HEFNER MODEL PARAMETERS FOR POWER IGBTS
UNDER PULSED POWER CONDITIONS
A Thesis
presented to
the Faculty of the Graduate School
at the University of Missouri
In Partial Fulfillment
of the Requirements for the Degree
Master of Science
By
JAMES A. VANGORDON
Dr. Scott D. Kovaleski, Thesis Supervisor
December 2010
The undersigned, appointed by the Dean of the Graduate School,
have examined the thesis entitled
HEFNER MODEL PARAMETERS FOR POWER IGBTS
UNDER PULSED POWER CONDITIONS
presented by James A. VanGordon,
a candidate for the degree of Master of Science, and hereby certify
that in their opinion it is worthy of acceptance.
________________________________
Dr. Scott Kovaleski, Thesis Supervisor
Electrical and Computer Engineering
________________________________
Dr. John Gahl
Chemical Engineering
________________________________
Dr. Naz Islam
Electrical and Computer Engineering
________________________________
Dr. Gregory Dale
Los Alamos National Laboratory
University of Missouri Graduate Faculty
Acknowledgements
I would like to thank Dr. Scott Kovaleski for his guidance throughout
this project.
I would also like to thank my committee members for their
insightful comments. I would especially like to thank Dr. Greg Dale. Without
his funding and guidance throughout the duration of the project, this
research would not have been possible.
Additionally, thank you to Dr.
Kovaleski and Dr. Dale for your patience, understanding, and direction.
Thanks to Duane Prussia at Powerex, Inc. for providing some of the
IGBTs utilized in this research. Thank you to Powerex, Inc. for your interest
in this research at various conferences.
I would like to thank many of my fellow students that have provided
both friendship and technical assistance for me while I’ve been a student
working on this project. Special thanks to Brian Hutsel and Andrew Benwell
for camaraderie and assistance in all aspects of life. Thanks to Mark Kemp,
Dustin Sullivan, Bill Zeier, Emily Baxter, Erik Becker, Brady Gall, Alec Hughs,
David Rice, Nelson DeSouza, Brett Scott, and any other fellow students that I
man have unintentionally omitted.
Finally, I would like to thank my parents, John VanGordon and Lynn
and Pam Allen, as well as my grandparents for their love and support
throughout all of my academic and personal endeavors. I would also like to
thank my siblings for their continued support and understanding: Amanda
VanGordon, Madison VanGordon, Emmalie VanGordon, Morgan VanGordon,
Aaron Allen, and Ashley Gash.
ii
Contents
List of Figures....................................................................................... v
List of Tables ...................................................................................... viii
1.
Introduction ................................................................................ 1
Solid-state Pulsed Power ..................................................................... 1
Solid-state Switching .......................................................................... 2
IGBT Switching .................................................................................. 3
Pulsed Power IGBT Characterization ..................................................... 4
Thesis Overview................................................................................. 5
2.
IGBT Modeling ............................................................................. 6
BJT/MOSFET Model............................................................................. 7
Hefner Model ..................................................................................... 9
Oziemkiewicz Implementation of the Hefner Model ............................... 12
Gate-Drain Overlap Area ................................................................ 13
Avalanche Multiplication Exponent ................................................... 13
Emitter Saturation Current Density .................................................. 14
Electron Mobility............................................................................ 15
Threshold Area.............................................................................. 15
Device Area .................................................................................. 16
Gate-Source Capacitance ............................................................... 16
Triode Region Factor ...................................................................... 17
Hole Mobility................................................................................. 17
Ambipolar Recombination Lifetime ................................................... 18
Gate-Drain Overlap Depletion Threshold........................................... 18
Avalanche Uniformity Factor ........................................................... 19
Gate-Drain Oxide Capacitance......................................................... 20
MOS Transconductance .................................................................. 20
Base Doping ................................................................................. 21
Transverse Field Factor .................................................................. 21
Metallurgical Base Width ................................................................ 22
3.
Experiment Setup ...................................................................... 23
iii
Overview ........................................................................................ 23
IGBT Test Circuit.............................................................................. 26
Diagnostics................................................................................... 30
Gate Drive Circuit............................................................................. 32
4.
Experiment Results .................................................................... 35
Least Squares Curve Fitting ............................................................... 35
IGBT Results ................................................................................... 37
Test Data ..................................................................................... 37
Curve Fitting................................................................................. 44
Linear Regression Analysis.............................................................. 51
AGD............................................................................................. 55
MUN ............................................................................................ 61
CGS............................................................................................. 68
MUP............................................................................................. 68
COXD........................................................................................... 73
KP ............................................................................................... 78
5.
Conclusions............................................................................... 82
6.
Future Work .............................................................................. 86
A.
Least Squares Curve Fitting MATLAB Code .................................... 89
B.
IGBT Test Circuit Bill of Materials ............................................... 115
C.
IGBT Test Circuit Printed Circuit Board Layout ............................. 116
D.
Raw Data Tables...................................................................... 121
E.
Coefficient of Determination, R2, Values for Simulated Waveforms . 138
Bibliography ..................................................................................... 155
iv
List of Figures
Figure 2.1: Simple IGBT model using only a pnp BJT and an n-channel
MOSFET............................................................................................... 8
Figure 2.2: Phenomonological IGBT equivalent circuit [44]. ..................... 10
Figure 3.1: Photograph of Powerex QIS4506001 IGBT (a) top view (b)
bottom view ....................................................................................... 24
Figure 3.2: Photograph of IXYS IXEL40N400 IGBT (a) top view (b) bottom
view .................................................................................................. 24
Figure 3.3: Experimental setup system diagram. ................................... 26
Figure 3.4: IGBT test circuit schematic. ................................................ 27
Figure 3.5: Photograph of the IGBT test circuit attached to the input
capacitor............................................................................................ 27
Figure 3.6: Plots for a 2 kV input voltage and a 15 Ω resistive load with a 24
V gate signal on a Powerex IGBT. (a) IGBT collector-emitter voltage and
collector current. (b) IGBT gate-emitter voltage and gate current. ........... 31
Figure 3.7: IGBT gate drive circuit schematic. ....................................... 34
Figure 4.1: Comparison of measured and simulated plots for a 3.5 kV input
voltage with 15 Ω resistive load and a 24 V gate signal. (a) IGBT gate
voltage, VGE. (b) IGBT collector-emitter voltage, VCE, and collector current, IC.
........................................................................................................ 38
Figure 4.2: Comparison of measured waveforms among 3 different Powerex
QIS4506001 IGBTs for (a) collector-emitter voltage (b) collector current and
(c) gate-emitter voltage....................................................................... 40
Figure 4.3: Example of measured waveforms on the IXYS IXEL40N400 with
a 15 Ω load and 1 kVin for (a) collector-emitter voltage (b) collector current
(c) gate-emitter voltage....................................................................... 41
Figure 4.4: Display of the apparent current limit for a Powerex QIS4506001
IGBT with a 3 Ω load, showing (a) collector-emitter voltage and (b) collector
current .............................................................................................. 42
Figure 4.5: Display of the apparent current limit for an IXYS IXEL40N40
IGBT with a 3 Ω load, showing (a) collector-emitter voltage and (b) collector
current .............................................................................................. 43
Figure 4.6: Micro-Cap 9 circuit schematic used for curve fitting ............... 46
Figure 4.7: Comparison of measured and simulated waveforms that match
up well for a Powerex QIS4506001 IGBT for (a) collector-emitter voltage
(R2=0.97572) (b) collector current (R2=0.97623) and (c) gate-emitter
voltage (R2=0.59535).......................................................................... 48
Figure 4.8: Comparison of measured and simulated waveforms that match
up well for a IXYS IXEL40N400 IGBT for (a) collector-emitter voltage
v
(R2=0.96877) (b) collector current (R2=0.96372) and (c) gate-emitter
voltage (R2=0.89290).......................................................................... 49
Figure 4.9: Comparison of measured and simulated waveforms that do not
match up well due to fixing some parameters for a Powerex QIS4506001
IGBT for (a) collector-emitter voltage (R2=0.96852) (b) collector current
(R2=0.96961) and (c) gate-emitter voltage (R2=0.73724)........................ 52
Figure 4.10: Comparison of measured and simulated waveforms that do not
match up well due to fixing some parameters for a IXYS IXEL40N400 IGBT
for (a) collector-emitter voltage (R2=0.95617) (b) collector current
(R2=0.95614) and (c) gate-emitter voltage (R2=0.82994)........................ 53
Figure 4.11: Combined AGD values for three different Powerex QIS4506001
IGBTs versus (a) collector current (b) collector-emitter voltage ................ 62
Figure 4.12: AGD values for an IXYS IXEL40N400 IGBT versus (a) collector
current (b) collector-emitter voltage...................................................... 63
Figure 4.13: Combined MUN values for three different Powerex QIS4506001
IGBTs versus collector current showing (a) all values (b) zoomed into smaller
values ............................................................................................... 64
Figure 4.14: Combined MUN values for three different Powerex QIS4506001
IGBTs versus collector-emitter voltage .................................................. 65
Figure 4.15: MUN values for an IXYS IXEL40N400 IGBT versus collector
current showing (a) all values (b) zoomed into smaller values .................. 66
Figure 4.16: MUN values for an IXYS IXEL40N400 IGBT versus collectoremitter voltage ................................................................................... 67
Figure 4.17: Combined CGS values for three different Powerex QIS4506001
IGBTs versus (a) collector current (b) collector-emitter voltage ................ 69
Figure 4.18: CGS values for an IXYS IXEL40N400 IGBT versus (a) collector
current (b) collector-emitter voltage...................................................... 70
Figure 4.19: Combined MUP values for three different Powerex QIS4506001
IGBTs versus (a) collector current (b) collector-emitter voltage ................ 71
Figure 4.20: MUP values for an IXYS IXEL40N400 IGBT versus (a) collector
current (b) collector-emitter voltage...................................................... 72
Figure 4.21: Combined COXD values for three different Powerex
QIS4506001 IGBTs versus collector current showing (a) all values (b)
zoomed into smaller values .................................................................. 74
Figure 4.22: Combined COXD values for three different Powerex
QIS4506001 IGBTs versus collector-emitter voltage ................................ 75
Figure 4.23: COXD values for an IXYS IXEL40N400 IGBT versus collector
current showing (a) all values (b) zoomed into smaller values .................. 76
Figure 4.24: COXD values for an IXYS IXEL40N400 IGBT versus collectoremitter voltage ................................................................................... 77
Figure 4.25: Combined KP values for three different Powerex QIS4506001
IGBTs versus (a) collector current (b) collector-emitter voltage ................ 80
vi
Figure 4.26: KP values for an IXYS IXEL40N400 IGBT versus (a) collector
current (b) collector-emitter voltage...................................................... 81
Figure C.1: IGBT test circuit schematic that correlates directly to the printed
circuit board layout ........................................................................... 116
Figure C.2: Gate drive circuit schematic that correlates directly to the
printed circuit board layout ................................................................ 117
Figure C.3: Circuit board layout shown with input capacitor that extends off
the board......................................................................................... 118
Figure C.4: PCB top copper layer and silkscreen .................................. 119
Figure C.5: PCB bottom copper layer and silkscreen ............................. 120
vii
List of Tables
Table 2.1: Micro-Cap 9 IGBT Modeling Parameters ................................. 11
Table 3.1: Test matrix for collecting data on the IGBTs........................... 25
Table 4.1: Parameters held constant for both IGBTs during least-squares
curve fitting ....................................................................................... 45
Table 4.2: IXYS IXLF19N250A parameter values in Micro-Cap 9 compared to
those determined at VCE=3 kV with a 3 Ω load for the Powerex QIS4506001
and IXYS IXEL40N400 IGBTs................................................................ 47
Table 4.3: Linear regression results versus collector current for the Powerex
QIS4506001 IGBT #1 parameters that were fit using least squares method56
Table 4.4: Linear regression results versus collector current for the Powerex
QIS4506001 IGBT #2 parameters that were fit using least squares method57
Table 4.5: Linear regression results versus collector current for the Powerex
QIS4506001 IGBT #3 parameters that were fit using least squares method58
Table 4.6: Linear regression results versus collector current for three
different Powerex QIS4506001 IGBTs combining parameters that were fit
using least squares method.................................................................. 59
Table 4.7: Linear regression results versus collector current for the IXYS
IXEL40N400 IGBT parameters that were fit using least squares method ... 60
Table B.1: Bill of materials for IGBT test circuit and gate drive circuit ..... 115
Table D.1: Powerex IGBT #1 raw data table........................................ 122
Table D.2: Powerex IGBT #1 raw data table (continued)....................... 123
Table D.3: Powerex IGBT #1 raw data table (continued)....................... 124
Table D.4: Powerex IGBT #1 raw data table (continued)....................... 125
Table D.5: Powerex IGBT #2 raw data table........................................ 126
Table D.6: Powerex IGBT #2 raw data table (continued)....................... 127
Table D.7: Powerex IGBT #2 raw data table (continued)....................... 128
Table D.8: Powerex IGBT #2 raw data table (continued)....................... 129
Table D.9: Powerex IGBT #3 raw data table........................................ 130
Table D.10: Powerex IGBT #3 raw data table (continued) ..................... 131
Table D.11: Powerex IGBT #3 raw data table (continued) ..................... 132
Table D.12: Powerex IGBT #3 raw data table (continued) ..................... 133
Table D.13: IXYS IGBT raw data table ................................................ 134
Table D.14: IXYS IGBT raw data table (continued) ............................... 135
Table D.15: IXYS IGBT raw data table (continued) ............................... 136
Table D.16: IXYS IGBT raw data table (continued) ............................... 137
Table E.1: Powerex IGBT #1 R2 values ............................................... 139
Table E.2: Powerex IGBT #1 R2 values (continued) .............................. 140
Table E.3: Powerex IGBT #1 R2 values (continued) .............................. 141
viii
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
E.4: Powerex IGBT #1 R2 values (continued) ..............................
E.5: Powerex IGBT #2 R2 values ...............................................
E.6: Powerex IGBT #2 R2 values (continued) ..............................
E.7: Powerex IGBT #2 R2 values (continued) ..............................
E.8: Powerex IGBT #2 R2 values (continued) ..............................
E.9: Powerex IGBT #3 R2 values ...............................................
E.10: Powerex IGBT #3 R2 values (continued) ............................
E.11: Powerex IGBT #3 R2 values (continued) ............................
E.12: Powerex IGBT #3 R2 values (continued) ............................
E.13: IXYS IGBT R2 values........................................................
E.14: IXYS IGBT R2 values (continued).......................................
E.15: IXYS IGBT R2 values (continued).......................................
E.16: IXYS IGBT R2 values (continued).......................................
ix
142
143
144
145
146
147
148
149
150
151
152
153
154
1. Introduction
Solid-state Pulsed Power
As pulsed power systems become increasingly compact, the need for
solid-state pulsed power becomes greater [1-3]. Systems that once took up
the floor space of a warehouse now are being compacted to the size of circuit
boards. This change necessitates a new field of study within pulsed power;
that is solid-state pulsed power.
Solid-state pulsed power systems are being implemented in a variety
of ways. Solid-state power modulators are being used as drivers for various
plasma applications, such as plasma drilling and plasma immersion ion
implantation [4-7].
Other pulsed power applications utilizing solid-state
1
systems include drivers for pulsed lasers and x-ray sources [8],[9]. These
solid-state pulsed power systems have a wide array of uses in addition to the
traditional pulsed power systems, such as compact magnetron power
supplies and combustion gas sensors [10],[11].
Solid-state Switching
Larger pulsed power systems commonly utilize relays or spark-gap
switches as the switching mechanisms [12-15].
However, as the systems
are becoming smaller, the size of traditional pulsed power switches becomes
the limiting factor in the compactness of the system.
Consequently,
semiconductor switches are the method of choice for designing compact,
solid-state pulsed power systems.
However, power semiconductor devices
are typically preferred due to their higher hold-off voltages and higher
current ratings. There are many types of power semiconductor switches that
can be utilized, including bipolar junction transistors (BJT), metal-oxidesemiconductor
field-effect
transistors
(MOSFET),
insulated-gate
bipolar
transistors (IGBT), and thyristors [16-19]. Even simple semiconductors such
as PiN diodes can be utilized in solid-state pulsed power systems.
Power semiconductor switches typically are silicon-based devices that
operate under the same semiconductor principles as their lower rated
counterparts.
This allows these devices to be easily triggered in their
respective systems using the turn-on properties of the respective device.
The high gate impedances of the MOSFET, IGBT, and gate turn-off thyristor
2
(GTO) can make them the switch of choice for most solid-state pulsed power
applications. This allows the low voltage, low current gate side of the device
to mostly be decoupled from the high voltage, high current switching side of
the device. However, despite the direct correlation of gate current to switch
current in the BJT and thyristor, these devices can be used for solid-state
pulsed power systems depending on the application [18],[20],[21].
The typically silicon based power semiconductor switches are also
being designed in silicon carbide (SiC) configurations [17],[22]. This allows
the devices to be more robust and handle higher powers than most silicon
devices. The SiC switches being designed include Schottky diodes, MOSFETs,
and thyristors [23].
However, the conversion of power IGBTs to SiC has
been limited due to the lack of increased performance compared with the SiC
MOSFETs increased performance [24].
IGBT Switching
IGBTs are commonly used as switches in many power electronics
applications. Similarly, their use as switches has also been transferred into
compact,
solid-state,
pulsed-power
applications
that
were
previously
described. While other semiconductor switches can be used at high switching
frequencies, the IGBT is typically used in applications where high voltage and
high current are more important than a high switching frequency [24].
Both continuous and switch-mode applications require the blocking
voltages and high input impedance gates of these devices. Thus, they have
3
become a popular choice for use in applications such as automotive ignitions
[25].
Additionally, the IGBT has been used successfully in motor drivers,
uninterruptible power supplies, and power inverters [26],[27].
Despite their usefulness in other power electronics applications, it is
the IGBT’s usage in solid-state pulsed power systems that is the focus of this
thesis.
There are many solid-state pulsed power applications that are
currently utilizing the IGBT as a switching mechanism [28-33].
Specific
designs such as compact Marx generators and transmission line transformers
are also taking advantage of the IGBT [34],[35].
Pulsed Power IGBT Characterization
While the continuous and switch-mode applications are common for
the IGBT, pulsed power applications have a few significant differences.
Pulsed power applications typically have shorter pulse widths and lower duty
factors than do switch mode applications.
Also, higher peak currents are
seen in pulsed power applications. The IGBT is capable of being used in the
pulsed power setting where these conditions are present as demonstrated in
the previous section.
Due to the differences between the traditional and
pulsed power applications of the IGBT, model parameters determined for the
IGBT under continuous or switch-mode conditions are not accurate under
pulsed power conditions.
Because
simulations,
circuit
the
models
circuit
allow
design
for accurate
process
4
can
be
and efficient
done
more
circuit
rapidly.
Specifically, having an IGBT model that utilizes parameters determined under
pulsed power conditions prevents the designer from having to use alternative
semiconductor switches to approximate the circuit behavior. Though many
IGBT models are provided in circuit simulation packages, many of them are
for low-voltage, low-current power electronics applications.
This thesis
presents a method for determining the modeling parameters necessary for
an IGBT model commonly used in circuit simulation packages.
Thesis Overview
Chapter
2
discusses
different
methods
for
modeling
IGBTs.
Specifically, it focuses on the Oziemkiewicz implementation of the Hefner
IGBT model to explain the seventeen modeling parameters for IGBTs in
Micro-Cap [36],[37].
Chapter 3 covers the setup of the IGBT test circuit.
The overall test system, test circuit design, and gate drive circuit. Chapter 4
covers the experimental and simulation results for both the Powerex
QIS4506001 and the IXYS IXEL40N400 IGBTs [38],[39]. The least squares
curve fitting method used to analyze the data is discussed along with
samples of measured and simulated waveforms.
Finally, the determined
modeling parameter values for each of the IGBTs are presented. Chapter 5
concludes the thesis and suggests possible future work.
Included in the
appendices are a copy of the least squares curve fitting MATLAB code, a bill
of materials for the IGBT test circuit, and the printed circuit board layout that
was used for the experiment [40].
5
2. IGBT Modeling
Due to the IGBT’s use in a variety of types of circuits, the modeling of
these devices has been studied by many people. As such, there are a few
different methods of modeling the behavior of an IGBT in a given circuit.
This chapter will highlight a few of the common IGBT circuit models: the
simple MOSFET/BJT model and the Hefner model.
The Oziemkiewicz
implementation of the Hefner IGBT model will be discussed in detail as it is
used in various circuit simulation software packages, such as Micro-Cap 9
[36].
Specifically, the origin of the seventeen modeling parameters in the
Oziemkiewicz implementation will be explained [37].
There are a few different ways to name the terminals of the IGBT.
Some authors, including Hefner and Oziemkiewicz, use the anode/cathode
nomenclature referring to the typical polarity of the input voltage on the
6
terminals of the device.
Others use the drain/source nomenclature of a
traditional MOSFET, since the source of the internal MOSFET is connected to
a terminal of the IGBT.
However, this thesis will use the collector/emitter
nomenclature as that is how it is specified on many IGBT manufacturers’
datasheets. Unfortunately, it can become cumbersome and confusing when
describing the terminals of an IGBT with respect to its internal semiconductor
equivalents.
Any deviation from the collector/emitter nomenclature in this
thesis will be when referring to the terminals of the equivalent internal
MOSFET and BJT.
BJT/MOSFET Model
The most basic circuit models that exist for IGBTs include only a
MOSFET and a BJT [41],[42]. A schematic for these simple models can be
seen in Figure 2.1. The IGBT’s collector terminal is connected to the emitter
of a pnp BJT, while the collector of that BJT is connected to the emitter
terminal of the IGBT.
The base of the BJT connects to the drain of an n-
channel MOSFET, with the MOSFET’s source connecting to the emitter
terminal of the IGBT. Simulating IGBTs using one of these models works to
determine how a circuit will react under general switching conditions.
However, these models in their simplest form neglect to include other effects
that are involved with having these two devices on the same substrate, such
as non-linear capacitances between terminals [43]. Therefore, these models
do not provide an adequate level of precision that may be necessary for
7
applications where the voltage and current transients may play an important
role in the behavior of the device. The transient voltages and currents that
occur within an IGBT simply cannot be modeled with only a MOSFET and BJT.
Given the highly transient nature of pulsed-power systems, this model would
cause difficulties distinguishing the transient nature of the circuit from that of
the IGBT being utilized.
Although these models can be expanded in their
degree of accuracy by adding external components to model specific
additional effects, one would have to choose which additional effects need to
be modeled beforehand.
There is, therefore, a tradeoff between the
simplicity of the model and the accuracy of the desired results.
Figure 2.1: Simple IGBT model using only a pnp BJT and an n-channel MOSFET
8
Hefner Model
One model of IGBTs was created by Dr. Allen Hefner at the National
Institute of Standards and Technology.
The model includes the inherent
MOSFET and BJT along with other circuit effects.
The configuration of the
MOSFET and BJT can be seen with other inherent equivalent circuit
components in Figure 2.2 [44].
This phenomenological equivalent circuit
allows for a model to be developed based upon general circuit analysis and
material properties.
Hefner’s model was the “…first one-dimensional (1-D) analytical,
charge controlled model suitable for circuit simulator implementation [43].”
Using semiconductor physics along with the known MOSFET and BJT
characteristics and his inherent equivalent circuit, Hefner was able to create
a device model for the IGBT that is widely used [43-48]. Hefner also went
on to study the model further, as well as IGBT drive requirements and
electro-thermal effects [49-55].
Hefner’s IGBT model has been verified not only by himself but also by
other researchers in the field [24],[43],[56-58]. Additionally, this model is
the basis for various implementations of the IGBT into circuit simulation
software [36],[37],[59].
One notable implementation of the Hefner model
was done by Gregory Oziemkiewicz for PSpice [37].
This implementation
along with its associated input parameters are described in the following
section.
9
Figure 2.2: Phenomonological IGBT equivalent circuit [44].
10
Table 2.1: Micro-Cap 9 IGBT Modeling Parameters
Parameter
Description
Units
AGD
Gate-drain overlap area
m2
BVN
Avalanche multiplication exponent
JSNE
Emitter saturation current density
A/cm2
MUN
Electron mobility
cm2/(V·s)
VT
Threshold area
V
AREA
Device area
m2
CGS
Gate-source capacitance per unit area
F/cm2
KF
Triode region factor
MUP
Hole mobility
cm2/(V·s)
TAU
Ambipolar recombination lifetime
s
VTD
Gate-drain overlap depletion threshold
V
BVF
Avalanche uniformity factor
COXD
Gate-drain oxide capacitance per unit area
F/cm2
KP
MOS transconductance
A/V2
NB
Base doping
cm-3
THETA
Transverse field factor
V-1
WB
Metallurgical base width
m
11
Oziemkiewicz Implementation of the Hefner Model
The implementation of the Hefner model by Gregory Oziemkiewicz is
used in many circuit simulation software packages such as Micro-Cap and
PSpice
[37].
This
implementation
allows
for
the
variation
of
the
semiconductor properties and the equivalent circuit components in the
Hefner IGBT model to simulate different IGBT models. The parameters that
can be varied in Micro-Cap version 9.0.2 for an individual IGBT are shown in
Table 2.1 [36].
Each manufactured IGBT should have different nominal values for the
parameters utilized in the Oziemkiewicz implementation. The determination
of these parameter values would allow one to simulate
commercially available IGBT.
any given
Additionally, experimental testing would
determine whether the equivalent circuit parameters vary under pulsed
power conditions.
The following subsections describe the origins of the seventeen input
modeling
parameters
in
the
Oziemkiewicz
implementation.
These
parameters are discussed with relation to both the Oziemkiewicz thesis and
the Micro-Cap 9 variation of the Oziemkiewicz implementation. As previously
mentioned, the parameters of the Oziemkiewicz implementation are defined
using the anode/cathode/gate nomenclature for the IGBT terminals rather
than the collector/emitter/gate nomenclature used in this thesis. However,
each of the input parameters is defined with regards to the internal
equivalent MOSFET and BJT.
Since the commercially available circuit
12
simulation software packages use variations of this terminology, the
parameters will be described with respect to these internal equivalent
components.
Gate-Drain Overlap Area
The gate-drain overlap area is the physical overlap area between the
gate and the drain of the internal MOSFET equivalent. In the Oziemkiewicz
thesis, this parameter is referred to as Agd. In Micro-Cap 9, this parameter is
referred to as AGD. Because the Micro-Cap parameters are those needed to
simulate the device, this parameter will be referred to as AGD within this
thesis if it is abbreviated.
The units of the gate-drain overlap area are
square centimeters (cm2) in the Oziemkiewicz thesis. However, Micro-Cap 9
implements this parameter with units of square meters (m2). Additionally,
this parameter should not be greater than the active device area, AREA, in
the simulations to prevent erroneous results.
Avalanche Multiplication Exponent
The avalanche multiplication exponent is an exponent term in the
equation used to calculate the avalanche multiplication factor. The equation
for the avalanche multiplication factor can be seen in Equation 2.1, where M
is the avalanche multiplication factor, Vds is the drain-source voltage on the
internal MOSFET, BVcbo is the open-emitter collector-base breakdown voltage,
and BVn is the avalanche multiplication exponent.
13
In the Oziemkiewicz
thesis, this parameter is referred to as BVn. In Micro-Cap 9, this parameter
is referred to as BVN. In both cases this is a unitless parameter. The MicroCap variation of BVN will be used in this thesis to denote the avalanche
multiplication exponent.
M=
1
 Vds
1 − 
 BVcbo



