Basics of SiC Power Devices
Basics of SiC Power Devices
Powered by
-Table of Contents-
Introduction ....................................................................................................................................... 1
necessary for device fabrication over wide ranges. For Further, the band gap is about three times that of Si,
these reasons, SiC is anticipated to be a material for use making possible operation at higher temperatures than
in power devices that go beyond the limits of Si. Below is for Si devices. At present, due to constraints imposed by
a summary of SiC's features. the thermal resistance of packages, operation is
guaranteed at 150 ℃ to 175 ℃ , but as package
IV-IV group compound semiconductors in which Si and
technology advances, guaranteed operation at 200 ℃
C bond 1-to-1
and above will become possible.
Close-packed structures in which Si C atom pairs form
Important points have been explained briefly. The
unit layers
discussion may have been a bit difficult for persons
Various polytypes exist; the 4H-SiC polytype is optimal without a background in material properties and
for power devices processes, but even without understanding all the details,
Bonding strength is extremely high ⇒ thermally, you will be able to make use of SiC power devices.
chemically, mechanically stable <Summary>
– Thermal stability: No liquid phase at ordinary The physical properties of SiC are well-suited to power
temperatures, sublimation at 2000℃ devices.
– Mechanical stability: Mohs hardness (9.3) near that Compared with Si semiconductors, losses are low, and
of diamond (10) dynamic characteristics in high-temperature
environments are excellent.
– Chemical stability: Inert in the presence of nearly all
acids and alkalis 2. Development Background and Advantages of SiC
Power Devices
1.2 Features of SiC Power Devices
SiC power devices are capable of high withstand voltages,
With dielectric breakdown electric field strength low resistances and fast operation exceeding Si power
approximately ten times higher than that of Si, SiC can devices, and can operate at higher temperatures as well.
achieve very high breakdown voltage from 600 V to We will talk about the background to development of SiC
thousands of volts. Doping concentrations can be made devices and their specific advantages.
higher than those in Si devices, and drift layers can be
made thin. Nearly all of the resistance component of a 2.1 Development Background of SiC Power Devices
high voltage power device is the resistance of the drift Previously we explained that by using SiC in power
layer, and the resistance value increases in proportion to devices, power conversion can be performed with lower
the thickness of the drift layer. When using SiC, the drift losses than is possible using conventional Si power
layer can be made thin, and so a device with a high devices. This is the main reason why SiC semiconductors
voltage and extremely low turn-on resistance per unit have been commercialized for power devices. And it is
area can be fabricated. Theoretically, for a given high easy to imagine that the background to this is the energy
voltage, the drift layer resistance per unit area can be conservation measures that the whole world is facing.
reduced to 1/300 of that for Si.
Taking as examples low-power DC/DC converters, the
In a Si power device, IGBTs (insulated-gate bipolar spread of mobile devices has resulted in conversion
transistors) and other minority-carrier devices (bipolar efficiencies that exceed 90% as a matter of course.
transistors) have mainly been used in the past in order to However, it is thought that there is still room for
alleviate the increase in turn-on resistance that improvement of the efficiency of high-voltage and large-
accompanies higher breakdown voltages. However, the current AC/DC converters. It is well known that directives
large switching losses give rise to heat generation for energy conservation, centered on the EU, have
problems, imposing limits on high-frequency driving. resulted in urgent demands for energy efficiency including
Using SiC, such fast majority-carrier devices as Schottky reduced standby power consumption of electric and
barrier diodes and MOSFETS can be designed for high electronic equipment.
voltages, making possible the simultaneous attainment of
Given such circumstances, reducing energy losses that
a high voltage, low turn-on resistance, and fast operation,
occur during power conversion is an urgent order of
parameters that entailed trade-offs in silicon devices.
business, and to this end, naturally a material that
transcends the limits of Si must be utilized in power reducing energy consumption as a result of installation
devices. locations, transportation and the like is another major goal.
By using SiC power devices, switching losses can be cut <Summary>
by 85% compared for example with IGBTs. As indicated SiC has been developed as one solution to energy-
in this example, there can be no doubt that SiC power related problems.
devices are one solution to energy problems.
SiC power devices have the great advantage of not
2.2 Advantages of SiC only reducing losses but also contributing to the
miniaturization of equipment.
