, Vol. 36, Issue 10, pp.
3,212–3,220, 1965.
▶ 10
H. Aoki, H. Sakairi, N. Kuroda, Y. Nakamura, K. Chikamatsu, and K. Nakahara.
“A Small Signal AC Model Using Scalable Drain Current Equations of AlGaN/
GaN MIS Enhancement HEMT,” IEEE RFIC2018, pp. 80–83, June 10–12, 2018.
▶ 11
Y. C. Fong and K. W. E. Cheng. “Experimental study on the electrical
characteristic of a GaN hybrid drain-embedded gate injection transistor (HD-
GIT),” 2017 IEEE PESA, 2017.
20 MARCH 2022 | www.powerelectronicsnews.comDESIGN
An Approach to
Thermal Management
of High-Power PCBs
By Stefano Lovati, technical writer for EEWeb
The entire power-electronic sector, including RF applications and systems involving high-speed
signals, is evolving toward solutions that offer increasingly complex functionalities in ever-smaller
spaces. Designers face increasingly demanding challenges to meet system size, weight, and
power requirements, which include effective thermal management, starting with the design of the
printed-circuit board.
High-integration–density active power devices, such as MOSFET transistors, can dissipate a
significant amount of heat and therefore require PCBs that can transfer heat from the hottest
components to ground planes or heat-dissipating surfaces, operating as efficiently and effectively
as possible. Thermal stress is one of the main causes of malfunctioning of power devices, as it
leads to a degradation of performance or even a possible malfunction or failure of the system. The
rapid growth of the power density of devices and the constant increase in frequencies are the main
reasons that cause excessive heating of electronic components. The increasingly widespread use of
semiconductors with reduced power losses and better thermal conductivity, such as wide-bandgap
materials, is not in itself sufficient to eliminate the need for effective thermal management.
MARCH 2022 | www.powerelectronicsnews.com 21Design
Current silicon-based power devices achieve a junction temperature between about 125˚C and
200˚C. However, it is always preferable to make the device operate below this limit, as this would
lead to a rapid degradation of the same and a reduction of its residual life. In fact, it has been
estimated that an increase of 20˚C in the operating temperature, caused by improper thermal
management, can reduce the residual life of the components by up to 50%.
LAYOUT APPROACH
An approach to thermal management commonly followed in many projects is to use substrates with
standard Flame Retardant Level 4 (FR-4), an inexpensive and easily workable material, focusing on
thermal optimization of the circuit layout.
The main adopted measures concern the provision of additional copper surfaces, the use of traces
with a greater thickness, and the insertion of thermal via beneath the components that generate
the greatest amount of heat. A more aggressive technique, capable of dissipating a greater amount
of heat, involves inserting into the PCB or applying on the outermost layers real copper blocks,
typically in the shape of a coin (hence the name “copper coins”). The copper coins are processed
separately and then soldered or attached directly to the PCB, or they can be inserted into the inner
layers and connected to the outer layers through thermal vias. Figure 1 shows a PCB in which a
special cavity has been made to house a copper coin.
Copper has a thermal conductivity
coefficient of 380 W/mK, compared
with 225 W/mK for aluminum and to
0.3 W/mK for FR-4. Copper is a relatively
cheap metal and already widely used in
PCB manufacturing; therefore, it is the
ideal choice for making copper coins,
thermal vias, and ground planes, all
solutions capable of improving heat
dissipation.
Proper positioning of the active
Figure 1: A PCB with a copper coin
components on the board is a crucial
factor in preventing the formation of hot
spots, thus ensuring that heat is distributed as evenly as possible along the entire board. In this
regard, the active components should be distributed in no particular order around the PCB to avoid
the formation of hot spots in a specific area. However, it’s better to avoid placing active components
that generate a significant amount of heat near the edges of the board. Conversely, they should
be positioned as close as possible to the center of the board, favoring an even heat distribution.