BVn
Equation 2.1
Emitter Saturation Current Density
The emitter saturation current density refers to the emitter current
density of the internal equivalent BJT at which the depletion region of that
internal BJT is saturated with carriers.
There is a slight variation between
the Oziemkiewicz thesis and Micro-Cap 9 as to how this parameter is
implemented. In the Oziemkiewicz thesis, this is represented by the emitter
electron saturation current, Isne.
This saturation current has the unit of
Amperes (A). In Micro-Cap 9, this parameter is represented by the emitter
saturation current density, JSNE. The saturation current density has units of
amperes per square centimeter (A/cm2).
The active device area, AREA, is
the area to which the current density is referenced, as that is the area
through which current flows.
14
Electron Mobility
The electron mobility refers to the semiconductor property of the same
name.
This would be dependent on the semiconductor material in the
device, as well as, the operating temperature.
Additionally, the electron
mobility can be a function of the applied electric field, and, thus, the voltage
across the device [60]. An analytical expression derived to fit empirical plots
can be seen in Equation 2.2, where µn is the electron mobility and E is the
electric field in V/cm [60].
referred to as µn.
In the Oziemkiewicz thesis, this parameter is
In Micro-Cap 9, this parameter is referred to as MUN.
Both cases have units of cm2/(V·s).
This thesis will refer to the electron
mobility as MUN.
µn =
1,375
  E 
1 + 
3 
8
×
10

 
2



1
2
Equation 2.2 [60]
Threshold Area
The threshold area is not actually an area in terms of physical
dimensions.
This is actually the MOSFET channel threshold voltage in the
Oziemkiewicz thesis, and is referred to as VT. However, Micro-Cap 9 refers to
this parameter as the threshold area, VT. In both cases, the units are volts
(V). This is the minimum voltage required for channel formation to occur in
the internal MOSFET. This thesis will refer to the threshold area as VT.
15
Device Area
The device area refers to the active area of the device.
This is the
current carrying area of the device. The Oziemkiewicz thesis refers to this
parameter as A, while Micro-Cap 9 refers to the device area as AREA. The
units of the device area are square centimeters (cm2) in the Oziemkiewicz
thesis.
However, Micro-Cap 9 implements this parameter with units of
square meters (m2). This thesis will refer to the active device area as AREA.
Gate-Source Capacitance
The gate-source capacitance is actually a combination of the MOSFET
source metallization capacitance, CM, and the capacitance due to the gate
oxide overlapping the source, Coxs. These two capacitances can be seen in
Figure 2.2 and are combined in parallel.
This equivalent capacitance is
referred to as Cgs in the Oziemkiewicz thesis and as CGS in Micro-Cap 9. The
Oziemkiewicz case has the units of farads (F).
However, Micro-Cap 9
implements this parameter as a capacitance per unit area which has units of
farads per square centimeter (F/cm2). The area is the physical overlap area
of the gate and the source of the internal MOSFET. This thesis will refer to
the gate-source capacitance as CGS.
16
Triode Region Factor
The triode region factor is the triode region MOSFET transconductance
factor.
This parameter is used to calculate the MOSFET channel current
during simulation, as well as, the conductance terms associated with the
MOSFET current (∂Imos/∂Vgs and ∂Imos/∂Vds).
The implementation of the
triode region factor in these calculations can be seen in the Oziemkiewicz
thesis, where the triode region factor is denoted as Kf. In Micro-Cap 9, the
triode region factor is denoted as KF. In both cases, the triode region factor
is a unitless parameter. This thesis will refer to the triode region factor as
KF.
Hole Mobility
The hole mobility refers to the semiconductor property of the same
name.
This would be dependent on the semiconductor material in the
device, as well as, the operating temperature. Additionally, the hole mobility
can be a function of the applied electric field, and, thus, the voltage across
the device [60]. An analytical expression derived to fit empirical plots can be
seen in Equation 2.3, where µp is the hole mobility and E is the electric field
in V/cm [60]. In the Oziemkiewicz thesis, this parameter is referred to as µp.
In Micro-Cap 9, this parameter is referred to as MUP. Both cases have units
of cm2/(V·s). This thesis will refer to the electron mobility as MUP.
17
µp =
487
E