The use of SiC power devices enables a significant
reduction in energy loss. This great advantage comes 3. What is SiC Schottky Barrier Diode?
from the characteristics of SiC power devices as
In this chapter, we will explain what SiC Schottky barrier
described above: low resistance, high-speed operation,
diodes (hereafter referred to as SiC SBDs) are and how
and high temperature operation.
they compare with Si diodes.
We proceed with a comparison with Si. "Low resistance"
3.1 Characteristics of SiC SBDs and Comparison
leads directly to low losses, but for the same resistance
with Si SBDs
value, the area of an element (chip) can be made smaller.
When handling large amounts of power, sometimes First, the structure of the SiC SBD is explained. As
power modules are used, in which multiple transistors shown in Figure 3, a metal is bonded (Schottky junction)
and diodes are incorporated into a module. As an to a semiconductor, SiC, to obtain a Schottky barrier.
example, for the same capacity, a SiC power module can The structure is basically the same as that of a Si
be designed to be about 1/10 the size of an Si module. Schottky barrier diode (hereafter referred to as Si SBD),
and the key feature is the same: high speed.
Where "fast operation" is concerned, by raising the
switching frequency, smaller size peripheral components The features of SiC SBDs are principally excellent high-
such as transformers, coils and capacitors can be used. speed operation combined with a high voltage. In order to
There are examples in which sizes could actually be raise the breakdown voltage of a Si SBD, the n-type layer,
shrunk to 1/10 or so. shown in Figure 3, is made thicker, and the carrier
concentration is lowered; but this means a high
"High-temperature operation" means that operation at
resistance value, a higher forward voltage, and other
higher temperatures is allowed, and so heat sinks and
changes resulting in large losses and characteristics not
other cooling mechanisms can be simplified.
suited to practical use. Therefore, the breakdown voltage
of Si SBDs is said to be almost limited to 200V. On the
other hand, SiC has a dielectric breakdown field intensity
ten times greater than that of silicon, and so SiC devices
can have high voltages while retaining characteristics
well-suited for practical use. The thicknesses of the n-
type layer of the Si SBD and the n-type layer of the SiC
SBD in Figure 3 are schematic images of how SiC can be
made thin but with high breakdown voltage.
3.3 Reverse Recovery Characteristics of SiC SBD The waveforms in Figure 6 show the current variation with
time during reverse recovery for SiC SBDs and FRDs,
The reverse recovery characteristics of SiC SBDs are
which are high-speed Si PNDs. It can be seen that the
explained based on a comparison with Si PNDs. Since
SiC SBD, indicated by the red line, has less reverse
the reverse recovery characteristic is a fundamental and
current and shorter trr. This characteristic is a
important parameter for diodes, especially for high-speed
consideration because the reverse current is a loss.
type diodes, it is useful to understand not only the
numerical comparison of trr, but also its waveform and Next, a cross-sectional view of each diode is used to
temperature characteristics when using the diode. illustrate the differences in operation. Figure 7 shows the
movement of electrons and holes when the bias of the Si
3.3.1 Differences in Reverse Recovery
PND is shifted from forward to reverse bias.
Characteristics between SiC SBDs and Si PNDs
Reverse recovery is a phenomenon in which the diode is
not immediately and completely turned off when it enters
a reverse bias state, but the reverse current flows for a
certain period of time. trr is the time during which the
reverse current flows. We have already mentioned that
the trr of SiC SBDs is faster than that of Si PNDs including
Si FRDs, but we will explain the reason and actual
characteristics.
Differences in the value of trr and the reverse recovery
characteristic are, to simplify, due to differences in the
diode structure. An explanation would involve a
discussion of electrons and holes moving through the
semiconductor, but we will begin by using waveform
diagrams to review the differences in the reverse
Figure 6. Comparison of trr characteristics
recovery characteristics of SiC SBDs and Si PNDs.
between SiC SBD and Si FRD
During a forward bias, carriers are injected and current These differences in reverse recovery times are all due to
flows through the recombination of holes and electrons. the diode structures. Hence reverse recovery is also rapid
When a reverse bias occurs, holes (the minority carrier) for a Si SBD. However, the withstand voltage of Si SBDs
in the n layer return to the p layer, but this takes some is about 200 V, and the devices cannot be used at higher
time, and a current flows until the holes finally return voltages. In contrast, by using SiC, SBDs having high
(some are annihilated due to the finite lifetime). This is the withstand voltages of over 600 V can be manufactured.
reverse recovery current. This is a major advantage of SiC SBDs.