If a high-power device is mounted near the edge of the board, it will build up heat on the edge,
22 MARCH 2022 | www.powerelectronicsnews.comDesign
increasing the local temperature. If, on the other hand, it is placed near the center of the board,
the heat will dissipate on the surface in all directions, reducing the temperature and dissipating the
heat more easily. Power devices should not be placed close to sensitive components and should be
properly spaced from each other.
The actions taken at the layout level can be further improved through the adoption of active or
passive cooling systems, such as heatsinks or fans, whose function is to remove heat from active
components rather than dissipate it directly in the board. In general, designers must find the right
compromise between different thermal-management strategies based on the requirements of the
specific application and the available budget.
PCB SUBSTRATE SELECTION
Due to its low thermal conductivity — between 0.2 and 0.5 W/mK — FR-4 is generally not suitable
for applications in which a large amount of heat needs to be dissipated. The heat that can build up
in high-power circuits is considerable, compounded by the fact that these systems often operate in
harsh environments and extreme temperatures. Using an alternative substrate material with higher
thermal conductivity may be a better choice than using the traditional FR-4.
Ceramic materials, for example, offer significant advantages for thermal management of high-power
PCBs. In addition to improved thermal conductivity, these materials offer excellent mechanical
properties that help compensate for the stress accumulated during repeated thermal cycling.
Additionally, ceramic materials have lower dielectric losses operating at frequencies up to 10 GHz.
For higher frequencies, it is always possible to opt for hybrid materials (such as PTFE), which offer
equally low losses with a modest reduction in thermal conductivity.
The higher the thermal conductivity of a material, the faster the heat transfer. It follows that metals
such as aluminum, in addition to being lighter than ceramic, offer an excellent solution for transferring
MARCH 2022 | www.powerelectronicsnews.com 23Design
heat away from components.
Aluminum particularly is
an excellent conductor,
has excellent durability, is
recyclable, and is non-toxic.
Thanks to their high thermal
conductivity, the metal layers
help to quickly transfer heat
throughout the board. Some
manufacturers also offer
metal-clad PCBs, wherein
both outer layers are metal-
clad, typically aluminum or
galvanized copper. From a
cost-per-unit-weight point
of view, aluminum is the best
choice, while copper offers
higher thermal conductivity.
Aluminum is widely used for
the construction of PCBs that
Figure 2: Example of aluminum PCB for high-power RGB LEDs
support high-power LEDs
(an example is shown in Figure
2), in which it is also particularly useful for its ability to reflect the light away from the substrate.
Even silver, thanks to its thermal conductivity approximately 5% higher than copper, can be used
to make tracks, via holes, pads, and metal layers. Also, if the board is used in a humid environment
where noxious gases are present, using a silver finish on exposed copper traces and pads will help
prevent corrosion, a typical threat found in these environments.
Metal PCBs, also known as insulating metal substrates (IMS), can be laminated directly into the
PCB, resulting in a board with FR-4 substrates and metal core with single-layer and double-layer
technology with depth control routing, which serves to transfer heat away from on-board components
and to less critical areas. In IMS PCBs, a thin layer of thermally conductive but electrically insulating
dielectric is laminated between a metal base and a copper foil. The copper foil is etched into the
desired circuit pattern and the metal base absorbs heat from this circuit through the thin dielectric.
24 MARCH 2022 | www.powerelectronicsnews.comDesign
The main advantages offered by IMS PCBs are the following:
▶ Heat dissipation is significantly higher than standard FR-4 constructions.
▶ The dielectrics are typically 5× to 10× more thermally conductive than normal epoxy
glass.
▶ Thermal transfer is exponentially more efficient than in a conventional PCB.
Besides LED technology (illuminated signs, displays, and lighting), IMS circuit boards are widely used
in the automotive industry (headlights, engine control, and power steering), in power electronics
(DC power supply, inverters, and engine control), in switches, and in semiconductor relays.
For More Information
▶ PCB Design
▶ PCBs for LEDs
MARCH 2022 | www.powerelectronicsnews.com 25TEST & MEASUREMENT
Test Solutions for EV
Inverters
By Giovanni Di Maria, technical writer for EEWeb
Tests on power-electronic systems are indispensable, but they can’t all happen in hardware.