1+ 
4 
1
.
95
×
10


Equation 2.3 [60]
Ambipolar Recombination Lifetime
The ambipolar recombination lifetime is the minority carrier lifetime
within the semiconductor material. This is a semiconductor property that can
be influenced by the material, operating temperature, electron mobility, and
hole mobility [16]. The ambipolar recombination lifetime is referred to in the
Oziemkiewicz thesis as the base high-level lifetime, τHL.
This parameter is
denoted as TAU in Micro-Cap 9. Both cases have the units of seconds (s).
This thesis will refer to the ambipolar recombination lifetime as TAU.
Gate-Drain Overlap Depletion Threshold
The gate-drain overlap depletion threshold is the negative voltage
required to turn off the normally open n-channel from the depletion-mode
MOSFET.
The internal MOSFET is in depletion mode during transient
conditions where a quasi-static approximation for the charge densities in the
region around the gate is not valid [24]. Thus, the threshold voltage is the
negative voltage required to close that depletion-mode channel [61].
This
gate-drain overlap depletion threshold voltage is referred to as VTd in the
18
Oziemkiewicz thesis and as VTD in Micro-Cap 9. Both cases have the units of
volts (V). This thesis will refer to the gate-drain overlap depletion threshold
as VTD.
Avalanche Uniformity Factor
The avalanche uniformity factor is a term used to calculate the openemitter collector-base breakdown voltage, BVcbo.
BVcbo is then used to
calculate avalanche multiplication factor along with BVn and Vds in Equation
2.1, where BVn is the avalanche multiplication exponent and Vds is the
internal
MOSFETs
drain-source
voltage.
BVcbo is
calculated
in
the
Oziemkiewicz thesis as shown in Equation 2.4, where BVf is the avalanche
uniformity factor and Nscl is the collector-base space charge concentration.
In Micro-Cap 9, the avalanche uniformity factor is denoted as BVF. In both
the Oziemkiewicz thesis and Micro-Cap 9, the avalanche uniformity factor is
unitless. This thesis will refer to the avalanche uniformity factor as BVF.
BVcbo =
(
BV f ⋅ 5.34 × 1013
0.75
N scl
Equation 2.4
19
)
Gate-Drain Oxide Capacitance
The gate-drain oxide capacitance is the capacitance formed by the
overlap of the gate oxide with the internal MOSFET’s drain.
In the
Oziemkiewicz thesis, the gate-drain oxide capacitance is referred to as Coxd,
while, in Micro-Cap 9, it is referred to as COXD. The Oziemkiewicz case has
the units of farads (F). However, Micro-Cap 9 implements this parameter as
a capacitance per unit area which has units of farads per square centimeter
(F/cm2).
The area is the physical overlap area of the gate oxide and the
drain of the internal MOSFET. This thesis will refer to the gate-drain oxide
capacitance as COXD.
MOS Transconductance
The MOS transconductance is the referred to as the MOSFET
transconductance
parameter
in
the
Oziemkiewicz
thesis.
The
transconductance is the gradient of the transfer characteristic at a given
temperature [24].
MOSFET.
overall
In this case, it is the transconductance of the internal
Despite the MOS transconductance playing a large part in the
IGBT
transconductance,
the
MOS
transconductance
does
not
completely account for the overall IGBT transconductance. The Oziemkiewicz
thesis refers to the MOS transconductance as Kp, while Micro-Cap 9 refers to
the parameter as KP.
Both cases have the units of A/V2.
refer to the MOS transconductance as KP.
20
This thesis will
Base Doping
The base doping refers to the concentration of the lightly-doped, ntype, drift region in the IGBT that corresponds to the base of the internal BJT
as shown in Figure 2.2. The Oziemkiewicz thesis refers to this parameter as
NB. Micro-Cap 9 refers to the base doping as NB. Both cases have the units
of inverse centimeters cubed (cm-3). This thesis will refer to the base doping
as NB.
Transverse Field Factor
The transverse field factor is referred to as the transverse field
transconductance factor in the Oziemkiewicz thesis. This parameter is used
in the semi-empirical formula to calculate the internal MOSFET channel
current during simulation accounting for the mobility reduction due to the
transverse electric field for high gate voltages [37]. Additionally, it is used to
calculate the conductance terms associated with the MOSFET current
(∂Imos/∂Vgs and ∂Imos/∂Vds). The implementation of the transverse field factor
in these calculations can be seen in the Oziemkiewicz thesis, where the
transverse field factor is denoted as θ. Micro-Cap 9 refers to the transverse
field factor as THETA. In both cases, the transverse field factor has units of
inverse volts (V-1).
This thesis will refer to the transverse field factor as
THETA.
21
Metallurgical Base Width
The metallurgical base width is the physically deposited width of the
base of the internal BJT.
This is used to calculate the quasi-neutral base
width by subtracting away the base-collector depletion width from the
metallurgical base width [37].
The Oziemkiewicz thesis refers to the
metallurgical base width as WB, while Micro-Cap 9 refers to it as WB. The
units
of
the
metallurgical
base
width
are
centimeters
(cm)
in
the
Oziemkiewicz thesis. However, Micro-Cap 9 implements this parameter with
units of meters (m). This thesis will refer to the metallurgical base width as
WB.
22
3. Experiment Setup
Overview
To determine the required IGBT parameters, a test circuit had to be
constructed for gathering experimental data on the IGBTs. Specifically, the
Powerex QIS4506001 and the IXYS IXEL40N400 are the IGBTs being tested
[38],[39]. Photos of these devices are shown in Figure 3.1 and Figure 3.2.
By testing each of these IGBTs under varying conditions, the parameters can
be determined for each of those conditions. This allows for the comparison
of the parameters to determine trends as the voltage and current on the
device are varied. Table 3.1 shows the test matrix that was used to test the
IGBT. Four resistive loads and collector-emitter voltages were tested for a
single 24 V gate-emitter voltage with a 10 µs pulsewidth.
23
(a)
(b)
Figure 3.1: Photograph of Powerex QIS4506001 IGBT (a) top view (b) bottom
view
(a)
(b)
Figure 3.2: Photograph of IXYS IXEL40N400 IGBT (a) top view (b) bottom view
24
Table 3.1: Test matrix for collecting data on the IGBTs.
Load [Ω]
Collector-Emitter
Voltage (Vce) [kV]
Gate-Emitter
Voltage (Vge) [V]
Pulse Width [µs]
15
1
24
10
9
2
---
---
4.5
3
---
---
3
3.5
---
---
The system diagram for the IGBT test circuit and its associated
components can be seen in Figure 3.3. A signal generator is used to create a
5 V pulse of a desired pulse width. This pulse is a -5 V to 0 V inverted pulse
because there is an inverting stage later in the gate drive circuitry.
This
inversion will be described later in the “Gate Drive Circuit” subsection in this
chapter. The 5 V pulse then triggers an electrical-to-optical converter that
connects via fiber optic to the gate drive circuit.
The electrical to optical
conversion allows the pulse to trigger the gate drive circuit without the
possibility of an accidental short to the signal generator. An external +5 VDC
power supply is needed to power the electrical-to-optical converter.
An optical-to-electrical converter is present at the input of the gate
drive circuit to provide the trigger signal to the integrated circuit (IC) gate
driver. Once the gate drive circuit is triggered, a +24 V signal is output for
the duration of the pulse width with an open or resistive load.
There is a
slight difference when the gate drive circuit is connected to the IGBT due to
the internal capacitances that will be discussed later in “IGBT Test Circuit”
subsection. Similarly to the electrical-to-optical converter, the gate drive
25
Signal
Generator
+5 VDC
+24 VDC
+HVDC
Electricalto-Optical
Converter
Gate Drive
IGBT Test
Circuit
Circuit
Diagnostics
Figure 3.3: Experimental setup system diagram.
circuitry requires an external +24 VDC power supply.
The signal from the
gate drive circuit triggers the IGBT in the IGBT test circuit. This circuit uses
the IGBT that is being tested to switch a DC voltage across a resistive load.
The DC voltage being switched across the load comes from Glassman highvoltage DC power supply in parallel with a 30 µF input capacitor. To measure
key voltages and currents within the IGBT, various diagnostics are utilized
that will be discussed in the “IGBT Test Circuit” subsection.
IGBT Test Circuit
The IGBT test circuit that has been constructed at the University of
Missouri is shown in Figure 3.4.
A photograph of the IGBT test circuit is
shown in Figure 3.5. The Glassman high-voltage DC power supply charges
the 30 µF input capacitor via a series of three protection diodes. The
protection diodes prevent any possible voltage reversal from sending current
26
Figure 3.4: IGBT test circuit schematic.
Figure 3.5: Photograph of the IGBT test circuit attached to the input capacitor.
27
back into the DC power supply. Three diodes are needed in series due to the
insufficient voltage rating of each diode.
When the IGBT closes, the capacitor discharges through a varying
equivalent resistive load that is comprised of fifteen resistors in series. The
fifteen resistors in series allow for higher power resistors to be used as well
as distribute the load power consumption among multiple load resistors. As
previously shown in Table 3.1, the resistive load varies from 3 Ω to 15 Ω
while the DC input voltages vary from 1 kV to 3.5 kV. A 2 MΩ bleed resistor
is in parallel with the input capacitor to dissipate any excess charge that
remains in the capacitor after the IGBT is triggered. This bleed resistor acts
as a safety feature to “bleed” the charge from the capacitor if it is not fully
discharged at the end of the experiment.
When the signal generator is used to trigger the gate drive circuitry, as
previously discussed, a +24 V pulse is sent to the gate of the IGBT.
However, there is gate impedance in the test circuit comprised of both the
internal gate impedance of the device and the damping resistance of the gate
drive circuitry shown in Figure 3.7.
The gate impedance of the IGBT is
frequency dependent due to the gate capacitance. At turn-on, the damping
resistance from the gate drive circuitry to the gate of the IGBT is 50 Ω.
However, at turn-off, the damping resistance is shorted via a diode to
decrease turn-off time. This shorted damping resistance led to ringing at the
gate during turn-off. This internal gate capacitance slows the rise time of the
gate signal as the charge time of an RC circuit. Consequently, if the pulse
28
width of the gate signal is less than the rise time of the signal, the gate
signal will be truncated before it reaches its maximum value.
The gate drive circuitry is connected from the gate to the emitter of
the IGBT via a 2 mΩ emitter-feedback resistor.
This emitter-feedback
resistor limits the peak current through the circuit in the event of a fault.
Using Kirchhoff’s Voltage Law around the loop between the gate and the
emitter, Equation 3.1 is found.
RE is the nominal value of the emitter-
feedback resistor, VG is the input gate voltage, VGE is the voltage drop from
the gate to the emitter, and IE,max is the maximum current allowed through
the device.
However, the gate-emitter voltage depends on a nonlinear
internal gate capacitance.
Rearranging Equation 3.1 and using a 2 mΩ
emitter feedback resistor, the maximum collector current ranges from 250 A
to 1,600 A for values of (VG – VGE) at 0.5 V and 3.2 V, respectively.
RE =
VG − VGE
I E ,max
Equation 3.1
Because a first generation printed circuit board (PCB) is being used,
stray inductance is present. This is represented with Lstray1 in the collector
current path and Lstray2 in the gate current path. These are determined using
the same method that is used to determine the seventeen IGBT modeling
29
parameters.
To prevent voltage reversals due to this inductance and
damage to the IGBT, an antiparallel diode has been place in the circuit with
the IGBT.
Voltage transients could also cause difficulty measuring the
collector-emitter voltage and possibly damage the device.
Thus, an RCD
snubber has been placed in parallel with the IGBT. The snubber consists of a
1 kΩ resistor in parallel with a diode where the parallel combination of the
two is in series with an 11.75 nF capacitor.
Diagnostics
Multiple diagnostic measurements need to be taken to adequately
measure the necessary voltages and currents. Despite the need for multiple
measurements, however, there are only a few types of diagnostics that were
used to take the measurements.
A Tektronix P6015A probe was used to
measure the collector-ground voltage, and a Tektronix P2220 10x probe was
used to measure the emitter-ground voltage.
Thus, a differential voltage
measurement was used to measure the collector-emitter voltage. This was
necessary due to the emitter-feedback resistor between the emitter and
ground. A Tektronix P2220 10x probe was also used to measure the gateground voltage. Similarly to the collector-emitter voltage, the gate-emitter
voltage was also a differential measurement.
measure the necessary currents.
Pearson coils were used to
A model 410 Pearson coil was used to
measure the collector current, while a model 2877 Pearson coil was used to
measure the gate current. The gate current should integrate to zero. As a
check, the sample current was found to integrate to -1.84 x 10-18 which is
30
Vce
Ic
Vce (kV)
180
130
80
Ic (A)
2.2
2
1.8
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
-0.2
30
-20
0
5
10
15
T ime (µs)
20
3
2.5
2
1.5
1
0.5
0
-0.5
-1
-1.5
-2
-2.5
-3
Vge (V)
30
25
20
15
10
5
0
-5
-10
-15
-20
-25
-30
Vge
0
5
Ig
10
15
T ime (µs)
Ig (A)
(a)
20
(b)
Figure 3.6: Plots for a 2 kV input voltage and a 15 Ω resistive load with a 24 V gate
signal on a Powerex IGBT. (a) IGBT collector-emitter voltage and collector current.
(b) IGBT gate-emitter voltage and gate current.
31
essentially equal to zero.
Sample measurements for a 2 kV input voltage
with a 15 Ω load and 24 V gate signal are shown in Figure 3.6.
Gate Drive Circuit
Because the optical-to-electrical converter is utilized to isolate the
signal generator, the opposite conversion does not provide enough current to
adequately trigger the IGBT. Thus, an IC gate driver is utilized to provide
the necessary current. The Micrel MIC4452 gate driver was used. However,
the optical-to-electrical converter is inverting, and the Micrel MIC4452 is a
non-inverting gate driver [62].
This was solved with the capability of the
signal generator to output an inverted signal, and the non-inverting gate
driver was left in place. The gate driver circuit provides the necessary bias
voltages for the optical-to-electrical converter and the IC gate driver to
operate.
A +24 V DC voltage is input to the gate drive circuit and then
stepped down with voltage regulators to values that each device requires. A
circuit schematic of the gate drive circuitry can be seen in Figure 3.7.
Initially, the gate drive circuit was designed to be on the underside of
the same PCB as the IGBT test circuit.
This, however, caused capacitive
coupling between the gate drive circuitry and the high voltage circuitry. This
capacitive coupling prevented the gate signal from predictably triggering the
IGBT. Consequently, the gate drive circuit was soldered to a separate PCB
and connected to the other using alligator clip connectors.
This wire
connection between the two circuit boards is what produces most of the stray
32
inductance between the gate driver and the gate of the IGBT which is
accounted for by using Lstray2 in the simulations. A coaxial transmission line
connection between the boards would have decreased the stray inductance.
However, because this was an unforeseen design change for the PCB, it was
not easily implemented.
33
Figure 3.7: IGBT gate drive circuit schematic.
34
4. Experiment Results
Least Squares Curve Fitting
A least squares curve fitting technique is used to actually determine
the seventeen modeling parameters for the IGBT.
The IGBT test circuit is
simulated in Micro-Cap using estimates for the IGBT modeling parameters
[36]. After the simulation is complete, a MATLAB code compares each of the
simulated data points to the measured values for a given input voltage and
load [40].
The error in the algorithm was calculated using the sum of
squares of the residuals between the simulated and measured values. Using
an algorithm called particle swarm analysis, the parameters are changed to
try to minimize the error [63]. The coefficient of determination, R2, was also
calculated using Equation 4.1 and Equation 4.2 where SSresiduals is the sum of
35
SS residuals
SStotal
R2 = 1−
Equation 4.1
SS total = ∑ ( y i − y )
2
i
Equation 4.2
squares of the residuals and SStotal is the sum of squares of the measured
values minus the mean of all measured values. These R2 values can be seen
in Appendix E.
Although some of the values for R2 in Appendix E are
negative, these values correspond to simulations where the sum of squares
of the error was very large due to the device not turning off in the
simulation. Once the error term falls below a certain level or the simulation
is found to be visually acceptable compared with the measured values, those
parameter values are determined to be the effective parameters for that load
and input voltage.
Sample comparisons of simulations and measured data
can be seen in Figure 4.1.
It is important to note, however, that these parameters may not be
the true values for the IGBT being tested. The effective values are those that
best match a simulation result to the measured result. For that reason, these
effective parameters are applicable only to the device that has been tested at
the given voltage and current values.
During this process, the values for
Lstray1 and Lstray2 were also determined to be 400 nH and 2.5 µH, respectively.
36
Additionally, the load resistance is swept over its tolerances to best fit the
measured data.
IGBT Results
Three Powerex QIS4506001 IGBTs were tested according to the test
matrix shown in Table 3.1. Each point in the matrix had a sample size, n, of
ten. This section will briefly discuss the measured results from the IGBT test
circuit and then discuss the least squares curve fitting results.
Finally, a
linear regression was performed on the curve fitting results to determine
whether the parameters stayed constant under the pulsed-power conditions.
One IXYS IXEL40N400 IGBT was also tested in the same manner with the
same sample size of ten shots at each data point.
Test Data
Using the IGBT test circuit described in the previous chapter, data
were taken on each of three Powerex IGBTs and one IXYS IGBT.
A
comparison of the measured waveforms for the three Powerex IGBTs can be
seen in Figure 4.2 for a 1 kV collector-emitter voltage with a 15 Ω load, while
Figure 4.3 shows a set of typical IXYS waveforms. Each of the waveforms is
approximately the same for each of the three different IGBTs with the
37
Vge (V)
30
25
20
15
10
5
0
-5
-10
-15
Simulated
Measured
0
5
10
15
T ime (µs)
20
25
(a)
Vce (kV)
Measured Voltage
Measured Current
350
3.5
300
3
250
2.5
200
2
150
1.5
100
1
Ic (A)
Simulated Voltage
Simulated Current
4
50
0.5
0
0
-50
0
5
10
15
T ime (µs)
20
25
(b)
Figure 4.1: Comparison of measured and simulated plots for a 3.5 kV input voltage
with 15 Ω resistive load and a 24 V gate signal. (a) IGBT gate voltage, VGE. (b) IGBT
collector-emitter voltage, VCE, and collector current, IC.
38
exception of some slight discrepancies in the rise and fall times. Figure 4.2
also provides an example of a typical shot taken on each of the IGBTs.
The
collector-emitter voltage shown in Figure 4.2a falls to its on-state voltage
and then rises back toward the initial charge voltage. The collector current
rises from zero to reach its on-state value that is determined by the load of
the circuit as shown in Figure 4.2b. The gate-emitter voltage in Figure 4.2c
rises to approximately 24 V.
Voltage undershoot, and then overshoot,
causes the device to turn on again briefly at the end of the signal.
This
ringing at turn-off is due to the diode-shorted damping resistance from the
gate drive circuitry. This ringing could have been reduced with a small nonzero damping resistance in series with the diode. The effects of this turn-on
can be seen in all of the waveforms.
Although Figure 4.2 shows an example of a typical Powerex shot for
most of the test matrix, there was a slightly different effect for the higher
currents that is shown in Figure 4.4.
The collector current reached its
saturation level for the applied gate-emitter voltage. The current appeared
to reach saturation at approximately 700 A, which caused the on-state
collector-emitter voltage from reaching its usual on-state value.
There is
also a similar effect for the IXYS IGBT that occurs at approximately 350 A.
This current limit initially appeared to be caused by the emitter-feedback
resistor being too large. However, after testing the IXYS IXEL40N400 IGBT,
this is likely not the case. The IXYS IGBT showed a similar effect at a much
lower current of approximately 350 A which would indicate that current
saturation level is being reached for each IGBT. The collector current as a
39
1200
Collector-Emitter Voltage (V)
1000
800
600
400
200
0
0
5
10
15
20
25
20
25
20
25
Time (µs)
IGBT 1
IGBT 2
IGBT 3
(a)
80
70
Collector Current (A)
60
50
40
30
20
10
0
0
5
10
15
-10
Time (µs)
IGBT 1
IGBT 2
IGBT 3
(b)
30
25
Gate-Emitter Voltage (V)
20
15
10
5
0
0
5
10
15
-5
-10
-15
-20
Time (µs)
IGBT 1
IGBT 2
IGBT 3
(c)
Figure 4.2:
Comparison of measured waveforms among 3 different Powerex
QIS4506001 IGBTs for (a) collector-emitter voltage (b) collector current and (c)
gate-emitter voltage
40
1200
Collector-Emitter Voltage (V)
1000
800
600
400
200
0
0
5
10
15
20
25
Time (µs)
(a)
80
70
Collector Current (A)
60
50
40
30
20
10
0
0
5
10
15
20
25
15
20
25
-10
Time (µs)
(b)
30
Gate-Emitter Voltage (V)
25
20
15
10
5
0
0
5
10
-5
Time (µs)
(c)
Figure 4.3: Example of measured waveforms on the IXYS IXEL40N400 with a 15 Ω
load and 1 kVin for (a) collector-emitter voltage (b) collector current (c) gateemitter voltage
41
4000
3500
Collector-Emitter Voltage (V)
3000
2500
2000
1500
1000
500
0
0
5
10
15
20
25
-500
Time (µs)
Vce=1 kV
Vce=2 kV
Vce=3 kV
Vce=3.5 kV
(a)
800
700
600
Collector Current (A)
500
400
300
200
100
0
0
5
10
15
20
25
-100
Time (µs)
Vce=1 kV
Vce=2 kV
Vce=3 kV
Vce=3.5 kV
(b)
Figure 4.4: Display of the apparent current limit for a Powerex QIS4506001 IGBT
with a 3 Ω load, showing (a) collector-emitter voltage and (b) collector current
42
4000
3500
Collector-Emitter Voltage (V)
3000
2500
2000
1500
1000
500
0
0
5
10
15
20
25
20
25
-500
Time (µs)
Vce=1 kV
Vce=2 kV
Vce=3 kV
Vce=3.5 kV
(a)
400
350
300
Collector Current (A)
250
200
150
100
50
0
0
5
10
15
-50
Time (µs)
Vce=1 kV
Vce=2 kV
Vce=3 kV
Vce=3.5 kV
(b)
Figure 4.5: Display of the apparent current limit for an IXYS IXEL40N40 IGBT with a
3 Ω load, showing (a) collector-emitter voltage and (b) collector current
43
function of gate-emitter voltage is shown in Equation 4.3, where gm is the
transconductance of the IGBT and Vth is the IGBT turn-on threshold voltage.
Curve Fitting
Once the measured data had been collected, the waveforms were used
as a reference to fit simulated waveforms using the least-squares curve
fitting method described at the beginning of this chapter.
Initially, all
seventeen parameters plus the stray inductances and load resistor tolerance
were fit using this method. However, fitting all of these parameters led to
impractically long simulation times.
Consequently, all but six of the
parameters were held constant.
It was determined that the gate-drain overlap area (AGD), electron
mobility (MUN), gate-source capacitance per unit area (CGS), hole mobility
(MUP), gate-drain oxide capacitance per unit area (COXD), and the MOS
transconductance (KP) were likely to be the most important parameters to
fit.
The carrier mobilities were chosen due to their dependence upon the
electric field in the device [60].
The other four parameters were chosen
because they were likely to affect or be affected by the collector current path
I C = g m (VGE − Vth )
Equation 4.3
44
Table 4.1: Parameters held constant for both IGBTs during least-squares curve
fitting
Parameter
Constant Values
Units
BVN
25
JSNE
4 x 10-11
A/cm2
VT
12
V
AREA
3.5 x 10-5
m2
KF
300
TAU
2 x 10-5
s
VTD
-22
V
BVF
4.5 x 104
NB
2 x 1015
cm-3
THETA
0.5
V-1
WB
5.0 x 10-4
m
within the device. AGD may vary based upon the size of the channel, while
COXD would be affected by AGD. CGS would be affected by the change in
AGD and the charge built up during the pulse. KP depends on COXD and the
size of the inversion layer channel that is formed. These constants can be
seen in Table 4.1 and were determined from the limited results of fitting all
seventeen parameters.
Because neither the Powerex QIS4506001 nor the IXYS IXEL40N400
IGBT was in the library for Micro-Cap 9, another IGBT’s parameters were
45
used as a starting point for the particle swarm algorithm. The IGBT that was
chosen was the IXYS IXLF19N250A that is rated for a collector-emitter
voltage of 2,500 V and a collector current of 32 A. That was the model of
IGBT with ratings closest to those IGBTs that were tested. The parameter
values for the IXYS IXLF19N250A can be seen in Table 4.2 compared to
determined parameter values at a collector-emitter voltage of 3 kV with a 3
Ω load for both the Powerex QIS4506001 and IXYS IXEL40N400.
The
parameters that were varied for this thesis are highlighted. Based on Table
4.2, the next parameter that would be interesting to vary would be the triode
region factor, KF, due to its large discrepancy between the IXLF19N250A and
the two IGBTs that were tested.
Figure 4.6 shows the circuit schematic equivalent of the netlist that
was used for simulating the circuit. An example where the simulated results
Figure 4.6: Micro-Cap 9 circuit schematic used for curve fitting
46
Table 4.2: IXYS IXLF19N250A parameter values in Micro-Cap 9 compared to those
determined at VCE=3 kV with a 3 Ω load for the Powerex QIS4506001 and IXYS
IXEL40N400 IGBTs
Parameter
IXLF19N250A
Value
QIS4506001
Value
IXEL40N400
Value
Units
AGD
6.4 x 10-6
7.23 x 10-6
8.43 x 10-6
m2
BVN
4
25
25
JSNE
6.5 x 10-13
4.0 x 10-11
4.0 x 10-11
A/cm2
MUN
1.0 x 105
1.03 x 106
1.55 x 109
cm2/(V—s)
VT
8.0678
12
12
V
AREA
1.6 x 10-5
3.5 x 10-5
3.5 x 10-5
m2
CGS
1.85 x 10-8
1.48 x 10-8
3.07 x 10-8
F/cm2
KF
0.72368
300
300
MUP
20
4.86 x 104
2.2 x 104
cm2/(V—s)
TAU
5.0148 x 10-8
2.0 x 10-5
2.0 x 10-5
s
VTD
-5
-22
-22
V
BVF
9.999
4.5 x 104
4.5 x 104
COXD
5.6429 x 10-7
6.58 x 10-4
1.79 x 10-1
F/cm2
KP
5.5219
8.83
14.6
A/V2
NB
2.0 x 1014
2.0 x 1015
2.0 x 1015
cm-3
THETA
0.02
0.5
0.5
V-1
WB
1.17 x 10-4
5.0 x 10-4
5.0 x 10-4
m
47
3500
3000
Collector-Emitter Voltage (V)
2500
2000
1500
1000
500
0
0
5
10
15
20
25
20
25
-500
Time (µs)
Measured
Simulated
(a)
350
300
Collector Current (A)
250
200
150
100
50
0
0
5
10
15
-50
Time (µs)
Measured
Simulated
(b)
40
30
Gate-Emitter Voltage (V)
20
10
0
0
5
10
15
20
25
-10
-20
-30
Time (µs)
Measured
Simulated
(c)
Figure 4.7: Comparison of measured and simulated waveforms that match up well
for a Powerex QIS4506001 IGBT for (a) collector-emitter voltage (R2=0.97572) (b)
collector current (R2=0.97623) and (c) gate-emitter voltage (R2=0.59535)
48
3500
3000
Collector-Emitter Voltage (V)
2500
2000
1500
1000
500
0
0
5
10
15
20
25
20
25
20
25
Time (µs)
Measured
Simulated
(a)
350
300
Collector Current (A)
250
200
150
100
50
0
0
5
10
15
-50
Time (µs)
Measured
Simulated
(b)
50
40
Gate-Emitter Voltage (V)
30
20
10
0
0
5
10
15
-10
-20
Time (µs)
Measured
Simulated
(c)
Figure 4.8: Comparison of measured and simulated waveforms that match up well
for a IXYS IXEL40N400 IGBT for (a) collector-emitter voltage (R2=0.96877) (b)
collector current (R2=0.96372) and (c) gate-emitter voltage (R2=0.89290)
49
matched the measured Powerex and IXYS results well can be seen in Figure
4.7 and Figure 4.8, respectively. The simulated collector-emitter voltage and
collector current waveforms match the measured waveforms. The simulated
gate-emitter voltage never matched its measured counterpart well. This may
be due to an omitted diode that provides no damping resistance at the gate
during turn-off. However, this is a difficult trace to curve fit due to the nonlinear capacitances within the device.
The gate-source capacitance at turn-on should theoretically be a four
part piecewise function as shown in Equation 4.4, but
the
Oziemkiewicz
implementation of the Hefner model utilized in Micro-Cap simplifies this a
constant value [24],[37]. Equation 4.4 varies with the gate-source voltage
of the internal MOSFET as a function of time throughout the pulse, where
td(on) is the time delay to turn-on, tr(on) is the rise time, td(off) is the time delay
to turn-off, and tf(off) is the fall time. The gate-source voltage of the internal
MOSFET is equal to the gate-emitter voltage for the IGBT. This discrepancy
for the gate-source capacitance did not allow the least squares curve fitting
technique to accurately find effective values to match the gate-emitter
voltage.
However, this method provided a close approximation of the
voltage at the gate of the device.
Since most of the parameters are being held constant, it was
impossible to get a good fit over the full range of testing. An example of this
can be seen in Figure 4.9 and Figure 4.10 for the Powerex and IXYS IGBTs,
respectively. The simulated on-state values for the collector-emitter voltage
50
CGS
td ( on )