Figure 8 shows transition to reverse bias of an SiC SBD. Next, we present data relating to the temperature
Because of the Schottky barrier structure, there is no PN dependence and the current dependence of the reverse
junction, and so there are no minority carriers, and during recovery characteristic in Figure 9.
reverse bias there is only the return of electrons, which
are the majority carrier in the n layer. Hence only a very
short reverse recovery time is necessary, and compared
with a Si PND, the device is turned off for a much shorter
time.
Figure 9. Temperature and current dependence of reverse recovery characteristics of Si FRD and SiC SBD
Figure 12. trr and VF vs. loss for Si FRD and SiC SBD
The faster the trr and the lower the VF value, the lower
are the total losses for the device. In the Si FRD, a faster
trr means the VF is higher. However, in a second-
generation SiC SBD, the VF is lowered from 1.5 V to 1.35
V while maintaining the fast trr value of conventional SiC
SBD devices. In this comparison, it can be seen that the
second generation SiC SBDs have the potential to reduce
losses the most.
<Summary> Figure 13. Evolution of SiC SBDs
The VF of an SiC SBD rises with temperature, whereas
the VF of an Si PND (FRD) falls. Through innovations in manufacturing processes,
second-generation SiC SBDs retain leakage currents and
The increase in the VF of an SiC SBD at high trr performance comparable to previous devices while
temperatures causes the IFSM to decline, but there is lowering VF by approx. 0.15 V. As a result, VF-induced
no thermal runaway, such as occurs in Si PNDs (FRDs), conduction losses are improved.
the VF of which declines.
In the third generation, a JBS (Junction Barrier Schottky)
The reduction of VF in SiC SBD contributes to loss structure was adopted with the aim of improving IFMS
reduction. and leakage current (IR). The JBS structure essentially is
3.6 Evolution of SiC SBDs effective with respect to IFSM and IR, but in addition the
low-VF characteristic achieved in the second generation
So far, we have explained the characteristics of SiC SBDs
was further improved. Typical values at Tj=25℃ are the
using Si diodes as a comparison to understand SiC SBDs.
same, but at Tj=150℃, VF is reduced 0.11 V compared
We have also shown that SiC SBDs themselves have
with second-generation devices. That is, there are further
evolved over the generations, and their performance has
advantages for operation under high-temperature
improved. Here, we describe the evolution of SiC SBDs.
conditions. A graph of the reverse voltage VR versus the
As of January 2023, more than 50 models of ROHM's SiC reverse (leakage) current IR is shown in Figure 14, and
SBDs of the second generation and more than 20 models Table 2 comparing numerical values of improved
of the third generation are in mass production and supply. parameters, for second- and third-generation SiC SBDs.
The graphs in Figure 13 compare the VF vs. IF and VF
vs. IFSM (surge current withstand capability)
characteristics of the first through third generations.
Figure 14. Comparison of 2nd and 3rd generation Table 2. Comparison of SiC SBD 2nd and
SiC SBDs with VR and IR 3rd generation characteristics values
Furthermore, a graph showing the temperature 3.7 Advantages of Using SiC SBDs
characteristics of VF is shown in Figure 15. The graph on
We have explained SiC SBDs by comparing their
the left is for the second-generation SCS210AG, and that
characteristics with those of Si diodes. Here, we
on the right is for the third-generation product SCS310AP,
summarize the advantages of using SiC SBDs. Basically,
both at 650 V/10 A. We see that the third-generation
the advantage of using SiC SBDs to replace Si
product has a steep VF-IF curve gradient at high
PNDs/FRDs is the high-speed performance of SiC SBDs.