Engineers can obtain exact operating models of inverters using virtual simulations of circuit
operation and, above all, simulations of defects, problems, and accidental events. To help them
on this, NI has announced new solutions and collaborations that improve test environments and
workflows for the development of electric-vehicle drive inverters.
The power circuits of EVs are extremely sophisticated due to their complex hardware and software,
and minor, unforeseen events can result in abnormal — often dangerous — operation. To ensure
safety and performance, designers carry out many types of tests under each profile of operation,
with simulations covering a wide range of scenarios, including the most remote and unexpected.
While this approach helps accelerate EV innovation by integrating testing early in the product
development life cycle, constantly changing models in simulation can be cumbersome and
time-consuming because engineers must recreate both the most realistic scenarios and those that
are difficult to replicate on the road using EV thruster simulations and hardware-in-the-loop tests.
26 MARCH 2022 | www.powerelectronicsnews.comTest & Measurement
To help on this task and improve test environments and workflows for validating EV traction
inverters, NI recently announced a new inverter test system and a collaboration agreement with
D&V Electronics for power inverter testing.
A NEW PERSPECTIVE
Thanks to the collaboration between NI and D&V, it is now possible to simulate and test the
operation of electric motors in direct current. This allows engineers to test the inverter
(see figure) at full power in a safe, cost-effective, and high-quality environment, making the whole
development process go faster without adding cost. “The simulation models are not constant,”
said Brandon Brice, principal solutions marketing manager at NI. “They are continually modified
by engineers, based on evolving EV performance and test requirements. These constant changes
introduce inefficiencies in the testing process that NI and D&V solutions help avoid.”
D&V’s innovative power emulators combined with NI’s high-speed test platforms offer new,
future-ready capabilities for EV and inverter testing. Both companies understand that an excellent
simulation must consider all the existing variables, including hardware, logic signal levels, and
power levels and that the simulation of the batteries includes not only normal operation in static
and dynamic conditions but also the characterizations that determine their longevity and reliability.
These battery tests also cover the accumulator cell’s charge/discharge process and allow for the
monitoring of system temperatures. Therefore, modifications and variations in operation can be
made to verify how environmental changes are addressed on the battery itself, especially for safety.
MARCH 2022 | www.powerelectronicsnews.com 27Test & Measurement
ENERGY AND SAFETY
One of the main purposes of
companies is to increase battery
capacity, and it is no doubt a
fundamental element in the EV
sector. However, another equally
crucial aspect, that of safety,
must be flanked and balanced
by this factor. It is an extremely
dangerous area where a battery
pack could catch fire, causing all
safety devices to malfunction.
For example, thermal leaks can
occur at any time, especially
with lithium-ion batteries, and
it is particularly difficult to
put out such fires. The critical
decision around energy and
safety is at what speed a battery
can charge and discharge so NI’s inverter test system
that it can provide all the energy
to the inverter efficiently. Due to
the various preliminary tests that allow engineers to simulate the entire system from the start and
have all clear ideas, as the tests are performed, cost reduction should be directed at achieving less
production waste and fewer faulty batteries.
LABVIEW
LabVIEW is powerful software that can help you execute a variety of tests because it allows engineers
to run models and mimic other components in real time. It is possible, for example, to run a motor
model with a particular battery, as if the whole system was real, complete with acceleration and
deceleration at extremely high rates. Furthermore, without having the physical motor, it is possible
to mimic and emulate hardware, as well as test the battery as if it were connected to the vehicle.
When the simulation of an inverter successfully passes a test, you move on to the control of all
other related components. “Eventually, the software will have different models for the batteries,
for the motors, for the power electronics, for the charger, and so on,” said Brice. “All these models
can be simulated together or separately in order to obtain a realistic final view. These models can
be perfectly imported and also acquired by other software, all in real time with extremely fast test
management.”
28 MARCH 2022 | www.powerelectronicsnews.comTest & Measurement
FAULT SIMULATION
A good simulator must include all of the necessary tools, and the test software must run in real time.