 VGS (max) − VGS (0) 


RG ln
V


GS (max) − VGS (t d ( on ) ) 


tr ( on ) − td ( on )


 VGS (max) − VGS (td ( on ) ) 

 RG ln


(
)
−
−
V
V
t
t

GS (max)
GS r ( on )
d ( on ) 

=
td ( off ) − tr ( on )




VGS (max) − VGS (tr ( on ) )

 RG ln


 VGS (max) − VGS (td ( off ) − tr ( on ) ) 

t f ( off ) − td ( off )



VGS (max) − VGS (td ( off ) )



 RG ln V

 GS (max) − VGS (t f ( off ) − td ( off ) ) 

0 ≤ t < td ( on )
td ( on ) ≤ t < tr ( on )
tr ( on ) ≤ t < td ( off )
td ( off ) ≤ t < t f ( off )
Equation 4.4 [24]
and the collector current never reach the measured values. Again, the gateemitter voltage does not match perfectly. This is quantified with the lower R2
values for each plot as shown in Figure 4.7 - Figure 4.10.
Linear Regression Analysis
Curve fitting of the points in the test matrix allowed parameter values
to be determined for each of the ten samples at each point for the three
IGBTs tested.
Using these values it was possible to perform a linear
regression on all of the data to determine whether the parameters were
constant as a function of voltage and/or current.
51
If a parameter is found to
1200
1000
Collector-Emitter Voltage (V)
800
600
400
200
0
0
5
10
15
20
25
-200
Time (µs)
Measured
Simulated
(a)
80
70
60
Collector Current (A)
50
40
30
20
10
0
0
5
10
15
20
25
20
25
-10
Time (µs)
Measured
Simulated
(b)
30
20
Gate-Emitter Voltage (V)
10
0
0
5
10
15
-10
-20
-30
Time (µs)
Measured
Simulated
(c)
Figure 4.9: Comparison of measured and simulated waveforms that do not match up
well due to fixing some parameters for a Powerex QIS4506001 IGBT for (a)
collector-emitter voltage (R2=0.96852) (b) collector current (R2=0.96961) and (c)
gate-emitter voltage (R2=0.73724)
52
1200
1000
Collector-Emitter Voltage (V)
800
600
400
200
0
0
5
10
15
20
25
20
25
-200
Time (µs)
Measured
Simulated
(a)
70
60
Collector Current (A)
50
40
30
20
10
0
0
5
10
15
-10
Time (µs)
Measured
Simulated
(b)
30
25
Gate-Emitter Voltage (V)
20
15
10
5
0
0
5
10
15
20
25
-5
Time (µs)
Measured
Simulated
(c)
Figure 4.10: Comparison of measured and simulated waveforms that do not match
up well due to fixing some parameters for a IXYS IXEL40N400 IGBT for (a) collectoremitter voltage (R2=0.95617) (b) collector current (R2=0.95614) and (c) gateemitter voltage (R2=0.82994)
53
not be constant, then the Oziemkiewicz implementation of the Hefner model
would only be valid for a given point where the parameters have been
determined experimentally.
The results of this regression analysis versus collector current for each
of the three tested IGBTs can be seen in Table 4.3 through Table 4.5. The
results when data from all three IGBTs were combined are shown in Table
4.6. Linear regression analysis is only valid when the data set has normal
distribution (i.e. the plot of the residuals is linear). Based upon the plot of
the residuals, MUN and COXD did not have normal distributions with respect
to collector current.
Therefore, the natural log of the values was taken to
normalize the data and make the linear regression analysis valid.
If the
natural log of the parameter has significance, then the parameter itself will
also have significance. The results of the analysis for the natural log of these
parameters are shown in the tables.
The p-value of slope for the linear
regression was used to determine whether the parameter was constant. Any
parameter found to have a p-value less than 0.05, has a slope when plotted
over the collector current.
0.05 is not constant.
Thus, any parameter with a p-value less than
A p-value of less than 0.05 indicates that the
parameter can be said to not be constant with at least 95% confidence. The
rows that are highlighted are those that were found to not have constant
values for the given parameter.
Linear regression analysis was also performed versus collector-emitter
voltage. However, due the nature of the test matrix, only a limited number
54
of test points could be compared to eliminate the effect of collector current in
the analysis (i.e. collector current increased as the collector-emitter voltage
increased). Therefore, the linear regression results versus collector current
are displayed because they are the most conclusive. Plots of the parameters
versus collector-emitter voltage will be shown later in this chapter.
AGD
The linear regression results tables for the Powerex IGBT indicate that
almost every parameter has a collector current dependence.
The gate-drain
overlap area, AGD, has a current dependence for every collector-emitter
voltage. This can be seen in Figure 4.11. Although the 1 kV results for all
IGBTs and the combination of them has a negative slope, the other voltages
have a positive slope. This result is likely due to current crowding through
the channel of the device until the device reaches its saturation current. As
the collector current through the device increases, the channel is rounded off
in the p+ region beneath the gate [64]. This rounding of the channel with
carriers effectively increases the overlap area between the gate and drain of
the internal MOSFET. It appears that the device reaches saturation beyond
approximately 400 A because there is a much smaller distribution at higher
collector currents.
This may account for the change in the plot versus
voltage because the data sets plotted are for 222 A and 333 A. The analysis
for the IXYS IGBT is similar to that for the Powerex IGBT but the current
values are lower. The IXYS results can be seen in Figure 4.12.
55
Table 4.3:
Linear regression results versus collector current for the Powerex
QIS4506001 IGBT #1 parameters that were fit using least squares method
Parameter
AGD
ln(MUN)
CGS
MUP
ln(COXD)
KP
CollectorEmitter
Voltage
(kV)
Parameter
Change per
Ampere
p-value
Lower 95%
Parameter
Change
Upper 95%
Parameter
Change
1
-2.10 x 10-8
4.20 x 10-3
-3.60 x 10-8
-7.2 x 10-9
2
2.31 x 10-8
2.48 x 10-8
1.64 x 10-8
2.98 x 10-8
3
2.91 x 10-8
3.87 x 10-21
2.60 x 10-8
3.22 x 10-8
3.5
2.79 x 10-8
1.07 x 10-27
2.60 x 10-8
2.99 x 10-8
1
0.0155
1.90 x 10-16
0.0133
0.0178
2
0.00829
5.28 x 10-17
0.00713
0.00945
3
0.0106
2.77 x 10-23
0.00965
0.0116
3.5
-0.00090
0.078
-0.0019
0.00011
-11
5.65 x 10-11
4.91 x 10
2
1.36 x 10-10
3.68 x 10-17
1.17 x 10-10
1.55 x 10-10
3
1.92 x 10-10
1.30 x 10-18
1.68 x 10-10
2.16 x 10-10
3.5
2.09 x 10-10
1.23 x 10-19
1.85 x 10-10
2.34 x 10-10
1
148.28
2.81 x 10-9
109.31
187.24
2
158.84
4.25 x 10-14
131.22
186.47
3
47.82
3.22 x 10-14
39.58
56.06
3.5
24.15
5.21 x 10-9
17.63
30.66
1
0.013
1.48 x 10-20
0.012
0.015
2
0.001
2.59 x 10-13
0.0077
0.011
3
0.008
8.91 x 10-10
0.0057
0.0095
3.5
0.002
4.80 x 10-20
0.0019
0.0024
1
0.049
4.35 x 10-15
0.041
0.057
0.039
1.70 x 10
-13
0.032
0.047
7.84 x 10
-35
0.057
0.062
3.30 x 10
-36
0.059
0.064
3
3.5
0.060
0.061
56
4.18 x 10
-11
1
2
4.10 x 10
-16
Table 4.4:
Linear regression results versus collector current for the Powerex
QIS4506001 IGBT #2 parameters that were fit using least squares method
Parameter
AGD
ln(MUN)
CGS
MUP
ln(COXD)
KP
CollectorEmitter
Voltage
(kV)
Parameter
Change per
Ampere
p-value
Lower 95%
Parameter
Change
Upper 95%
Parameter
Change
1
-8.20 x 10-9
0.098
-1.80 x 10-8
1.58 x 10-9
2
2.72 x 10-8
7.47 x 10-17
2.34 x 10-8
3.11 x 10-8
3
2.41 x 10-8
4.21 x 10-16
2.05 x 10-8
2.77 x 10-8
3.5
2.17 x 10-8
4.13 x 10-8
1.53 x 10-8
2.81 x 10-8
1
0.0177
2.35 x 10-22
0.0160
0.0194
2
0.0081
1.62 x 10-16
0.0069
0.0093
3
0.0072
3.38 x 10-31
0.0068
0.0076
3.5
0.0046
3.34 x 10-5
0.0026
-11
7.64 x 10-11
6.97 x 10
2
1.51 x 10-10
1.19 x 10-17
1.30 x 10-10
1.71 x 10-10
3
1.92 x 10-10
3.78 x 10-20
1.70 x 10-10
2.13 x 10-10
3.5
2.32 x 10-10
9.50 x 10-21
2.07 x 10-10
2.57 x 10-10
1
141.14
1.55 x 10-11
110.94
171.34
2
159.61
2.18 x 10-15
134.43
184.78
3
52.94
4.54 x 10-19
46.51
59.37
3.5
7.53
0.125
-2.19
17.26
1
0.013
1.95 x 10-20
0.012
0.015
2
0.0089
3.80 x 10-12
0.0071
0.011
3
0.0074
5.67 x 10-9
0.0054
0.0094
3.5
0.0080
5.12 x 10-8
0.0056
0.010
1
0.047
1.58 x 10-16
0.040
0054
0.035
2.72 x 10
-13
0.029
0.042
4.00 x 10
-36
0.051
0.056
6.79 x 10
-37
0.058
0063
3
3.5
0.054
0.061
57
6.29 x 10
0.0066
-11
1
2
2.30 x 10
-22
Table 4.5: Linear regression results versus collector current for the Powerex
QIS4506001 IGBT #3 parameters that were fit using least squares method
Parameter
AGD
ln(MUN)
CGS
MUP
ln(COXD)
KP
CollectorEmitter
Voltage
(kV)
Parameter
Change per
Ampere
p-value
Lower 95%
Parameter
Change
Upper 95%
Parameter
Change
1
-1.40 x 10-8
5.34 x 10-3
-2.40 x 10-8
-4.40 x 10-8
2
2.12 x 10-8
9.34 x 10-11
1.64 x 10-8
2.61 x 10-8
3
2.35 x 10-8
1.98 x 10-19
2.07 x 10-8
2.63 x 10-8
3.5
1.69 x 10-8
3.05 x 10-8
1.20 x 10-8
2.19 x 10-8
1
0.0172
2.72 x 10-20
0.0153
0.0191
2
0.0081
3.62 x 10-15
0.0068
0.0094
3
0.0053
7.68 x 10-7
0.0035
0.0071
3.5
0.0048
1.37 x 10-5
0.0028
-11
6.90 x 10-11
6.14 x 10
2
1.48 x 10-10
5.12 x 10-21
1.32 x 10-10
1.64 x 10-10
3
1.86 x 10-10
1.70 x 10-28
1.73 x 10-10
1.98 x 10-10
3.5
2.42 x 10-10
1.43 x 10-20
2.16 x 10-10
2.69 x 10-10
1
135.75
2.76 x 10-12
108.47
163.02
2
169.59
359 x 10-17
146.09
193.10
3
52.60
2.68 x 10-22
47.46
57.75
3.5
-1.32
0.873
-17.91
15.27
1
0.013
2.19 x 10-20
0.011
0.014
2
0.0076
4.74 x 10-12
0.0061
0.0092
3
0.0072
1.16 x 10-8
0.0052
0.0093
3.5
0.0078
5.42 x 10-8
0.0055
0.0101
1
0.049
1.89 x 10-18
0.043
0.055
0.034
1.01 x 10
-13
0.028
0.040
2.82 x 10
-36
0.051
0.055
8.09 x 10
-27
0.061
0.071
3
3.5
0.053
0.066
58
5.38 x 10
0.0067
-11
1
2
9.51 x 10
-19
Table 4.6: Linear regression results versus collector current for three different
Powerex QIS4506001 IGBTs combining parameters that were fit using least squares
method
Parameter
AGD
ln(MUN)
CGS
MUP
ln(COXD)
KP
CollectorEmitter
Voltage
(kV)
Parameter
Change per
Ampere
p-value
Lower 95%
Parameter
Change
Upper 95%
Parameter
Change
1
-1.40 x 10-8
7.38 x 10-5
-2.10 x 10-8
-7.40 x 10-9
2
2.39 x 10-8
1.61 x 10-29
2.08 x 10-8
2.70 x 10-8
3
2.56 x 10-8
1.06 x 10-52
2.37 x 10-8
2.74 x 10-8
3.5
2.21 x 10-8
5.48 x 10-28
1.91 x 10-8
2.51 x 10-8
1
0.0168
1.05 x 10-55
0.0157
0.0180
2
0.0082
3.90 x 10-47
0.0075
0.0088
3
0.0077
6.95 x 10-37
0.0069
0.0085
3.5
0.0027
7.21 x 10-6
0.0016
0.0039
1
6.01 x 10-11
1.65 x 10-47
5.52 x 10-11
6.50 x 10-11
2
1.45 x 10-10
8.68 x 10-54
1.35 x 10-10
1.56 x 10-10
3
1.90 x 10-10
3.71 x 10-62
1.79 x 10-10
2.01 x 10-10
3.5
2.27 x 10-10
6.16 x 10-58
2.13 x 10-10
242 x 10-10
1
141.41
1.16 x 10-29
123.04
159.78
162.69
7.81 x 10
-45
148.49
176.90
1.11 x 10
-41
46.26
55.89
-3
3.48
17.57
2
3
51.08
3.5
10.53
3.73 x 10
1
0.013
4.36 x 10-60
0.012
0.014
2
0.0086
4.57 x 10-35
0.0077
0.0096
3
0.0074
2.31 x 10-25
0.0063
0.0085
3.5
0.0059
1.69 x 10-15
0.0046
0.0071
1
0.048
7.20 x 10-48
0.044
0.052
2
0.036
8.54 x 10-38
0.032
0.040
3
0.055
3.90 x 10-101
0.054
0.057
3.5
0.063
3.18 x 10-92
0.061
0.065
59
Table 4.7:
Linear regression results versus collector current for the IXYS
IXEL40N400 IGBT parameters that were fit using least squares method
Parameter
AGD
ln(MUN)
CGS
MUP
ln(COXD)
KP
CollectorEmitter
Voltage
(kV)
Parameter
Change per
Ampere
p-value
Lower 95%
Parameter
Change
Upper 95%
Parameter
Change
1
4.66 x 10-8
2.49 x 10-10
3.55 x 10-8
5.76 x 10-8
2
3.92 x 10-8
2.07 x 10-9
2.90 x 10-8
4.94 x 10-8
3
3.71 x 10-8
6.49 x 10-7
2.45 x 10-8
4.98 x 10-8
3.5
4.31 x 10-8
1.08 x 10-15
3.65 x 10-8
4.98 x 10-8
1
0.018154
5.02 x 10-17
0.015611
0.020696
2
0.033742
6.02 x 10-22
0.030366
0.037118
3
0.018882
2.56 x 10-3
0.007044
0.030721
3.5
0.059137
9.54 x 10-13
0.047696
0.070578
-10
3.51 x 10-10
3.11 x 10
2
3.63 x 10-10
1.27 x 10-24
3.33 x 10-10
3.94 x 10-10
3
4.68 x 10-10
3.37 x 10-20
4.16 x 10-10
5.21 x 10-10
3.5
6.03 x 10-10
7.61 x 10-23
5.46 x 10-10
6.59 x 10-10
1
52.25
2.85 x 10-15
43.94
60.57
2
-13.50
0.1626
-32.69
5.69
3
-18.22
0.2833
-52.10
15.66
3.5
-59.60
1.56 x 10-3
-95.00
-24.19
1
0.020
8.37 x 10-32
0.019
0.021
2
0.032
1.34 x 10-39
0.031
0.034
3
0.021
9.49 x 10-8
0.014
0.027
3.5
0.051
2.48 x 10-23
0.047
0.056
1
0.078
2.19 x 10-42
0.076
0.080
0.089
1.62 x 10
-24
0.081
0.096
6.46 x 10
-17
0.073
0.098
2.99 x 10
-17
0.086
0.113
3
3.5
0.086
0.099
60
2.70 x 10
-10
1
2
4.33 x 10
-18
MUN
Similar to the results for AGD, the Powerex device has a collector
current dependence on the electron mobility, MUN, shown in the regression
tables.
This is confirmed in Figure 4.13 and Figure 4.14.
shows MUN
plotted versus collector current.
Figure 4.13a
There appears to bealmost
exponential growth in MUN at the maximum current values.
Figure 4.13b
shows a zoomed in plot of the smaller MUN values that are difficult to see on
the full plot. Trendlines have been deleted for this plot to avoid confusion,
since not all MUN values can be seen on the zoomed in plot. This is likely
also due to the current reaching its saturation level in the device.
The
electric field in the internal BJT increases as the current crowding worsens
[64]. Therefore, MUN should also increase [60].
Figure 4.14 shows MUN plotted versus collector-emitter voltage. MUN
has a decreasing trend for increasing collector-emitter voltage.
For
increasing collector-emitter voltages, the length of the depletion region in the
device before turn-on also increases.
This increase in distance across the
depletion region would correspond to a decreased electric field within the