temperatures and a lower VF for the same IF, an
improvement over the second-generation device. The trr is fast, so that recovery losses can be
dramatically reduced, for higher efficiency
<Summary>
ROHM SiC SBDs have already evolved to the third For a similar reason, the reverse current is small so that
generation. noise is low, and noise/surge suppression components
can be eliminated, enabling enhanced miniaturization
Third-generation products offer improved TFMS and
reduced leakage currents, and further reduce the low High frequency operation enables miniaturization of
VF values achieved in the second generation. inductors and other peripheral components
Figure 15. Comparison of VF vs IF temperature characteristics of 2nd and 3rd generation SiC SBDs
Further, due to their very stable operation with respect to performance, reliability is another important factor. In
temperature, these devices are compatible with particular, power devices are intended to handle large
automotive applications, and the advantages of SiC amounts of power, and so must attain satisfactory
SBDs are being exploited in actual HV/EV/PHV onboard reliability.
charging circuits. Figures 16 and 17 show examples and
Because SiC does not have a long history as a
images.
semiconductor material, and because it does not have
much of a track record compared with Si power devices,
there may not be much awareness of the level of its
reliability. Therefore, ROHM basically conducts reliability
tests in accordance with the test conditions of JEITA ED-
4701.
JEITA ED-4701 is a standard for "environmental and
endurance test methods for semiconductor devices", and
describes testing methods for evaluating semiconductor
devices for industry and consumers. This standard is
widely adopted in Japan. In other words, it has been
confirmed that sufficient reliability is ensured in the same
reliability tests as those for existing Si transistors and ICs.
By the way, where Si SBDs are concerned, you may have
heard that there are failure modes relating to dV/dt or to
dI/dt. With regard to the former, when a high dV/dt is
applied, the failure mode entails breakdown of the outer
peripheral structure of the SiC SBD. However, in studies
conducted up to this time, ROHM SiC SBDs have not
been found to breakdown in this mode even for dV/dt
Figure 16. Advantages of using SiC SBDs: values up to approximately 50 kV/μs.
Example of a PFC circuit
As for dI/dt, in Si FRDs the recovery current Irr is large
when dI/dt is high, so that the failure mode is caused by
current concentration. Hence there are concerns about
the possibility of a similar failure mode. But in SiC SBDs,
the occurrence of such failure is regarded as unlikely
because the recovery current is extremely small.
4. What is SiC MOSFET?
The role of transistors in power conversion circuits is vital,
and various efforts are being made to improve transistors
in order to reduce losses and enable smaller application
size. The advantages of SiC semiconductors have been
described earlier, including low losses, fast switching, and
high temperature operation, all of which are very useful in
their benefit power conversion applications. In this
chapter, we will discuss SiC MOSFETs in comparison
with other power transistors.
Figure 17. Advantages of using SiC SBDs:
Example of on-board charging circuit
4.1 SiC MOSFET Features
3.8 Reliability of SiC SBDs Figure 18 shows the breakdown voltage region of a
typical Si power transistor and SiC MOSFET, a
When evaluating semiconductor devices, in addition to comparison with the same intent as Figure 5, also used
the electrical and mechanical specifications and in the SiC SBD chapter.
Table 3. Features of power devices with rated voltages ranging from 600 V to 2000 V
4.2 Comparison of Various Power Transistor Figure 20 shows a graph of on-resistance and withstand
Structures and Features voltage based on the physical properties and process
specifications used by each transistor. The graph shows
Figure 19 compares the structures, rated voltages, on-
that, theoretically, SiC DMOS can make transistors with
resistances, and switching speeds of various power
higher withstand voltage and lower on-resistance.
transistors.
However, the current SiC DMOS has characteristics in
Even in the bracket of power transistors, structures differ the range circled by the oval, and further performance
depending on the process technology used, and the improvement is expected with future advancements.
electrical characteristics are also different. Incidentally,
DMOS is a general element structure that is a planar-type
MOSFET. However, in recent years, with Si power
MOSFETs, both a high voltage and a reduced ON-
resistance can be obtained, and super junction structure
MOSFETs (hereafter SJ MOSFETs) have come into
widespread use. We have presented a DMOS structure
as an example of a SiC MOSFET, but at present, ROHM
is mass-producing trench-structure SiC MOSFETs with
further improved characteristics. We will be describing
these devices in more detail in the future.