The possibility of having many I/O ports as well as signal-conditioning tools and numerous digital
buses (CAN and others) is extremely important. However, one option that must not be overlooked
is the ability to introduce an intended problem into the system. The ability to program a failure in
tests is extremely important to see how the controller might react in the most critical areas of the
vehicle. To meet high safety standards, it is necessary to anticipate as many conceivable scenarios
as possible, including the most unlikely ones. If the engine fails, the driver must still be able to
drive the vehicle home, albeit at a reduced speed and performance, but in total safety. The tests
must also implement these types of controls and forecasts. Both the hardware and the software
must be able to simulate mistakes and unanticipated difficulties.
CONCLUSION
Performing extensive tests early in the product development life cycle accelerates innovation.
The research is based on complex simulation models of all the power components in EVs. The
collaboration of people, ideas, and technology enables humanity to successfully address the world’s
greatest issues, such as reducing carbon emissions through the massive adoption of EVs.
For More Information
▶ Electric Vehicles
▶ NI’s Inverter Test System
MARCH 2022 | www.powerelectronicsnews.com 2930 MARCH 2022 | www.powerelectronicsnews.com
TEST & MEASUREMENT
Wide-Bandgap
Materials Address
EV Power and Efficiency
Requirements
By Maurizio Di Paolo Emilio, editor-in-chief of Power Electronics News
Silicon carbide and gallium nitride technologies have grown enormously over the past few years, proving
to be commercially available energy-saving technologies. During the last “Wide Bandgap Devices and
Applications Short Course,” a virtual event organized by PowerAmerica on Nov. 16–17, 2021, instructors
coming from leading semiconductor companies, universities, and institutions explained how
MARCH 2022 | www.powerelectronicsnews.com 31Test & Measurement
wide-bandgap semiconductors enable clean energy manufacturing, high technology, job creation,
and energy savings.
The first speaker, Peter Friedrichs, senior director of SiC at Infineon Technologies, has talked about
SiC power device technology, focusing on aspects such as device design, reliability, and system
benefits. According to Friedrichs, SiC has still a substantially higher price than silicon, mainly due
to the substrate (wafer) manufacturing process and to its higher defect density. However, by using
multiple substrates and dropping down the defect density, Infineon has been able to reduce the
overall production costs.
“The more cells, or the more channel widths, you can place into a given area, the more efficient
your device becomes; it means also the concept that the best volume utilization is favored,” said
Friedrichs.
The first step, which is already productive now on Infineon sites, has been achieved through the
innovative Cold Split technology, which can process crystal material efficiently and with minimal
waste of resources. Today, traditional wire sawing wastes up to 75% of raw material, while the
already-deployed SiC boule splitting is able to reduce raw material losses by 50%. In the near
future, Infineon will use this technology to split entire SiC wafers, thus doubling the number of
chips out of one wafer, as shown in Figure 1.
Figure 1: Cold Split technology reduces raw material losses during SiC manufacturing. (Source: Infineon Technologies)
32 MARCH 2022 | www.powerelectronicsnews.comTest & Measurement
In SiC planar MOSFETs, the channel resistance is normally pretty high. That means we can achieve
a low on-resistance as a final device only if we apply a significantly higher electric field across a
gate oxide. Today, nearly all common MOSFETs have more than a 3-MW/cm electric field applied to
the gate oxide. Planar is relatively easy and cheap processing that allows you to achieve a very good
shielding of the gate oxide in blocking mode. However, it has lower channel mobility and limited
device area shrink options. On the other side is the trench design, which brings benefits such
as lower on-resistance, smaller parasitic capacitance, and improved switching performance. The
drawback, however, is a reduced short-circuit tolerance due to the lower on-resistance.
“SiC system benefits and value propositions, which we believe are striking and unique, include
the solar inverter — with a significant increase in power-handling capability while keeping volume
and rate nearly constant — motor drives, and EV charging, especially with the ultra-high–power
charging up to 350 kW, very high voltages, very high currents, and fast switching,” said Friedrichs.