depletion region. This may account for the decrease in MUN for increasing
collector-emitter charge voltage. IXYS results can be seen in Figure 4.15 and
Figure 4.16.
61
2.50E-05
2.00E-05
2
AGD (m )
1.50E-05
1.00E-05
5.00E-06
0.00E+00
0
100
200
300
400
500
600
700
800
Collector Current (A)
Vce=1kV
Vce=2kV
Vce=3kV
Vce=3.5kV
(a)
2.00E-05
1.80E-05
1.60E-05
1.40E-05
AGD (m2)
1.20E-05
1.00E-05
8.00E-06
6.00E-06
4.00E-06
2.00E-06
0.00E+00
0
500
1000
1500
2000
2500
3000
3500
Collector-Emitter Voltage (V)
Ic=222 A
Ic=333 A
(b)
Figure 4.11: Combined AGD values for three different Powerex QIS4506001 IGBTs
versus (a) collector current (b) collector-emitter voltage
62
2.50E-05
2.00E-05
2
AGD (m )
1.50E-05
1.00E-05
5.00E-06
0.00E+00
0
50
100
150
200
250
300
350
400
Collector Current (A)
Vce=1 kV
Vce=2 kV
Vce=3 kV
Vce=3.5 kV
2000
2500
(a)
2.5000E-05
2.0000E-05
AGD (m2)
1.5000E-05
1.0000E-05
5.0000E-06
0.0000E+00
0
500
1000
1500
3000
3500
Collector-Emitter Voltage (V)
Ic=222 A
Ic=333 A
(b)
Figure 4.12: AGD values for an IXYS IXEL40N400 IGBT versus (a) collector current
(b) collector-emitter voltage
63
8.00E+08
7.00E+08
6.00E+08
4.00E+08
2
MUN (cm /(V*s))
5.00E+08
3.00E+08
2.00E+08
1.00E+08
0.00E+00
0
100
200
300
400
500
600
700
800
700
800
-1.00E+08
Collector Current (A)
Vce=1kV
Vce=2kV
Vce=3kV
Vce=3.5kV
(a)
5.00E+07
2
MUN (cm /(V*s))
4.00E+07
3.00E+07
2.00E+07
1.00E+07
0.00E+00
0
100
200
300
400
500
600
Collector Current (A)
Vce=1kV
Vce=2kV
Vce=3kV
Vce=3.5kV
(b)
Figure 4.13: Combined MUN values for three different Powerex QIS4506001 IGBTs
versus collector current showing (a) all values (b) zoomed into smaller values
64
7.00E+07
6.00E+07
MUN (cm2/(V*s))
5.00E+07
4.00E+07
3.00E+07
2.00E+07
1.00E+07
0.00E+00
0
500
1000
1500
2000
2500
3000
3500
Collector-Emitter Voltage (V)
Ic=222A
Ic=333A
Figure 4.14: Combined MUN values for three different Powerex QIS4506001 IGBTs
versus collector-emitter voltage
65
2.00E+12
1.00E+12
2
MUN (cm /(V*s))
1.50E+12
5.00E+11
0.00E+00
0
50
100
150
200
250
300
350
400
-5.00E+11
Collector Current (A)
Vce=1 kV
Vce=2 kV
Vce=3 kV
Vce=3.5 kV
(a)
2.00E+09
1.80E+09
1.60E+09
1.20E+09
2
MUN (cm /(V*s))
1.40E+09
1.00E+09
8.00E+08
6.00E+08
4.00E+08
2.00E+08
0.00E+00
0
50
100
150
200
250
300
350
400
Collector Current (A)
Vce=1 kV
Vce=2 kV
Vce=3 kV
Vce=3.5 kV
(b)
Figure 4.15: MUN values for an IXYS IXEL40N400 IGBT versus collector current
showing (a) all values (b) zoomed into smaller values
66
1.2000E+09
1.0000E+09
MUN (cm2/(V*s))
8.0000E+08
6.0000E+08
4.0000E+08
2.0000E+08
0.0000E+00
0
500
1000
1500
2000
2500
3000
3500
Collector-Emitter Voltage (V)
Ic=222A
Figure 4.16:
voltage
Ic=333A
MUN values for an IXYS IXEL40N400 IGBT versus collector-emitter
67
CGS
The gate-source capacitance, CGS, has a collector current dependence as
shown in the regression tables.
This can be seen in Figure 4.17 for both
current and voltage. The increase in collector current causes an increase in
CGS.
Using C = Q/V, the increase in collector current would account for
more charge in the region during the pulse width. Thus, there would also be
increased capacitance for increasing collector current. Similar to AGD, there
appears to be a threshold of about 400 A, beyond which CGS increases
significantly.
This may account for the change from slightly positive to
slightly negative slope for the plot versus collector-emitter voltage as the two
data sets are at 222 A and 333 A.
Similar to AGD, this may be due to
current reaching its saturation level in the device. The results for the IXYS
IGBT are seen in Figure 4.18. The current in the IXYS plot is lower but has
approximately the same maximum values.
The voltage plot differs in that
there was found to be no statistical difference versus voltage for the 222 A
data.
Also, the values for CGS decreased an order of magnitude from the
Powerex to the IXYS IGBTs.
MUP
The hole mobility, MUP, has positive slopes that decrease in magnitude
for increasing collector-emitter voltages.
Again, this appears to have a
current threshold of approximately 300 A, beyond which the current
dependence decreases.
Because hole mobility is much less than electron
68
1.60E-07
1.40E-07
1.20E-07
2
CGS (F/cm )
1.00E-07
8.00E-08
6.00E-08
4.00E-08
2.00E-08
0.00E+00
0
100
200
300
400
500
600
700
800
Collector Current (A)
Vce=1kV
Vce=2kV
Vce=3kV
Vce=3.5kV
(a)
4.00E-08
3.50E-08
3.00E-08
2
CGS (F/cm )
2.50E-08
2.00E-08
1.50E-08
1.00E-08
5.00E-09
0.00E+00
0
500
1000
1500
2000
2500
3000
3500
Collector-Emitter Voltage (V)
Ic=222 A
Ic=333 A
(b)
Figure 4.17: Combined CGS values for three different Powerex QIS4506001 IGBTs
versus (a) collector current (b) collector-emitter voltage
69
1.60E-07
1.40E-07
1.20E-07
CGS (F/cm2)
1.00E-07
8.00E-08
6.00E-08
4.00E-08
2.00E-08
0.00E+00
0
50
100
150
200
250
300
350
400
Collector Current (A)
Vce=1 kV
Vce=2 kV
Vce=3 kV
Vce=3.5 kV
2000
2500
(a)
1.2000E-07
1.0000E-07
CGS (F/cm2)
8.0000E-08
6.0000E-08
4.0000E-08
2.0000E-08
0.0000E+00
0
500
1000
1500
3000
3500
Collector-Emitter Voltage (V)
Ic=222 A
Ic=333 A
(b)
Figure 4.18: CGS values for an IXYS IXEL40N400 IGBT versus (a) collector current
(b) collector-emitter voltage
70
1.20E+05
1.00E+05
2
MUP (cm /(V*s))
8.00E+04
6.00E+04
4.00E+04
2.00E+04
0.00E+00
0
100
200
300
400
500
600
700
800
Collector Current (A)
Vce=1kV
Vce=2kV
Vce=3kV
Vce=3.5kV
(a)
8.00E+04
7.00E+04
5.00E+04
2
MUP (cm /(V*s))
6.00E+04
4.00E+04
3.00E+04
2.00E+04
1.00E+04
0.00E+00
0
500
1000
1500
2000
2500
3000
3500
Collector-Emitter Voltage (V)
Ic=222 A
Ic=333 A
(b)
Figure 4.19: Combined MUP values for three different Powerex QIS4506001 IGBTs
versus (a) collector current (b) collector-emitter voltage
71
3.00E+04
2.50E+04
2
MUP (cm /(V*s))
2.00E+04
1.50E+04
1.00E+04
5.00E+03
0.00E+00
0
50
100
150
200
250
300
350
400
Collector Current (A)
Vce=1 kV
Vce=2 kV
Vce=3 kV
Vce=3.5 kV
(a)
3.0000E+04
2.5000E+04
2
MUP (cm /(V*s))
2.0000E+04
1.5000E+04
1.0000E+04
5.0000E+03
0.0000E+00
0
500
1000
1500
2000
2500
3000
3500
Collector-Emitter Voltage (V)
Ic=222 A
Ic=333 A
(b)
Figure 4.20: MUP values for an IXYS IXEL40N400 IGBT versus (a) collector current
(b) collector-emitter voltage
72
mobility, it is possible that current saturation affects MUP less than MUN.
However, there is still a current dependence for the Powerex IGBT. Despite
the dependence for the Powerex IGBT, the IXYS IGBT only had statistical
significance for the 1 kV and 3.5 kV data sets as shown in Table 4.7. This
discrepancy is likely due to the decreased sample size for the IXYS IGBT,
since only one IXYS IGBT was tested. Unlike MUN for both types of IGBTs,
the two data sets for MUP on the plots versus collector-emitter voltage have
different polarity slopes.
Additionally, the IXYS and Powerex plots versus
voltage have reversed slopes for the data sets.
This may be due to
compensation for some other parameter during the least squares curve
fitting process.
It is also possible that the different manufacturer’s dyes
affect this parameter differently.
COXD
The gate-drain oxide capacitance, COXD, has results similar to that of
MUN.
There is a positively sloped collector current dependence and a
negatively sloped collector-emitter voltage dependence as shown in Figure
4.21 and Figure 4.22.
Similar to MUN, there is a seemingly exponential
increase in COXD at high currents. However, despite the similarities between
the COXD and MUN plots, it is likely the increase in AGD that causes the
increase in COXD.
COXD refers to basically the same overlapping area as
AGD. Therefore, the increase in area would directly correlate to an increase
in capacitance from C=εA/d. Similar to CGS, COXD decreases in magnitude
73
2.00E-01
2
COXD (F/cm )
1.50E-01
1.00E-01
5.00E-02
0.00E+00
0
100
200
300
400
500
600
700
800
700
800
-5.00E-02
Collector Current (A)
Vce=1kV
Vce=2kV
Vce=3kV
Vce=3.5kV
(a)
3.00E-02
2
COXD (F/cm )
2.50E-02
2.00E-02
1.50E-02
1.00E-02
5.00E-03
0.00E+00
0
100
200
300
400
500
600
Collector Current (A)
Vce=1kV
Vce=2kV
Vce=3kV
Vce=3.5kV
(b)
Figure 4.21: Combined COXD values for three different Powerex QIS4506001 IGBTs
versus collector current showing (a) all values (b) zoomed into smaller values
74
2.50E-02
COXD (F/cm2)
2.00E-02
1.50E-02
1.00E-02
5.00E-03
0.00E+00
0
500
1000
1500
2000
2500
3000
3500
Collector-Emitter Voltage (V)
Ic=222 A
Ic=333 A
Figure 4.22: Combined COXD values for three different Powerex QIS4506001 IGBTs
versus collector-emitter voltage
75
70
60
50
2
COXD (F/cm )
40
30
20
10
0
0
50
100
150
200
250
300
350
400
-10
Collector Current (A)
Vce=1 kV
Vce=2 kV
Vce=3 kV
Vce=3.5 kV
200
250
(a)
3.00E+00
2.50E+00
2
COXD (F/cm )
2.00E+00
1.50E+00
1.00E+00
5.00E-01
0.00E+00
0
50
100
150
300
350
400
Collector Current (A)
Vce=1 kV
Vce=2 kV
Vce=3 kV
Vce=3.5 kV
(b)
Figure 4.23: COXD values for an IXYS IXEL40N400 IGBT versus collector current
showing (a) all values (b) zoomed into smaller values
76
2.5
COXD (F/cm2)
2
1.5
1
0.5
0
0
500
1000
1500
2000
2500
3000
3500
Collector-Emitter Voltage (V)
Ic=222 A
Figure 4.24:
voltage
Ic=333 A
COXD values for an IXYS IXEL40N400 IGBT versus collector-emitter
77
for increasing collector emitter voltage. This would also be due to a decrease
in voltage for approximately the same amount of charge.
By C=Q/V, the
capacitance should decrease for increasing voltage. The IXYS results can be
seen in Figure 4.23 and Figure 4.24.
KPsat =
Zµ ni COXD
(VG − VTH )
LCH
Equation 4.5 [64]
KP
The results for the MOS transconductance, KP, show a strong current
dependence throughout the test matrix.
This can be seen in the linear
regression tables, as well as, Figure 4.25. The current dependence shown in
Figure 4.25a for the Powerex IGBT and Figure 4.26a for the IXYS IGBT
corresponds to relationship between COXD and KP shown in Equation 4.5,
where KPsat is the MOS transconductance in current saturation mode, Z is the
channel width orthogonal to the cross section, µni is the inversion layer
electron mobility, VG is the gate voltage, VTH is the threshold voltage, and LCH
is the channel length [64]. Figure 4.25b and Figure 4.26b show plots of KP
versus
collector-emitter
voltage
for
the
Powerex
and
IXYS
IGBTs,
respectively. Despite the correlation between KP and COXD, the plots for KP
versus voltage do not correspond to those for COXD. According to Equation
4.5, this discrepancy would be associated with changing channel length or
78
width when the voltage is held constant.
Although, these variables would
also change for increasing current, COXD must dominate when the collectoremitter voltage is held constant and the collector current is increased.
79
50
45
40
35
2
KP (A/V )
30
25
20
15
10
5
0
0
100
200
300
400
500
600
700
800
Collector Current (A)
Vce=1kV
Vce=2kV
Vce=3kV
Vce=3.5kV
(a)
25
20
2
KP (A/V )
15
10
5
0
0
500
1000
1500
2000
2500
3000
3500
Collector-Emitter Voltage (V)
Ic=222A
Ic=333 A
(b)
Figure 4.25: Combined KP values for three different Powerex QIS4506001 IGBTs
versus (a) collector current (b) collector-emitter voltage
80
40
35
30
2
KP (A/V )
25
20
15
10
5
0
0
50
100
150
200
250
300
350
400
Collector Current (A)
Vce=1 kV
Vce=2 kV
Vce=3 kV
Vce=3.5 kV
(a)
30
25
2
KP (A/V )
20
15
10
5
0
0
500
1000
1500
2000
2500
3000
3500
Collector-Emitter Voltage (V)
Ic=222A
Ic=333 A
(b)
Figure 4.26: KP values for an IXYS IXEL40N400 IGBT versus (a) collector current
(b) collector-emitter voltage
81
5. Conclusions
At the University of Missouri, a test stand was constructed for
measuring the performance of two different manufacturer’s IGBTs under
pulsed-power conditions. These data were used to determine the modeling
input parameters for the Oziemkiewicz implementation of the Hefner IGBT
model under varied pulsed input voltages and currents. This was done in an
attempt to extend the Oziemkiewicz implementation utilized in some
common circuit simulation software packages, such as Micro-Cap 9, beyond
the typical voltage and current levels into the pulsed-power regime [36].
The Powerex QIS4506001 and IXYS IXEL40N400 IGBTs were the
devices being tested. Each IGBT was tested according to a test matrix that
involved collector-emitter voltages up to 3.5 kV with resistive loads as low as
3 Ω. All tests performed had 10 µs pulsewidth and an approximately 24 V
82
gate-emitter voltage. Although loads as small as 3 Ω were tested for each
IGBT, all of the tested IGBTs reached their saturation current levels before
the anticipated maximums were reached. Both models of IGBTs should have
reached theoretical collector currents of approximately 1.17 kA, but both
models reached saturation collector currents well below this level.
The
Powerex IGBT reached a maximum collector current at approximately 700 A,
while the IXYS IGBT reached its maximum collector current at approximately
350 A. This is likely due to the different current saturation levels in each of
the devices.
Overall, the Powerex IGBT performed better for the higher
currents required in the pulsed power regime for the 24 V gate signal that
was applied.
A least-squares curve fitting algorithm was implemented to determine
the modeling input parameters from the experimental data. Although fitting
all seventeen of the modeling parameters took an impractically long time, a
limited set of parameters were fit while holding the remaining parameters