Figure 20. On-resistance and withstand
Features of the different element types are as described voltage of each transistor based on
in the radar charts above. ON-resistance is an issue with physical properties and process specifications
Si DMOS’s, and as explained above, by adopting a SJ
MOSFET structure, the ON-resistance is improved. <Summary>
IGBTs have superior ON-resistances and rated voltages, The features of power transistors differ depending on
but switching speed is a problem. SiC DMOS’s are the materials and structures.
switching elements which excel with respect to rated There are various advantages and disadvantages
voltage, ON-resistance, and switching speed alike, and where characteristics are concerned, but SiC MOSFETs
have the further major advantage of satisfactory exhibit excellent characteristics overall.
operation at high temperatures.
4.3 Differences with Si MOSFET the gate resistance value is inversely proportional to the
chip size, so that the resistance is higher for smaller chips.
Rather than just presenting detailed data, readers who
Because, for devices with equivalent performance, a SiC
have not yet used SiC MOSFETs will probably have
MOSFET chip is small compared with an Si device, the
questions about how they differ from Si MOSFETs with
gate capacitance is small, but the internal gate resistance
respect to driving methods and so on. Here we will explain
is higher. For example, a device rated at 1200 V and 80
two important points relating to driving when comparing
mΩ (an S2301 bare-die product) has an internal gate
SiC MOSFETs with Si MOSFETs.
resistance of about 6.3 Ω (see Figure 22).
4.3.1 Driving Voltage
Compared with their Si counterparts, SiC MOSFETs have
a lower drift layer resistance but a higher channel
resistance, and so the higher the gate-source voltage Vgs,
which is the driving voltage, the lower is the on-resistance.
Figure 21 indicates the relation between on-resistance
and Vgs for SiC MOSFETs.
By developing a proprietary process that prevents breakdown tends to be lower than for Si devices. In the
expansion of stacking faults, ROHM succeeded in case of a 1200 V class MOSFET in a TO247 package, the
securing reliability with respect to body diode conduction. short-circuit withstand time at Vdd=700 V and Vgs=18 V
Figure 33 shows the results of conduction tests for 1000 is roughly 8 to 10 μs. As Vgs falls, the saturation current
hours of an 8 A DC current through the body diode of a becomes smaller, so that the withstand time is longer. And,
second-generation 1200 V, 80 Ω SiC MOSFET product. when Vdd is lower, there is less heat generated, so that
It was verified that there is no fluctuation in any of the the withstand time is longer.
characteristics, including the on-resistance and the
Because the time required to turn off a SiC MOSFET is
leakage current.
extremely short, when the Vgs shutoff speed is fast, a
steep dI/dt can result in a large surge voltage. A soft
turnoff function to gradually lower the gate voltage or
some other method should be used to achieve shutoff
while avoiding overvoltage conditions.
4.7.6 dV/dt Breakdown
In a Si MOSFET, there is a mode in which a high dV/dt
causes a transient current to flow through the capacitance
Cds and turn on the parasitic bipolar transistor, leading to
device breakdown. Because ROHM SiC MOSFETs have
a low current amplification factor (hFE) of the parasitic
bipolar transistor, current amplification is thought not to
occur, and investigations up till now have found no
evidence of this breakdown mode even in operation up to
about 50 kV/μs. The recovery current of a SiC MOSFET
is extremely small, and dI/dt during recovery is low, so
that the dV/dt of the body diode during recovery does not
increase too much either, and therefore it is thought that
this fault mode does not readily occur.
4.7.7 Cosmic Ray Neutron-induced Single-event
Effects
In high-altitude applications, there are cases in which
single-event effects of semiconductor devices, due to
neutrons, heavy ions and other particles that are caused
by cosmic rays that only rarely reach the earth, can cause
problems. As a result of white neutron irradiation tests
(energy: 1 to 400 MeV, conducted at the Research Center
for Nuclear Physics, Osaka University (RCNP)) on SiC
MOSFETs (n=15), no failures were found to occur due to
single-event effects at Vds=1200 V (100% of the rated
breakdown voltage) under neutron irradiation of 1.45×109
neutrons/cm2/s. The failure rate at sea level is 0.92 FIT,
and even at an altitude of 4000 m is 23.3 FIT, which is 3
to 4 orders of magnitude lower than that of equivalent Si
Figure 33. SiC MOSFET: body diode conduction
IGBTs and Si MOSFETs. With a high effective rated
test results
voltage and an ample margin, the risk of failure due to
4.7.5 Short-circuit Rating neutrons originating in cosmic rays can be reduced in
Because of the small chip area and high current density uses at high altitudes and in numerous quantities.