The next speaker was Victor Veliadis, executive director and CTO of PowerAmerica, who talked
about the SiC market outlook and some of the key applications. Power devices are large discrete
transistors capable of switching high currents and blocking high voltages. The critical electric fields
and energy gaps for SiC and GaN are much higher than those of silicon. Because breakdown voltage
is inversely proportional to the critical electric field, if we increase the critical electric field tenfold,
the thickness of this drift layer becomes 10× smaller, reducing the resistance of the device we are
fabricating. For a specific breakdown voltage, the resistance will be proportional to the inverse
third power of the critical electric field. So if we have a critical electric field that is 10× bigger, the
resistance contribution of this layer will be 1,000× smaller.
“Large critical electric fields allow you to make high-voltage devices with much thinner layers
than what you would have in silicon,” said Veliadis. “That reduces the resistance, the associated
conduction losses, and the overall capacitance. That allows you to operate at higher frequencies
and temperatures, with higher efficiency, and it simplifies a lot of the magnetic circuits, the volume,
the weight.”
While silicon is still competitive at lower voltages up to 650 V, SiC and GaN offer efficient
high-frequency and high-current operation at higher voltages. The big battleground among Si, SiC,
and GaN plays around 650 V, where all the devices are suitable for the 400-V EV bus voltage.
“Looking at some of the opportunities, the first one is automotive with the electric vehicles,” said
Veliadis. “UPS for data centers is another big area where silicon carbide can play a significant
role. Other applications include green infrastructure — basically, photovoltaic and wind energy
— electric motor drives, micro grid, and fast-charging stations. This is where 6.5-kV and 10-kV
MOSFETs are going to be needed.”
MARCH 2022 | www.powerelectronicsnews.com 33Test & Measurement
The SiC device market is projected to be a $3.2 billion market by 2025, with a phenomenal CAGR
growth over the years, up to 50%.
Wide-bandgap devices can be of a lateral or vertical configuration (see Figure 2). The larger the
separation between drain and gate, the higher the breakdown voltage the device can tolerate.
However, if we increase this distance so much, the device will take up too much space on the wafer,
increasing the overall cost. The solution is to go vertical. Instead of having a large gate to drain
separation in the horizontal direction and take up space on the wafer, we do that in the vertical
direction. That’s the reason why the vast majority of SiC devices have a vertical configuration.
Figure 2: Lateral versus vertical configuration (Source: Victor Veliadis)
Burak Ozpineci, section head for vehicle and mobility systems research at the Oak Ridge National
Lab (ORNL) in Knoxville, Tennessee, gave a presentation on power electronics for EVs. “We are still
focused on pure EVs and looking at electric vehicles going beyond the 200-mile range, with 60-kWh
or higher energy storage,” he said. “We are currently looking at ways of integrating the motors and
power electronics inside the chassis.”
ORNL’s roadmap defines the pathway to achieving 2025 targets, which consist of an increase of
the power density, power level, and vehicle reliability/lifetime, halving the total cost per kilowatt.
Ozpineci has presented five major keystone projects developed at ORNL in this area:
▶ Look for technologies that help us achieve higher power densities. These technologies
include new materials and substrates (such as insulated metal substrate with insertion
of thermal pyrolytic graphite or direct bonded copper), a genetic algorithm for heatsink
optimization, and reduction of volume for DC-link capacitors.
34 MARCH 2022 | www.powerelectronicsnews.comTest & Measurement
▶ New topologies for electric motors. Because ORNL is provided with a supercomputer
facility, it can be used to generate a high-fidelity model of the motor, such as the outer
rotor motor, which has the stator inside and the rotor outside.
▶ The outer rotor motor integrates the inverter right into the motor, eliminating the
connectors and long cables and reducing the size of the motor by up to 30%. This is the
third keystone project; that is, the integrated electric drive.