constant. The parameters that were fit using the least-squares curve fitting
algorithm were the gate-drain overlap area (AGD), electron and hole
mobilities (MUN and MUP), gate-source capacitance per unit area (CGS),
gate-oxide drain capacitance (COXD), and MOS transconductance (KP). This
decrease in simulation size led to more practical simulation times with the
tradeoff of having some points of the test matrix that could not be fit as well
as others. Once the parameters were determined for all points of the test
matrix, a linear regression analysis was performed on the data. This allowed
83
for the determination of whether or not the parameters remained constant
throughout the test range.
Using the linear regression analysis on the determined parameters, it
was found that all of the parameters had a non-zero slope when plotted
versus collector current. Additionally, most of the parameters that were fit
also have a non-zero slope with respect to collector-emitter voltage.
This
shows that the existing modeling input parameters cannot be held constant
throughout the pulsed-power regime.
cannot
be
used
for
IGBT
modeling
Although constant nominal values
parameters
under
pulsed-power
conditions, it is still possible to utilize the existing model to simulate these
devices. This was shown in the previous chapter by comparing the simulated
waveforms for the collector-emitter voltage, collector current, and gateemitter voltage with the measured waveforms.
By utilizing plots of the
parameters versus collector current and collector-emitter voltage similar to
those found in this thesis, it would be possible to interpolate approximate
parameter values for the conditions being modeled.
It has been shown that the existing model in Micro-Cap 9 and other
similar circuit simulation software packages can be utilized to simulate power
IGBTs under pulsed-power conditions.
Despite having to empirically
determine the parameters necessary to model these devices, adequate
simulations can be performed for future designs utilizing the IGBT. However,
this thesis has shown that modifications will have to be made to the
84
Oziemkiewicz implementation of the Hefner IGBT model if simulations are to
be performed using constant input parameters in the pulsed-power regime.
85
6. Future Work
Although this thesis has shown valuable insight into the modeling of
IGBTs under pulsed-power conditions, there is still more work that would be
beneficial for accurately modeling these devices. The foremost problems that
must be addressed are the mathematical relationships for the Oziemkiewicz
implementation.
Although Oziemkiewicz’s implementation of the Hefner
model works well for modeling IGBTs for more common low-voltage, lowcurrent switching applications, there are current and voltage dependences
that are unaccounted for in his implementation. As discussed in Chapter 4,
the effects of current crowding at the gate of these devices plays a role until
the device reaches its saturation current. The collector current reaching its
saturation level for the given gate signal also played an important role at
86
higher currents.
The mathematical incorporation of these effects into the
model would be an excellent beginning for looking into the discrepancies of
the model for normal operating conditions compared to pulsed-power
conditions.
The lack of fitness for the gate-emitter voltage waveforms indicates
that additional analysis needs to be performed on the gate of the device in
the model.
Additional gate inductances or shunt capacitances may be
necessary to better model the effects at the turn-off of the devices. Curve
fitting using the existing model with varying damping resistances at the gate
would determine the effects of damping at the gate and the omitted diode in
the gate circuitry utilized for simulating in this thesis. Testing with the IXYS
19N250A would show the difference between the existing IGBT model input
parameters in Micro-Cap 9 and the determined pulsed power parameters.
Additionally, it would be interesting to look into the effects additional
variables in the IGBT test stand. Some of these variables would include the
gate-emitter voltage magnitude, gate-emitter voltage waveform shape,
operating temperature, and operating frequency.
The effects of the gate-
emitter voltage magnitude and waveform shape would have significant
effects for the turn-on and turn-off of the device.
Testing different gate
voltages to look at varying saturation current levels would provide insight as
to the effects due to current saturation. The operating temperature would be
appropriate for looking at the robustness of these devices under such pulsedpower operating conditions.
Also, the operating temperature could give
87
insight as the maximum operational collector current that can be pushed
through the device.
For power modulator systems, the switches must be
repetitively pulsed open and closed under voltages and currents that are
typically beyond the manufacturers suggested ratings.
Therefore, the
maximum operating frequency under these conditions would be a beneficial
piece of information for power modulator system designers.
88
A. Least Squares Curve Fitting MATLAB Code
clear;
close;
clc;
%%%% simulation parameters
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
total_simulations=1;
generations_to_simulate=20;
%50
end_fitness=1e4;
%kunits
number_of_bugs=250;
%400
show_figure=2;
%0->nofigure,1->very small
figure, 2->full screen figure
parameter_variation=0.1;
%fraction allowed variability in
the parameters
number_of_shots=1;
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%
%%%% swarm behavior parameters. Probably don't mess with.
accel_constant_1=3;
%between 0 and 4. This is
associated with personal best. Global is (4 - this number).
%3 was better then 1.
inertia=0.5;
%less than 1. big=explore,
small=smooth (0.5>0.8>0.2)
number_of_elements=11;
%parameters being varied
total_elements=20;
%total IGBT parameters and stray
L's
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%
best_bugs=zeros(total_simulations,number_of_elements);
best_bugs_fitness=zeros(1,total_simulations);
shot_location='C:\Jim\';
estimated_charge_voltage=estimated_charge_voltage*1e3;
%convert to equiv. total estimated charge
simulation_number=1;
k=1;
temp0={'fitness' 'AGD' 'MUN' 'VT' 'CGS' 'MUP' 'TAU' 'COXD' 'KP'
'NB' 'THETA' 'RL' 'BVN' 'JSNE' 'AREA' 'KF' 'VTD' 'BVF' 'WB'
'Lstray1' 'Lstray2'};
while simulation_number<total_simulations+1
[global_best_position,global_best_position_constants,global_best_
89
fitness1,global_best_shot_fitness]=calibration_script(simulation_
number,
generations_to_simulate,number_of_bugs,accel_constant_1,inertia,n
umber_of_elements,shot_location,training_data_offset,show_figure,
estimated_charge_voltage,parameter_variation,end_fitness,training
_shot);
l=1;
while l<number_of_shots+1
global_best_fitness(l,1)=global_best_fitness1;
l=l+1;
end
temp1=[global_best_shot_fitness,global_best_position,global_best_
position_constants];
temp2(simulation_number+(k1)*number_of_shots:simulation_number+k*number_of_shots1,:)=temp1;
w=1;
while w<22
temp3(simulation_number+(k1)*number_of_shots:simulation_number+k*number_of_shots1,w)=cellstr(num2str(temp2(simulation_number+(k1)*number_of_shots:simulation_number+k*number_of_shots-1,w)));
w=w+1;
end
temp=[temp0;temp3];
xlswrite('C:\Jim\sim20bestbugs.tab', temp)
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
simulation_number=simulation_number+1;
k=k+1;
end;
90
function
[global_best_position,global_best_position_constants,global_best_
fitness,global_best_shot_fitness]=calibration_script(simulation_n
umber,
generations_to_simulate,number_of_bugs,accel_constant_1,inertia,n
umber_of_elements,shot_location,training_data_offset,show_figure,
estimated_charge_voltage,parameter_variation,end_fitness,training
_shot)
%%%%%%%% CONTROL VARIABLES %%%%%%%%
samplerate=24.999e-6/2500;
%sample
rate for simulations
simend=24.999e-6;
%when do the
simulations stop?
training_simend=24.999e-6;
train_on_Vce=1;
%0-no, 1-yes
train_on_Vge=1;
train_on_Ic=1;
training_weight=[1 100 100];
%weight to put on [Vce Ic Vge];
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
temp=size(training_shot);
number_of_shots=1;
shot_number=1;
time_delay=2.72e-6;
V_charge=3e3;
%%%%%%%% SET LIMITS FOR CIRCUIT ELEMENT VALUES %%%%%%%%
%
As with the bug_position and bug_velocity vectors, the
columns in the element_limits matrix are the
%
elements in the circuit model. The order and size is the same
in all three matrixes. The (2) rows in
%
the element_limits matrix are: 1- the upper limit of the
individual element value
%
2- the lower limit of the
individual element value
element_limits=zeros(2*number_of_shots,number_of_elements);
%initialize the matrix
while shot_number<number_of_shots+1
AREA = 3.5e-5;
guess=1.85e-5; %AGD 1
element_limits(shot_number*2-1,1)= AREA;
%upper limit of 1st element
element_limits(shot_number*2,1)= guess*(1.0parameter_variation);
%lower limit of 1st element
guess=1.44e8;
%MUN 2
91
element_limits(shot_number*2-1,2)=
guess*(1+parameter_variation);
%upper limit of 2nd
element
element_limits(shot_number*2,2)= guess*(1parameter_variation);
%lower limit of 2nd element
guess=12;%VT 3
element_limits(shot_number*2-1,3)= 12;
%upper limit of 3rd element
element_limits(shot_number*2,3)= 12;
%lower limit of 3rd element
guess=6.54e-8; %CGS 4
element_limits(shot_number*2-1,4)=
guess*(1+parameter_variation);
%upper limit of 4th element
element_limits(shot_number*2,4)= guess*(1parameter_variation);
%lower limit of 4th element
guess=9.14e3;
%MUP 5
element_limits(shot_number*2-1,5)=
guess*(1+parameter_variation);
%upper limit of 5th
element
element_limits(shot_number*2,5)= guess*(1parameter_variation);
%lower limit of 5th element
guess=2.0e-5;
%TAU 6
element_limits(shot_number*2-1,6)= 2e-5;
%upper limit of 6th element
element_limits(shot_number*2,6)= 2e-5;
%lower limit of 6th element
guess=8.91e-2;%COXD 7
element_limits(shot_number*2-1,7)=
guess*(1+parameter_variation);
%upper limit of 7th
element
element_limits(shot_number*2,7)= guess*(1parameter_variation);
%lower limit of 7th element
guess=27.02;%KP 8
element_limits(shot_number*2-1,8)=
guess*(1+parameter_variation);
%upper limit of 8th
element
element_limits(shot_number*2,8)= guess*(1parameter_variation);
%lower limit of 8th element
guess=2e15;%NB 9
element_limits(shot_number*2-1,9)= 2e15;
%upper limit of 9th element
element_limits(shot_number*2,9)= 2e15;
%lower limit of 9th element
guess=0.5;%Theta 10
92
element_limits(shot_number*2-1,10)= 0.5;
%upper limit of 11th element
element_limits(shot_number*2,10)= 0.5;
%lower limit of 11th element
guess=9.0;
%Rload 11
element_limits(shot_number*2-1,11)= guess+0.1*guess;
%upper limit of 12th element
element_limits(shot_number*2,11)= guess-0.1*guess;
%lower limit of 12th element
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
shot_number=shot_number+1;
end
%%%%%%%% USED BY PROGRAM %%%%%%%%
% Variable initialization
accel_constant_2=4-accel_constant_1;
bug_num=1;
global_best_position = zeros(number_of_shots,number_of_elements);
global_best_fitness = 1e30;
shot_fitness = zeros(number_of_shots,1);
global_best_shot_fitness = zeros(number_of_shots,1);
personal_best_position =
zeros(number_of_bugs*number_of_shots,number_of_elements);
personal_best_fitness(number_of_bugs,1) = 10e30;
personal_best_fitness(:,1) = 10e30;
bug_fitness(number_of_bugs,1) = 10e30;
xaxis=[samplerate:samplerate:simend];
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%
Initialize the position and velocity for the bugs. The
velocity is zero and initialize_bugs takes care of the position.
bug_position =
zeros(number_of_bugs*number_of_shots,number_of_elements);
bug_position =
initialize_bugs(number_of_bugs,number_of_shots,element_limits);
bug_velocity =
zeros(number_of_bugs*number_of_shots,number_of_elements);
%%%%%%%%%%%%%%%%%%%%%%%%%%%
% %%%%% GET TEST DATA %%%%%%%%
shot_number=1;
training_data_saved=zeros(floor(training_simend/samplerate),numbe
r_of_shots*4);
size(training_data_saved);
training_data_saved_temp1=xlsread('C:\Jim\shot7\1');
training_data_saved_temp2=xlsread('C:\Jim\shot7\2');
93
training_data_saved_temp3=xlsread('C:\Jim\shot7\3');
training_data_saved_temp4=xlsread('C:\Jim\shot7\4');
training_data_saved_temp5=xlsread('C:\Jim\shot7\5');
training_data_saved_temp6=xlsread('C:\Jim\shot7\6');
training_data_saved_temp7=xlsread('C:\Jim\shot7\7');
training_data_saved_temp8=xlsread('C:\Jim\shot7\8');
training_data_saved_temp9=xlsread('C:\Jim\shot7\9');
training_data_saved_temp10=xlsread('C:\Jim\shot7\10');
while shot_number<number_of_shots+1
if shot_number==1
n=1;
while n<5
training_data_saved(:,n+4*(shot_number1))=training_data_saved_temp1(:,n);
n=n+1;
end
end
if shot_number==2
n=1;
while n<5
training_data_saved(:,n+4*(shot_number1))=training_data_saved_temp2(:,n);
n=n+1;
end
end
if shot_number==3
n=1;
while n<5
training_data_saved(:,n+4*(shot_number1))=training_data_saved_temp3(:,n);
n=n+1;
end
end
if shot_number==4
n=1;
while n<5
training_data_saved(:,n+4*(shot_number1))=training_data_saved_temp4(:,n);
n=n+1;
end
end
if shot_number==5
n=1;
while n<5