of a SiC MOSFET compared with a S -MOSFET, the
ability to withstand short-circuits that can cause thermal
4.7.8 Electrostatic Discharge Rating 5.1 Switching Losses in Full-SiC Power Modules
A feature of SiC MOSFETs is the ability to reduce chip Full-SiC power modules have two important advantages
sizes relative to Si MOSFETs, but on the other hand, their over IGBT modules: 1) the ability to dramatically reduce
ability to withstand electrostatic discharge (ESD) failure is switching losses, and 2) overall loss reduction that
limited. Hence adequate electrostatic countermeasures becomes more significant as the switching frequency
must be taken when handling these devices. rises.
■Examples of ESD protection measures Figure 35 compares a 1200 V/300 A full-SiC power
Eliminate static electricity from human body, devices, module BSM300D12P2E001 with a functionally
and the work environment using ionizers equivalent IGBT module. In comparison based on data
(recommended) sheet specifications; the switch-on loss Eon, switch-off
loss Eoff, and recovery loss Err are all much smaller, and
Eliminate static electricity from human body and work
Err is negligible because Irr flows very little. As a result,
environment using wristbands and grounding (This
the overall switching loss is reduced by 77%. This is the
measure is ineffective against static charge on devices,
first of the above-described advantages.
so this measure alone is insufficient)
4.7.9 General Reliability Test Conditions
ROHM SiC MOSFETs are based on reliability tests in
accordance with JEITA ED-4701, which is widely adopted
in Japan as a test method for semiconductor evaluation.
Please contact us if you need reliability test results.
5. What is Full-SiC Power Module?
A full-SiC power module is a module that consists entirely
of SiC power devices. ROHM has begun the world's first
mass production of full-SiC power modules, which use
the company's own SiC MOSFETs and SiC SBDs.
Compared with Si IGBT power modules, these products
are capable of faster switching and dramatically lower
losses. The reader will grasp this from the features of SiC
SBDs and SiC MOSFETs in the previous chapters. The
SiC power modules currently being manufactured by Figure 35. Loss comparison between
ROHM are half-bridge type and step-up chopper type full SiC power module and IGBT module
products.
Figure 34. Full SiC power module image and internal circuit examples (2 types)
Figure 36 shows totals for the switching loss and As one would expect, this small switching loss is due to
conduction loss at switching frequencies of 5 kHz and 30 the characteristics of the SiC elements that constitute the
kHz in loss simulations for the case of a PWM inverter. full-SiC module. Full-SiC power modules are capable of
Compared with an IGBT module, total losses are reduced dramatic reductions in switching losses compared with
by roughly 22% at 5 kHz. The orange areas represent equivalent IGBT modules, and the higher the switching
switching losses; nearly all of the reduction in losses is frequency, the greater is the difference in losses relative
due to reduced switching losses. At 30 kHz, the switching to IGBT modules. This means that full-SiC power
loss for the IGBT is greatly increased; this is a well-known modules are capable of high-speed switching while
issue with high-speed switching in IGBTs. The full-SiC sharply reducing losses, in contrast with IGBT modules,
power module also exhibits an increase in switching loss, which are ill-suited to high-speed switching operation.
but the proportion is smaller than the relative increase for
<Summary>
the IGBT module. As a result, we see that at 30 kHz the
Full-SiC power modules are capable of dramatic cuts in
total loss can be reduced by about 60%. This is the
switching losses compared with IGBT modules.
second of the above-mentioned advantages.
The difference is particularly stark at higher switching
frequencies.
5.2 Tips for Practical Use: Gate Driving
From here, we will explain the key points to take full
advantage of the excellent performance of full-SiC power
modules.