▶ Medium-duty and heavy-duty electric drives. This project aims to extend the research area
from passenger vehicle electric drives and component technologies to medium-duty and
heavy-duty electric drives. That means higher-voltage batteries (1,000–1,500 V), higher
current levels, and higher power requirements for charging (greater than 1 MW).
▶ Wireless charging. Right now, the research is focused on 200-kW–plus stationary or static
wireless charging. The goal is to go to 270 kW, a power level that can be achieved only
with SiC devices, looking also at dynamic wireless charging.
Iqbal Husain of North Carolina State University has talked about wide-bandgap power electronics
driving high-speed electric machines for EVs. The four major areas of power conversion that are
used in the electric powertrain are the inverter, the DC/DC converter, the converter that supplies
the low-voltage electronics, and an on-board charger. So SiC devices offer this opportunity for
enabling either smaller batteries or longer driving range with smaller, cooler, and lighter systems in
the various converters.
“In all of these areas, there is the opportunity for using silicon carbide devices because of the
advancements and the stage they are in, in terms of their availability and commercial production,”
said Husain. “Our ultimate goal is to improve both the efficiency and the power density.”
For More Information
▶ PowerAmerica Institute
▶ Silicon Carbide Book
MARCH 2022 | www.powerelectronicsnews.com 35ENERGY STORAGE
Battery Technology
for Automotive: An
Interview with Vicor’s
Nicolas Richard
By Maurizio Di Paolo Emilio, editor-in-chief of Power Electronics News
Today’s automobile battery must do a lot more than just start the car and keep the radio on for
the journey. In the last 10 years, the amount of electricity required by car features and electronic
gadgets has doubled, and it is expected to increase again in the next five years. The 12-V battery is
responsible for powering everything from heated seats and entertainment systems to cutting-edge
safety features like pedestrian-detection systems, as well as sustaining the car’s electrical network
during the trip.
A high-voltage (HV) battery that drives the powertrain and an improved 12-V battery that performs
essential offloads are required for electric automobiles. When the HV system is turned off, the
12-V battery is used to safely initialize the HV battery and to power vehicle applications. During the
driving cycle, the HV battery recharges the 12-V battery and maintains loads.
36 MARCH 2022 | www.powerelectronicsnews.comEnergy Storage
Let’s examine the characteristics of this technology
with Nicolas Richard, director of EMEA automotive
business development at Vicor.
Power Electronics News: What is the future
of the 12-V lead-acid battery?
Nicolas Richard: Europe has decreed that no new
cars will have lead-acid batteries after 2030, creating
a considerable challenge for OEMs to find alternative
solutions.
While this may seem like a daunting task, it also
presents a tremendous opportunity to eliminate the
Vicor’s Nicolas Richard
environmentally toxic lead-acid battery while also
reducing weight in a vehicle and improving overall efficiency. It can also reduce the battery warranty
costs faced by manufacturers.
PEN: What is the function of the 12-V battery in a vehicle today?
Richard: The most essential role of the 12-V battery has been to provide a reservoir of power for
loads that require a lot of power. The typical load in a vehicle will have two types of current draw:
one for start, particularly cold-crank start, and one for steady-state operation. Both require a large
amount of current either to charge a capacitor or to turn an armature but then drop down as a
steady-state reservoir.
PEN: What are the technical considerations that need to be supported if the 12-V
battery is eliminated?
Richard: Slew rate or transient response. Replacing the 12-V battery in a vehicle with a traditional
converter may cause the load voltage to drop low enough that the load turns off, thus causing a
reboot in a vehicle. Automobile manufacturers typically require 250 A/ms for their fastest loads,
which 12-V batteries can achieve (75 A/30 µs).
PEN: What alternative 12-V battery technologies exist?
Richard: Replacing the 12-V lead-acid battery with a 12-V Li-ion battery is one option. While this
slightly reduces weight, it retains the decades-old legacy of the 12-V Power Distribution Network
(PDN), which yields limited additional benefit. The other option is to support a 12-V PDN powered
from the primary 400-V or 800-V battery in electric vehicles and hybrid EVs/plug-in hybrid EVs.