training_data_saved(:,n+4*(shot_number1))=training_data_saved_temp5(:,n);
n=n+1;
end
end
if shot_number==6
n=1;
94
while n<5
training_data_saved(:,n+4*(shot_number1))=training_data_saved_temp6(:,n);
n=n+1;
end
end
if shot_number==7
n=1;
while n<5
training_data_saved(:,n+4*(shot_number1))=training_data_saved_temp7(:,n);
n=n+1;
end
end
if shot_number==8
n=1;
while n<5
training_data_saved(:,n+4*(shot_number1))=training_data_saved_temp8(:,n);
n=n+1;
end
end
if shot_number==9
n=1;
while n<5
training_data_saved(:,n+4*(shot_number1))=training_data_saved_temp9(:,n);
n=n+1;
end
end
if shot_number==10
n=1;
while n<5
training_data_saved(:,n+4*(shot_number1))=training_data_saved_temp10(:,n);
n=n+1;
end
end
shot_number=shot_number+1;
end
%%%%%%%%%initialize all paramters being held
constant%%%%%%%%%%%%%%%%%%
num_constants=9;
bug_position1 =
zeros(number_of_bugs*number_of_shots,num_constants);
x=1;
while x < number_of_bugs*number_of_shots+1
95
bug_position1(x,1)=25;
%BVN
bug_position1(x,2)=4e-11;
%JSNE
bug_position1(x,3)=3.5e-5;
%AREA
bug_position1(x,4)=300;
%KF
bug_position1(x,5)=22;
%VTD
bug_position1(x,6)=4.5e4;
%BVF
bug_position1(x,7)=5e-4;
%WB
bug_position1(x,8)=4e-7;
%Lstray1
bug_position1(x,9)=2.5e-6;
%Lstray2
x=x+1;
end
% repeat for all generations
generation=1;
while (generation<generations_to_simulate+1) &
(global_best_fitness > end_fitness)
%%%%%%write all netlist files%%%%%%%%%
shot_number=1;
while shot_number < number_of_shots+1
bug_num=1;
while bug_num < number_of_bugs+1
write_netlist(bug_position,bug_position1,bug_num,samplerate,simen
d,number_of_bugs,shot_number,time_delay,V_charge);
bug_num=bug_num+1;
end
shot_number=shot_number+1;
end
%%%%%%%simulate all netlists in batch
file%%%%%%%%%%%%%%%%%%%%%%%%%%
test_bat(number_of_bugs*number_of_shots);
%write batch
file with all netlist names
system('mc9 @test.bat');
%run bugs in Micro-Cap
%compare results for each bug
bug_num=1;
while bug_num<number_of_bugs+1
bug_fitness=0;
shot_number=1;
shot_fitness = zeros(number_of_shots,1);
while shot_number<number_of_shots+1
file_name =
sprintf('C:\\Jim\\bug_files\\test%03d.TNO', bug_num +
number_of_bugs*(shot_number-1));
if exist(file_name)==0
swarm_data = ones(2500,4)*-5e9;
96
bad = 1
else
swarm_data = load('-ascii', file_name);
delete(file_name);
end
if train_on_Vce==1
bug_fitness = bug_fitness+sum((swarm_data(:,2)training_data_saved(:,2+(shot_number1)*4)).^2)*training_weight(1);
shot_fitness(shot_number,1) =
shot_fitness(shot_number,1) + sum((swarm_data(:,2)training_data_saved(:,2+(shot_number1)*4)).^2)*training_weight(1);
end
if train_on_Vge==1
bug_fitness = bug_fitness+sum((swarm_data(:,3)training_data_saved(:,3+(shot_number1)*4)).^2)*training_weight(2);
shot_fitness(shot_number,1) =
shot_fitness(shot_number,1) + sum((swarm_data(:,2)training_data_saved(:,2+(shot_number1)*4)).^2)*training_weight(1);
end
if train_on_Ic==1
bug_fitness = bug_fitness+sum((swarm_data(:,4)training_data_saved(:,4+(shot_number1)*4)).^2)*training_weight(3);
shot_fitness(shot_number,1) =
shot_fitness(shot_number,1) + sum((swarm_data(:,2)training_data_saved(:,2+(shot_number1)*4)).^2)*training_weight(1);
end
shot_number=shot_number+1;
end
%is it a personal best?
if bug_fitness< personal_best_fitness(bug_num,1)
personal_best_fitness(bug_num,1)=bug_fitness;
shot_number=1;
while shot_number<number_of_shots+1
personal_best_position(bug_num +
number_of_bugs*(shot_number-1),:)=bug_position(bug_num +
number_of_bugs*(shot_number-1),:);
shot_number=shot_number+1;
end
end;
97
%is it a global best?
if bug_fitness < global_best_fitness
global_best_fitness=bug_fitness;
shot_number=1;
while shot_number<number_of_shots+1
global_best_position(shot_number,:)=bug_position(bug_num+number_o
f_bugs*(shot_number-1),:);
global_best_position_constants(shot_number,:)=bug_position1(bug_n
um+number_of_bugs*(shot_number-1),:);
global_best_shot_fitness = shot_fitness;
shot_number=shot_number+1;
end
end
%the "two equations" for swarm optimization. update
velocity and position
shot_number=1;
while shot_number < number_of_shots+1
bug_velocity(bug_num+number_of_bugs*(shot_number1),:)= bug_velocity(bug_num+number_of_bugs*(shot_number1),:)*inertia +
rand(1)*accel_constant_1*(personal_best_position(bug_num+number_o
f_bugs*(shot_number-1),:)bug_position(bug_num+number_of_bugs*(shot_number-1),:)) +
rand(1)*accel_constant_2*(global_best_position(shot_number,:)bug_position(bug_num+number_of_bugs*(shot_number-1),:));
bug_velocity(bug_num+number_of_bugs*(shot_number1),:)=
constrain_velocity(bug_velocity,element_limits,bug_num,shot_numbe
r,number_of_bugs);
shot_number=shot_number+1;
end
shot_number=1;
while shot_number < number_of_shots+1
bug_position(bug_num+number_of_bugs*(shot_number1),:)= bug_position(bug_num+number_of_bugs*(shot_number-1),:) +
bug_velocity(bug_num+number_of_bugs*(shot_number-1),:);
[bug_position(bug_num+number_of_bugs*(shot_number1),:),element_limits]=
constrain_position(bug_position,element_limits,bug_num,shot_numbe
r,number_of_bugs);
shot_number=shot_number+1;
end
%output to screen
simulation_number
generation
bug_num
98
global_best_fitness
bug_num=bug_num+1;
end;
close all;
shot_number=1;
while shot_number<number_of_shots+1
%Offset and get training data %%%%
training_data_saved_temp=zeros(floor(training_simend/samplerate),
4);
n=1;
while n<5
%
training_data_saved_temp(:,n)=training_data_saved(:,n+4*(shot_num
ber-1));
training_data(:,n)=training_data_saved(:,n+4*(shot_number-1));
n=n+1;
end
%
training_data=offset_training_data(training_data_saved_temp,globa
l_best_position,shot_number,simend,samplerate,1,number_of_bugs);
%%%%%%%%%%%%%%%%%%%%%%%%%%
swarm_data1 =
simulation(global_best_position,global_best_position_constants,1,
samplerate,simend,1,shot_number,time_delay,V_charge);
if show_figure==0
else
% Output best swarm solution so far and compare to
the training data
figure;
temp=get(0,'ScreenSize');
if show_figure==1
set(gcf,'Position',[1 1 10 10])
else
set(gcf,'Position',[10 10 temp(3)*.8 temp(4)*.8])
end
if train_on_Vce==1
subplot(2,2,1)
plot(swarm_data1(:,1)*1e6,swarm_data1(:,2),'b')
99
hold on;
grid on;
plot(training_data(:,1)*1e6,training_data(:,2),'r')
legend('swarm solution','Raw Probe Data')
xlabel('Time(us)')
ylabel('Voltage(V)')
title('Vce')
axis([0 simend*1e6 0.9*min(training_data(:,2))
1.1*max(training_data(:,2))])
end
if train_on_Vge==1
subplot(2,2,2)
plot(swarm_data1(:,1)*1e6,swarm_data1(:,3),'b')
hold on;
grid on;
plot(training_data(:,1)*1e6,training_data(:,3),'r')
legend('swarm solution','Raw Probe Data')
xlabel('Time(us)')
ylabel('Voltage(V)')
title('Vge')
axis([0 simend*1e6 0.9*min(training_data(:,3))
1.1*max(training_data(:,3))])
end;
if train_on_Ic==1
subplot(2,2,3)
plot(swarm_data1(:,1)*1e6,swarm_data1(:,4),'b')
hold on;
grid on;
plot(training_data(:,1)*1e6,training_data(:,4),'r')
legend('swarm solution','Raw Probe Data')
xlabel('Time(us)')
ylabel('Current(A)')
title('Ic')
axis([0 simend*1e6 0.9*min(training_data(:,4))
1.1*max(training_data(:,4))])
text(0,max(training_data(:,4))/7,strcat(mat2str(global_best_shot_fitness
(shot_number,1)/1e8,4), ' e8'))
text(0,max(training_data(:,4))/4,mat2str(global_best_position(shot_numbe
r,:),3))
end;
pause(1);
%pause for a second
end
100
save(strcat('C:\Jim\training',num2str(shot_number),'.tab'),'train
ing_data','-tabs','-ascii');
save(strcat('C:\Jim\swarm',num2str(shot_number),'.tab'),'swarm_da
ta1','-tabs','-ascii');
shot_number=shot_number+1;
end
generation;
generation=generation+1;
end;
101
function bug_position =
initialize_bugs(number_of_bugs,number_of_shots,element_limits)
% This function initializes the bug positions. The method we
chose is to randomly
% choose values between the predefined limits. The returned
vector is the initialized
% positions.
temp=size(element_limits);
number_of_elements=temp(2);
bug_position(number_of_bugs*number_of_shots,number_of_elements)=0
;
bug_position2(number_of_bugs*number_of_shots,number_of_elements)=
0;
VT = [12 12 12 12]; %added
TAU = [2e-5 2e-5 2e-5 2e-5]; %added
NB = [2e15 2e15 2e15 2e15]; %added
THETA = [0.5 0.5 0.5 0.5]; %added
shot_number=1;
while shot_number<number_of_shots+1
bug_num=1;
while bug_num<number_of_bugs+1
%Do for every element in the row
n=1;
while n < number_of_elements+1
bug_position2(bug_num +(shot_number1)*number_of_bugs,n)=unifrnd(element_limits(shot_number*2,n),elem
ent_limits(shot_number*2-1,n));
n=n+1;
end;
n=1;
bug_position2(bug_num +(shot_number1)*number_of_bugs,n)=unifrnd(element_limits(shot_number*2,n),bug_
position2(bug_num,6));
n=3; %added
random_num=randi([1 4],1,1,'double'); %added
bug_position2(bug_num +(shot_number1)*number_of_bugs,n)= VT(1, random_num); %added
n=6; %added
random_num=randi([1 4],1,1,'double'); %added
bug_position2(bug_num +(shot_number1)*number_of_bugs,n)= TAU(1, random_num); %added
n=9; %added
random_num=randi([1 4],1,1,'double');
102
%added
bug_position2(bug_num +(shot_number1)*number_of_bugs,n)= NB(1, random_num); %added
n=10; %added
random_num=randi([1 4],1,1,'double'); %added
bug_position2(bug_num +(shot_number1)*number_of_bugs,n)= THETA(1, random_num); %added
bug_num=bug_num+1;
end;
shot_number=shot_number+1;
end
bug_position=bug_position2;
103
function
[temp1,new_element_limits]=constrain_position(bug_position,elemen
t_limits,bug_num,shot_number,number_of_bugs)
temp=size(element_limits);
number_of_elements=temp(2);
element_num=1;
while element_num <number_of_elements+1
if bug_position(bug_num+number_of_bugs*(shot_number1),element_num)>element_limits(shot_number*2-1,element_num)
bug_position(bug_num+number_of_bugs*(shot_number1),element_num)=element_limits(shot_number*2-1,element_num);
end;
if bug_position(bug_num+number_of_bugs*(shot_number1),element_num)<element_limits(shot_number*2,element_num)
bug_position(bug_num+number_of_bugs*(shot_number1),element_num)=element_limits(shot_number*2,element_num);
end;
element_num=element_num+1;
end;
element_limits(shot_number*2-1,1)=
bug_position(bug_num+number_of_bugs*(shot_number-1),6);
upper limit of AGD to AREA
new_element_limits = element_limits;
%set
if bug_position(bug_num+number_of_bugs*(shot_number1),1)>element_limits(shot_number*2-1,1)
bug_position(bug_num+number_of_bugs*(shot_number1),1)=element_limits(shot_number*2-1,1);
end;
temp1=bug_position(bug_num+number_of_bugs*(shot_number-1),:);
104
function
temp1=constrain_velocity(bug_velocity,element_limits,bug_num,shot
_number,number_of_bugs)
temp=size(element_limits);
number_of_elements=temp(2);
element_num=1;
while element_num <number_of_elements + 1
lowerl=element_limits(shot_number*2,element_num);
upperl=element_limits(shot_number*2-1,element_num);
if bug_velocity(bug_num+number_of_bugs*(shot_number1),element_num)>lowerl*10^(log10(upperl/lowerl)/3)
bug_velocity(bug_num+number_of_bugs*(shot_number1),element_num)=lowerl*10^(log10(upperl/lowerl)/3);
end;
if bug_velocity(bug_num+number_of_bugs*(shot_number1),element_num)<-lowerl*10^(log10(upperl/lowerl)/3)
bug_velocity(bug_num+number_of_bugs*(shot_number1),element_num)=-lowerl*10^(log10(upperl/lowerl)/3);
end;
element_num=element_num+1;
end;
temp1=bug_velocity(bug_num+number_of_bugs*(shot_number-1),:);
105
function sim_data =
write_netlist(bug_position,bug_position1,bug_num,samplerate,simen
d,number_of_bugs,shot_number,time_delay,V_charge)
file_name=sprintf('c:\\Jim\\bug_files\\test%03d.ckt', bug_num +
number_of_bugs*(shot_number-1));
fid=fopen(file_name,'w');
fprintf(fid,'Test circuit');
fprintf(fid,'\r\n');
%%%%%%%%%% IGBT Test Circuit %%%%%%%%%%%%
fprintf(fid, 'C1 Vin 0 %18.18f IC=%18.18f', 30e-6, V_charge);
fprintf(fid, '\r\n');
fprintf(fid, 'C2 6 Ve %18.18f', 11.75e-9);
fprintf(fid, '\r\n');
fprintf(fid, 'D1 4 Vin $GENERIC');
fprintf(fid, '\r\n');
fprintf(fid, 'D2 Vc 6 $GENERIC');
fprintf(fid, '\r\n');
fprintf(fid, 'D3 Ve Vc $GENERIC');
fprintf(fid, '\r\n');
fprintf(fid, 'L1 8 Vg %18.18f', bug_position1(bug_num +
number_of_bugs*(shot_number-1),9));
fprintf(fid, '\r\n');
fprintf(fid, 'L2 VL- Vc %18.18f', bug_position1(bug_num +
number_of_bugs*(shot_number-1),8));
fprintf(fid, '\r\n');
fprintf(fid, 'RL Vin VL- %18.18f', bug_position(bug_num +
number_of_bugs*(shot_number-1),11));
fprintf(fid, '\r\n');
fprintf(fid, 'Rsnub 6 Vc %18.18f', 1e3);
fprintf(fid, '\r\n');
fprintf(fid, 'Rfback 0 Ve %18.18f', 2e-3);
fprintf(fid, '\r\n');
fprintf(fid, 'Rbleed 0 Vin %18.18f', 200e6);
fprintf(fid, '\r\n');
fprintf(fid, 'Rg 3 8 %18.18f', 50);
fprintf(fid, '\r\n');
fprintf(fid, 'V1 3 0 DC %18.18f AC %18.18f %18.18f PULSE %18.18f
%18.18f %18.18f %18.18f %18.18f %18.18f %18.18f', 0, 0, 0, 0, 25,
time_delay, 10e-9, 10e-9, 10e-6, 50e-6);
fprintf(fid, '\r\n');
fprintf(fid, 'V2 4 0 DC %18.18f AC %18.18f %18.18f', V_charge, 1,
0);
fprintf(fid, '\r\n');
fprintf(fid, 'Z1 Vc Vg Ve %s', 'JIMTEST');
fprintf(fid, '\r\n');
fprintf(fid, '*');
fprintf(fid, '\r\n');
106
fprintf(fid, '.MODEL $GENERIC D (AF=%18.18f BV=%18.18f
CJO=%18.18f EG=%18.18f FC=%18.18f IBV=%18.18f IBVL=%18.18f', 1,
5400, 2.5e-12, 1.11, 500e-3, 100e-12, 0);
fprintf(fid, '\r\n');
fprintf(fid, '+ IKF=%18.18f IS=%18.18f ISR=%18.18f KF=%18.18f
M=%18.18f N=%18.18f NBV=%18.18f NBVL=%18.18f NR=%18.18f
RS=%18.18f TBV1=%18.18f TBV2=%18.18f', 0, 8e-9, 0, 0, 10e-3, 2,
1, 1, 2, 400e-3, 0, 0);
fprintf(fid, '\r\n');
fprintf(fid, '+ TIKF=%18.18f TRS1=%18.18f TRS2=%18.18f TT=%18.18f