5.2.1 Issues Relating to Gate Driving: False Gate
Turn-On
"False gate turn-on" is a phenomenon in a configuration
using a high-side SiC MOSFET and a low-side SiC
MOSFET in which, at the time of SiC MOSFET transition
(switching), ringing occurs in the gate voltage of the high-
side SiC MOSFET, or the gate voltage of the low-side SiC
MOSFET is raised, causing the SiC MOSFET to operate
erroneously. The essence of this phenomenon can be
Figure 36. Loss simulation of full SiC power module
understood easily by referring to Figure 37.
and IGBT module
Figure 42. Gate drive evaluation circuit with false gate turns on suppression circuit
5.3.4 Example of Snubber Capacitor Effect in pathways of large loops, with large inductance.
In the three examples given in the previous section, the As a result, in the case of ceramic capacitors, we can say
surge waveforms occurring between drain and source that the surge suppression effect is greater for a smaller
when the SiC MOSFET is turned off and current is cut off number of parallel groups and for larger total capacitance.
are shown in Figure 51. When a capacitor is added near
Finally, the effect of large-value capacitors is checked. In
the terminals of the SiC MOSFET, there is a surge
Figure 53, Surges are compared when large-capacitance
suppression effect, and the effect differs depending on
capacitors are and are not present, in a state in which five
the type of capacitor.
groups of two series-connected ceramic capacitors are
connected in parallel.
Figure 51. Examples of snubber capacitor effect Figure 53. Snubber capacitor: Effect of
large-capacitance capacitors
Next, waveforms comparing the suppression effect for the
ceramic capacitor example (Figure 51 in red) in three As indicated by the red waveform for the case in which a
under different conditions are shown in Figure 52. large-capacitance capacitor is present, we see that a
large-value capacitor is effective in suppressing long-
period ringing.
In these examples, by connecting a large-value capacitor
and ceramic capacitors close to the terminals of the SiC
MOSFET as snubbers, a considerable suppression effect
is obtained. However, the position of capacitor installation,
the combination of parallel and series connections of
capacitors, and slight differences affect the surge
waveform, and so it is necessary to verify the suppression
effect in actual equipment.
<Summary>
Figure 52. Snubber capacitor: Capacitance
To take advantage of high-speed switching performance,
and effect of ceramic capacitors
snubber capacitors are needed to minimize parasitic
The ringing t1 at a frequency of about 23 MHz inductance in the electrical wiring.
immediately after turn-off attenuates more rapidly for Connect a capacitor near the terminal for power to reduce
fewer capacitor groups in parallel (blue A). This is thought wiring inductance.
to be because the capacitor ESR is inversely proportional
to the number of parallel groups, that is, the smaller the The effect varies depending on the type of capacitor,
number of parallel capacitor groups, the greater the ESR. mounting position, and configuration, so evaluation with
actual equipment is necessary.
When the capacitance of a ceramic capacitor is small,
ringing t2 at a frequency of about 2 MHz occurs (green B,
blue A). This is attributed to charge that has accumulated
5.4 Tips for Practical Use: The Effects of Specialized 5.4.2 Specialized Gate Driver Board and Ceramic
Gate Drivers and Snubber Modules Snubber Capacitor Module
In addition to the snubber capacitors that were discussed Figure 55 shows a specialized gate driver and snubber
in the previous section we explain improvements to capacitor for addition to the basic configuration, and
switching characteristics when a specialized gate driver Figure 56 shows an example installation. They are made
is used. available for use in evaluating full-SiC modules.
5.4.1 Driving Specifications and Basic Configuration
of Full-SiC Modules
The characteristics resulting when using snubber
capacitors and specialized gate drivers are compared for
conditions in Figure 54 and circuit configuration. The
circuit configuration is provided with the electrolytic
capacitors and film capacitors described in the section.
When a specialized gate driver is not used, as a false gate
turn-on countermeasure, a capacitance CGS of microfarad
order is added, and -5 V is applied to VGS as a negative
bias (for the case of a full-SiC module equipped with
second-generation SiC MOSFETs).
Figure 54. Driving specifications and basic configuration
of full-SiC modules
Figure 55. Specialized gate driver and snubber capacitor to be added to the basic configuration
Figure 56. Example of specialized gate driver and snubber capacitors installation
5.4.3 Effect of Specialized Gate Driver and Snubber mJ to 5.3 mJ, while Eoff is reduced from 5.3 mJ to 3.8 mJ.
Module This is because with decreasing inductance, which exerts
a large influence, Eon increases while Eoff decreases. On
To begin with, waveforms are compared when the full-SiC
comparing the overall loss (Eon+Eoff), we see that loss is
power module turned on, both before and after installing
reduced by 0.4 mJ.
the specialized gate driver and the snubber capacitors
(Figure 57). In conclusion, we can say that in order to fully utilize the
performance of a full-SiC module, it is a good idea to add
a snubber and use a gate driver with a specialized design.