MARCH 2022 | www.powerelectronicsnews.com 37Energy Storage
PEN: What are the benefits and disadvantages of these technologies?
Richard: Simply replacing the 12-V lead-acid battery with a 12-V Li-ion battery saves approximately
55% in weight; however, it has a high impact on costs. The 12-V Li-ion battery needs a battery
management system to control the charging and maintain the full battery operation over the vehicle’s
life.
Furthermore, adding a bulky DC/DC converter from HV to 12 V (with a voltage and current regulation
feature) is needed to recharge the 12-V Li-ion battery and supply the electrical loads.
However, this adds limited benefits. Conversely, it adds weight, vehicle packaging complexity, and
system cost while also reducing overall vehicle reliability.
By contrast, eliminating the 12-V battery altogether removes 13 kg from the vehicle and can improve
the cargo space by 2.4%.
PEN: What loads will continue to rely on 12 V?
Richard: Most of the safety loads and very low-power loads, such as LED reading lights for the cabin
and electric windows, will stay on 12 V. All the power loads, such as water and oil pumps, will move
to 48 V or HV.
There is a lot of research going on to improve ride comfort. As vehicles become more autonomous,
the travel experience for drivers and passengers will become more like riding on a train, which they
expect to be very smooth. Achieving this smooth ride requires equipment such as electric anti-rolling
bars or active suspension, which need a lot of power.
PEN: Can an EV’s main traction battery be used to supply these loads?
Richard: It makes sense to use the traction motor battery, which is the largest energy source in the
vehicle, to down-convert to different safe voltages. Typically, the traction motor battery in an EV is
either 400 V or 800 V.
A better approach to solving this problem is to completely rethink the PDN in a vehicle: Eliminate the
physical 12-V battery and replace it with a 12-V “virtual” battery from the primary EV battery.
The ideal vehicle architecture would be one HV battery used to power the powertrain and all the
auxiliary loads, provided it can meet the necessary transient response requirements. Vicor high-
density bus converter module technology enables this approach by virtualizing a low-voltage battery
(48 V or 12 V) directly from the HV battery.
38 MARCH 2022 | www.powerelectronicsnews.comEnergy Storage
A modular power approach combined with innovative topologies allows you to far exceed the slew
rate — the transient response — of a 12-V lead-acid battery (75 A/30 µs). Using power modules and
a sine-amplitude converter, you can process thousands of amperes from the HV battery to the load,
eliminating any dips or loads falling out of regulation.
PEN: What are the advantages of doing this?
Richard: The major advantage is that there is only one battery in the vehicle, saving weight. It also
simplifies the architecture and reduces the packing requirements, which can be complicated for a
heavy 12-V battery.
This method also reduces the warranty costs for the vehicle OEM. The 12-V battery is usually under
warranty for three years. During this time, the responsibility of changing it if it fails rests with the
vehicle manufacturer, leading to potentially considerable costs.
The Vicor modular approach allows engineers to achieve approximately 300 combinations of power
delivery by using just three to four scalable building-block modules of various types. Vehicles ranging
in size from a small city car to a large SUV can all be powered from a common power platform using
the same DC/DC converter. Scaling up simply involves the addition of more power modules.
This design approach saves hundreds of hours of time and a large amount of resources, allowing
OEMs to get out in front in the electrification race.
PEN: What are the main considerations to bear in mind when using this method?
Richard: A traditional PDN based on a discrete DC/DC converter design can consist of over 200
bulky components, whereas Vicor advanced technology provides a single high-density power module.
The time savings for an engineering design team are significant, with qualification required for only
one module compared with over 200 individual components for the same function.
PEN: What extra architecture or components are needed to achieve this?
Richard: No extra components are necessary, as an HV to 12-V DC/DC converter is already an
essential component for EVs. Vicor’s technology brings additional features because of the fast
transient response that results from the removal of the 12-V battery.
Vicor is also conducting research into redundancy of components to improve functional safety.
MARCH 2022 | www.powerelectronicsnews.com 39You can also read