Up till now, taking as our theme important points in
utilizing full-SiC modules, we have considered gate
drivers and explained the effect of snubbers, and, in this
article, the use of specialized gate drivers. In order to
perform extremely high-speed switching of high voltages
and large currents, the use of such complementary
components, as well as adjustments based on
evaluations, are important points. Particularly during
initial evaluations, using an evaluation board and other
tools can greatly smooth development.
<Summary>
Figure 57. Comparison turn-on waveforms both before Surges and ringing can be dramatically suppressed by
and after installing the specialized gate driver using a specialized gate driver and a snubber module.
and the snubber capacitors
Where losses are concerned, there is an increase in
In order from the top, the ID, VD, and VG waveforms are Eon and a decrease in Eoff. When comparing overall
shown in Figure 57, with red and orange representing the losses (Eon+Eoff), losses are reduced.
waveforms for cases in which the specialized gate driver
5.5 Support Tools: Full-SiC Module Loss Simulator
and the snubber module are installed, and the blue and
green representing waveforms when they are not A Loss simulator that can be used in the study and
installed. It is clearly seen that surges and ringing are evaluation of full-SiC modules is offered at no cost.
suppressed. Next, the turn-off waveforms are shown in Simply by selecting a full-SiC module and setting the
Figure 58. input conditions, the losses and temperatures of the
transistors and diodes within the module can be
simulated.
Figure 59 is an example of an input screen. G is the power
factor, H is the modulation ratio, K is the heat sink
temperature, and other quantities are as indicated in the
screen.
Revision History
Notes
1. The information contained herein is subject to change without notice.
2. Before you use our Products, please contact our sales representative and verify the latest specifications.
3. Although ROHM is continuously working to improve product reliability and quality, semiconductors can break down
and malfunction due to various factors. Therefore, in order to prevent personal injury or fire arising from failure,
please take safety measures such as complying with the derating characteristics, implementing redundant and fire
prevention designs, and utilizing backups and fail-safe procedures. ROHM shall have no responsibility for any
damages arising out of the use of our Products beyond the rating specified by ROHM.
4. Examples of application circuits, circuit constants and any other information contained herein are provided only to
illustrate the standard usage and operations of the Products. The peripheral conditions must be taken into account
when designing circuits for mass production.
5. The technical information specified herein is intended only to show the typical functions of and examples of
application circuits for the Products. ROHM does not grant you, explicitly or implicitly, any license to use or exercise
intellectual property or other rights held by ROHM or any other parties. ROHM shall have no responsibility
whatsoever for any dispute arising out of the use of such technical information.
6. The Products specified in this document are not designed to be radiation tolerant.
7. For use of our Products in applications requiring a high degree of reliability (as exemplified below), please contact
and consult with a ROHM representative: transportation equipment (i.e., cars, ships, trains), primary communication
equipment, traffic lights, fire/crime prevention, safety equipment, medical systems, servers, solar cells, and power
transmission systems.
8. Do not use our Products in applications requiring extremely high reliability, such as aerospace equipment, nuclear
power control systems, and submarine repeaters.
9. ROHM shall have no responsibility for any damages or injury arising from non-compliance with the recommended
usage conditions and specifications contained herein.
10. ROHM has used reasonable care to ensure the accuracy of the information contained in this document. However,
ROHM does not warrant that such information is error-free, and ROHM shall have no responsibility for any damages
arising from any inaccuracy or misprint of such information.
11. Please use the Products in accordance with any applicable environmental laws and regulations, such as the RoHS
Directive. For more details, including RoHS compatibility, please contact a ROHM sales office. ROHM shall have
no responsibility for any damages or losses resulting non-compliance with any applicable laws or regulations.
12. When providing our Products and technologies contained in this document to other countries, you must abide by
the procedures and provisions stipulated in all applicable export laws and regulations, including without limitation
the US Export Administration Regulations and the Foreign Exchange and Foreign Trade Act.
13. This document, in part or in whole, may not be reprinted or reproduced without prior consent of ROHM.
Powered by