VJ=%18.18f XTI=%18.18f)', 0, 0, 0, 1e-9, 700e-3, 3);
fprintf(fid, '\r\n');
fprintf(fid, '.MODEL JIMTEST NIGBT (AGD=%18.18f AREA=%18.18f
BVF=%18.18f BVN=%18.18f CGS=%18.18f', bug_position(bug_num +
number_of_bugs*(shot_number-1),1), bug_position1(bug_num +
number_of_bugs*(shot_number-1),3), bug_position1(bug_num +
number_of_bugs*(shot_number-1),6), bug_position1(bug_num +
number_of_bugs*(shot_number-1),1), bug_position(bug_num +
number_of_bugs*(shot_number-1),4));
fprintf(fid, '\r\n');
fprintf(fid, '+ COXD=%18.18f JSNE=%18.18f KF=%18.18f KP=%18.18f
MUN=%18.18f MUP=%18.18f NB=%18.18f', bug_position(bug_num +
number_of_bugs*(shot_number-1),7), bug_position1(bug_num +
number_of_bugs*(shot_number-1),2), bug_position1(bug_num +
number_of_bugs*(shot_number-1),4), bug_position(bug_num +
number_of_bugs*(shot_number-1),8), bug_position(bug_num +
number_of_bugs*(shot_number-1),2), bug_position(bug_num +
number_of_bugs*(shot_number-1),5), bug_position(bug_num +
number_of_bugs*(shot_number-1),9));
fprintf(fid, '\r\n');
fprintf(fid, '+ TAU=%18.18f THETA=%18.18f VT=%18.18f VTD=-%18.18f
WB=%18.18f)', bug_position(bug_num + number_of_bugs*(shot_number1),6), bug_position(bug_num + number_of_bugs*(shot_number-1),10),
bug_position(bug_num + number_of_bugs*(shot_number-1),3),
bug_position1(bug_num + number_of_bugs*(shot_number-1),5),
bug_position1(bug_num + number_of_bugs*(shot_number-1),7));
fprintf(fid, '\r\n');
fprintf(fid, '*');
fprintf(fid, '\r\n');
fprintf(fid, '.OPTIONS ACCT LIST OPTS ABSTOL=%18.18f
CHGTOL=%18.18f DEFL=%18.18f DEFW=%18.18f DEFNRD=%18.18f', 1e-6,
1e-9, 100e-6, 100e-6, 0);
fprintf(fid, '\r\n');
fprintf(fid, '+ DEFNRS=%18.18f DEFPD=%18.18f DEFPS=%18.18f
DIGDRVF=%18.18f DIGDRVZ=%18.18f DIGERRDEFAULT=%18.18f
DIGERRLIMIT=%18.18f', 0, 0, 0, 2, 20e3, 20, 0);
fprintf(fid, '\r\n');
fprintf(fid, '+ DIGFREQ=%18.18f DIGINITSTATE=%18.18f
DIGIOLVL=%18.18f DIGMNTYMX=%18.18f DIGMNTYSCALE=%18.18f
DIGOVRDRV=%18.18f', 10e9, 0, 2, 2, 0.4, 3);
fprintf(fid, '\r\n');
107
fprintf(fid, '+ DIGTYMXSCALE=%18.18f GMIN=%18.18f ITL1=%18.18f
ITL2=%18.18f ITL4=%18.18f PIVREL=%18.18f PIVTOL=%18.18f', 1.6,
1e-9, 200, 50, 50, 1e-3, 0.1e-12);
fprintf(fid, '\r\n');
fprintf(fid, '+ RELTOL=%18.18f TNOM=%18.18f TRTOL=%18.18f
VNTOL=%18.18f WIDTH=%18.18f', 10e-3, 27, 7, 1e-3, 80);
fprintf(fid, '\r\n');
fprintf(fid, '*');
fprintf(fid, '\r\n');
fprintf(fid, '.LIB %s', '"C:\Program Files\Spectrum
Software\MC9\library\NOM.LIB"');
fprintf(fid, '\r\n');
fprintf(fid, '*');
fprintf(fid, '\r\n');
fprintf(fid, '.TEMP %18.18f', 27);
fprintf(fid, '\r\n');
fprintf(fid, '*');
fprintf(fid, '\r\n');
fprintf(fid, '.TRAN %18.18f %18.18f', 25e-6/2500, 24.999e-6);
fprintf(fid, '\r\n');
fprintf(fid, '.PRINT TRAN (V([VC])-V([VE])) (V([VG])-V([VE]))
I(RL)');
fprintf(fid, '\r\n');
fprintf(fid, '*');
fprintf(fid, '\r\n');
fprintf(fid, '.PROBE');
fprintf(fid, '\r\n');
fprintf(fid, '.END');
fprintf(fid, '\r\n');
fprintf(fid, '%s', ';$SpiceType=PSPICE');
fprintf(fid, '\r\n');
fclose(fid);
%%%%%%%%%%%%%%%%%Troubleshooting%%%%%%%%%%%%%%%%%%%%%%%%%
% if bug_position(bug_num,1) <= bug_position(bug_num,6)
%
system('mc9 @test.bat');
%
% else
%
sim_data = ones(2500,4)*-5e9;
%
bad = 0
%
% end
%
%
% if exist('C:\Jim\test.tno')==0
%
sim_data = ones(2500,4)*-5e9;
%
bad = 1
%
if bug_position(bug_num,1) > bug_position(bug_num,6)
%
agd = 0
%
end
% else
108
%
%
%
%
% %
%
%
%
%
%
%
%
%
%
% end
%
%
%
sim_data = load('-ascii', 'C:\Jim\test.TNO');
if bug_position(bug_num,1) > bug_position(bug_num,6)
agd = 2
end
size(sim_data)
if size(sim_data) == [2500,4]
else
sim_data = ones(2500,4)*-5e9;
bad = 2
if bug_position(bug_num,1) > bug_position(bug_num,6)
agd = 1
end
end
109
function batch = test_bat(number_of_bugs_n_shots)
fid=fopen('c:\Program Files\Spectrum Software\MC9\test.bat','w');
fprintf(fid,'@noecho');
fprintf(fid,'\r\n');
%%%%%%%%%% Batch File to Run Bugs in Micro-Cap 9 %%%%%%%%%%%%
n=1;
while n < number_of_bugs_n_shots+1
fprintf(fid, 'c:\\Jim\\bug_files\\test%03d /T /S',n);
fprintf(fid, '\r\n');
n = n+1;
end
fclose(fid);
110
function sim_data =
simulation(bug_position,bug_position1,bug_num,samplerate,simend,n
umber_of_bugs,shot_number,time_delay,V_charge)
file_name=sprintf('c:\\Jim\\test_final.ckt');
fid=fopen(file_name,'w');
fprintf(fid,'Generational Final Test Circuit');
fprintf(fid,'\r\n');
%%%%%%%%%% IGBT Test Circuit %%%%%%%%%%%%
fprintf(fid, 'C1 Vin 0 %18.18f IC=%18.18f', 30e-6, V_charge);
fprintf(fid, '\r\n');
fprintf(fid, 'C2 6 Ve %18.18f', 11.75e-9);
fprintf(fid, '\r\n');
fprintf(fid, 'D1 4 Vin $GENERIC');
fprintf(fid, '\r\n');
fprintf(fid, 'D2 Vc 6 $GENERIC');
fprintf(fid, '\r\n');
fprintf(fid, 'D3 Ve Vc $GENERIC');
fprintf(fid, '\r\n');
fprintf(fid, 'L1 8 Vg %18.18f', bug_position1(bug_num +
(shot_number-1),9));
fprintf(fid, '\r\n');
fprintf(fid, 'L2 VL- Vc %18.18f', bug_position1(bug_num +
(shot_number-1),8));
fprintf(fid, '\r\n');
fprintf(fid, 'RL Vin VL- %18.18f', bug_position(bug_num +
(shot_number-1),11));
fprintf(fid, '\r\n');
fprintf(fid, 'Rsnub 6 Vc %18.18f', 1e3);
fprintf(fid, '\r\n');
fprintf(fid, 'Rfback 0 Ve %18.18f', 2e-3);
fprintf(fid, '\r\n');
fprintf(fid, 'Rbleed 0 Vin %18.18f', 200e6);
fprintf(fid, '\r\n');
fprintf(fid, 'Rg 3 8 %18.18f', 50);
fprintf(fid, '\r\n');
fprintf(fid, 'V1 3 0 DC %18.18f AC %18.18f %18.18f PULSE %18.18f
%18.18f %18.18f %18.18f %18.18f %18.18f %18.18f', 0, 0, 0, 0, 25,
time_delay, 10e-9, 10e-9, 10e-6, 50e-6);
fprintf(fid, '\r\n');
fprintf(fid, 'V2 4 0 DC %18.18f AC %18.18f %18.18f', V_charge, 1,
0);
fprintf(fid, '\r\n');
fprintf(fid, 'Z1 Vc Vg Ve %s', 'JIMTEST');
fprintf(fid, '\r\n');
fprintf(fid, '*');
fprintf(fid, '\r\n');
111
fprintf(fid, '.MODEL $GENERIC D (AF=%18.18f BV=%18.18f
CJO=%18.18f EG=%18.18f FC=%18.18f IBV=%18.18f IBVL=%18.18f', 1,
5400, 2.5e-12, 1.11, 500e-3, 100e-12, 0);
fprintf(fid, '\r\n');
fprintf(fid, '+ IKF=%18.18f IS=%18.18f ISR=%18.18f KF=%18.18f
M=%18.18f N=%18.18f NBV=%18.18f NBVL=%18.18f NR=%18.18f
RS=%18.18f TBV1=%18.18f TBV2=%18.18f', 0, 8e-9, 0, 0, 10e-3, 2,
1, 1, 2, 400e-3, 0, 0);
fprintf(fid, '\r\n');
fprintf(fid, '+ TIKF=%18.18f TRS1=%18.18f TRS2=%18.18f TT=%18.18f
VJ=%18.18f XTI=%18.18f)', 0, 0, 0, 1e-9, 700e-3, 3);
fprintf(fid, '\r\n');
fprintf(fid, '.MODEL JIMTEST NIGBT (AGD=%18.18f AREA=%18.18f
BVF=%18.18f BVN=%18.18f CGS=%18.18f', bug_position(bug_num +
(shot_number-1),1), bug_position1(bug_num + (shot_number-1),3),
bug_position1(bug_num + (shot_number-1),6), bug_position1(bug_num
+ (shot_number-1),1), bug_position(bug_num + (shot_number-1),4));
fprintf(fid, '\r\n');
fprintf(fid, '+ COXD=%18.18f JSNE=%18.18f KF=%18.18f KP=%18.18f
MUN=%18.18f MUP=%18.18f NB=%18.18f', bug_position(bug_num +
(shot_number-1),7), bug_position1(bug_num + (shot_number-1),2),
bug_position1(bug_num + (shot_number-1),4), bug_position(bug_num
+ (shot_number-1),8), bug_position(bug_num + (shot_number-1),2),
bug_position(bug_num + (shot_number-1),5), bug_position(bug_num +
(shot_number-1),9));
fprintf(fid, '\r\n');
fprintf(fid, '+ TAU=%18.18f THETA=%18.18f VT=%18.18f VTD=-%18.18f
WB=%18.18f)', bug_position(bug_num + (shot_number-1),6),
bug_position(bug_num + (shot_number-1),10), bug_position(bug_num
+ (shot_number-1),3), bug_position1(bug_num + (shot_number-1),5),
bug_position1(bug_num + (shot_number-1),7));
fprintf(fid, '\r\n');
fprintf(fid, '*');
fprintf(fid, '\r\n');
fprintf(fid, '.OPTIONS ACCT LIST OPTS ABSTOL=%18.18f
CHGTOL=%18.18f DEFL=%18.18f DEFW=%18.18f DEFNRD=%18.18f', 1e-6,
1e-9, 100e-6, 100e-6, 0);
fprintf(fid, '\r\n');
fprintf(fid, '+ DEFNRS=%18.18f DEFPD=%18.18f DEFPS=%18.18f
DIGDRVF=%18.18f DIGDRVZ=%18.18f DIGERRDEFAULT=%18.18f
DIGERRLIMIT=%18.18f', 0, 0, 0, 2, 20e3, 20, 0);
fprintf(fid, '\r\n');
fprintf(fid, '+ DIGFREQ=%18.18f DIGINITSTATE=%18.18f
DIGIOLVL=%18.18f DIGMNTYMX=%18.18f DIGMNTYSCALE=%18.18f
DIGOVRDRV=%18.18f', 10e9, 0, 2, 2, 0.4, 3);
fprintf(fid, '\r\n');
fprintf(fid, '+ DIGTYMXSCALE=%18.18f GMIN=%18.18f ITL1=%18.18f
ITL2=%18.18f ITL4=%18.18f PIVREL=%18.18f PIVTOL=%18.18f', 1.6,
1e-9, 200, 50, 50, 1e-3, 0.1e-12);
fprintf(fid, '\r\n');
fprintf(fid, '+ RELTOL=%18.18f TNOM=%18.18f TRTOL=%18.18f
VNTOL=%18.18f WIDTH=%18.18f', 10e-3, 27, 7, 1e-3, 80);
112
fprintf(fid, '\r\n');
fprintf(fid, '*');
fprintf(fid, '\r\n');
fprintf(fid, '.LIB %s', '"C:\Program Files\Spectrum
Software\MC9\library\NOM.LIB"');
fprintf(fid, '\r\n');
fprintf(fid, '*');
fprintf(fid, '\r\n');
fprintf(fid, '.TEMP %18.18f', 27);
fprintf(fid, '\r\n');
fprintf(fid, '*');
fprintf(fid, '\r\n');
fprintf(fid, '.TRAN %18.18f %18.18f', 25e-6/2500, 24.999e-6);
fprintf(fid, '\r\n');
fprintf(fid, '.PRINT TRAN (V([VC])-V([VE])) (V([VG])-V([VE]))
I(RL)');
fprintf(fid, '\r\n');
fprintf(fid, '*');
fprintf(fid, '\r\n');
fprintf(fid, '.PROBE');
fprintf(fid, '\r\n');
fprintf(fid, '.END');
fprintf(fid, '\r\n');
fprintf(fid, '%s', ';$SpiceType=PSPICE');
fprintf(fid, '\r\n');
fclose(fid);
if bug_position(bug_num + (shot_number-1),1) <=
bug_position1(bug_num + (shot_number-1),3)
system('mc9 @test_final.bat');
else
sim_data = ones(2500,4)*-5e9;
bad = 0
end
if exist('C:\Jim\test_final.tno')==0
sim_data = ones(2500,4)*-5e9;
bad = 1
if bug_position(bug_num + (shot_number-1),1) >
bug_position(bug_num + (shot_number-1),6)
agd = 0
end
else
sim_data = load('-ascii', 'C:\Jim\test_final.TNO');
if bug_position(bug_num + (shot_number-1),1) >
bug_position(bug_num + (shot_number-1),6)
agd = 2
end
%
size(sim_data)
113
if size(sim_data) == [2500,4]
else
sim_data = ones(2500,4)*-5e9;
bad = 2
if bug_position(bug_num + (shot_number-1),1) >
bug_position1(bug_num + (shot_number-1),3)
agd = 1
end
end
end
114
B. IGBT Test Circuit Bill of Materials
Table B.1: Bill of materials for IGBT test circuit and gate drive circuit
ITEM QTY REF DES
1
15 R1-R15
DESCRIPTION
VENDOR
WSR Resistors for load
Vishay-Dale
VALUE
1
PART NO
WSR21R000FEA
2
1
Rbleed
2Mohm, 5kV resistor
Ohmite
3
2
Rsnub
500ohm, 2kV resistor
Riedon
500
NPS 2-T126 500.000 OHM 1%
4
1
Rfbac k
2mohm WSR resistor for emitter feedback resistor
Vishay-Dale
2m
WSR32L000FEA
5
1
Rcvr
1mohm WSR resistor for CVR
Vishay-Dale
1m
WSR31L000FEA
6
1
R16
Potentiometer
Bourns
10k
3224W-1-103E
7
1
R17
1/4W, 240 ohm, 1206 case resistor
Rohm
240
MCR18EZHF2400
8
1
R18
1/4W, 560 ohm, 1206 case resistor
Rohm
560
MCR18EZHF5600
9
1
R19
1/4W, 280 ohm, 1206 case resistor
Rohm
280
MCR18EZHF2800
10
9
D
Dprotect1-3, Dsnub1-3, Dap1-3, 63A, 1800V
IXYS
DSDI60-18A
11
4
D1-D4
50V, 1A diodes
Fairchild
S1A
12
3
D5-D7
75V, 150mA diodes
Comchip Tech
13
1
Cin
Custom SBE cap
SBE, Inc
30u
14
4
Csnub
47nF, 1kV caps
AVX
47n
15
3
C1-C3
1uF, 50V ceramic, 1206 case
AVX
1u
16
1
C4
0.1uF, 50V c eramic, 1206 cas e
Kemet
0.1u
C1206C104K5RACTU
17
2
C5-C6
15uF, 35V tantalum, 7343 case
Kemet
15u
T491D156K035AT
18
6
C7-C12
10uF, 35V tantalum, 7343 case
Kemet
10u
19
1
Q1
IGBT
Powerex
20
1
U1
5V linear voltage regulator
Linear Tech
LT1121CST-5
21
1
U2
1.2-37V voltage regulator D2 Pak, SMD-220-3, TO-263-3
STMicroelectronics
LM317D2T-TR
22
1
U3
optical receiver
Avago
HFBR-2412Z
23
1
U4
Inverting gate driver
Micrel
MIC4451
24
2
J1-J2
4.5kV input
Phoenix Contact
1986628
25
1
J3
24V input
Phoenix Contact
1711026
26
2
J4-J5
VCE connectors
Phoenix Contact
1986628
27
5
Loops to connect leads on
Keystone Electronics
1040
Total Parts Used: 72
115
2meg MC102822004JE
CDSF4148
2220AC473KAT1A
12065G105ZAT2A
T491D106K035AT
QIS4506001
C. IGBT Test Circuit Printed Circuit Board Layout
Figure C.1: IGBT test circuit schematic that correlates directly to the printed circuit
board layout
116
Figure C.2: Gate drive circuit schematic that correlates directly to the printed circuit
board layout
117
Figure C.3: Circuit board layout shown with input capacitor that extends off the
board
118
Figure C.4: PCB top copper layer and silkscreen
119
Figure C.5: PCB bottom copper layer and silkscreen
120
D. Raw Data Tables
121
Table D.1: Powerex IGBT #1 raw data table
3Ω
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
304
304
304
304
304
304
304
304
304
304
592
592
592
592
592
592
592
592
592
592
720
720
720
720
720
720
720
720
720
720
736
736
736
736
736
736
736
736
736
736
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
AGD
MUN
CGS
MUP
COXD
KP
1.34E-05 5.02E+07 3.06E-08 4.26E+04 2.02E-02 1.66E+01
8.75E-06 4.97E+07 2.70E-08 3.84E+04 1.93E-02 1.67E+01
1.23E-05 4.87E+07 2.92E-08 4.19E+04 1.96E-02 1.62E+01
1.03E-05 4.75E+07 2.99E-08 3.90E+04 1.94E-02 1.62E+01
1.32E-05 5.04E+07 2.94E-08 3.94E+04 1.83E-02 1.66E+01
9.10E-06 4.89E+07 3.15E-08 4.18E+04 1.91E-02 1.62E+01
1.21E-05 5.03E+07 3.00E-08 3.97E+04 1.89E-02 1.65E+01
9.44E-06 5.42E+07 3.12E-08 4.16E+04 1.96E-02 1.57E+01
1.53E-05 4.91E+07 3.02E-08 4.09E+04 1.90E-02 1.59E+01
1.48E-05 5.18E+07 3.10E-08 3.89E+04 1.95E-02 1.65E+01
1.86E-05 5.24E+07 9.20E-08 8.39E+04 1.55E-02 2.68E+01
1.82E-05 5.40E+07 8.58E-08 8.41E+04 1.57E-02 2.74E+01
1.84E-05 6.29E+07 9.29E-08 8.51E+04 1.55E-02 2.67E+01
1.99E-05 6.30E+07 8.02E-08 9.72E+04 1.64E-02 2.42E+01
1.82E-05 5.57E+07 9.02E-08 9.01E+04 1.64E-02 2.56E+01
1.94E-05 5.44E+07 8.50E-08 8.76E+04 1.58E-02 2.51E+01
1.82E-05 6.22E+07 7.85E-08 9.33E+04 1.60E-02 2.43E+01
1.85E-05 5.71E+07 9.06E-08 9.31E+04 1.63E-02 2.52E+01
1.86E-05 5.65E+07 9.20E-08 8.73E+04 1.46E-02 2.63E+01
1.89E-05 5.68E+07 8.66E-08 9.53E+04 1.55E-02 2.43E+01
1.98E-05 5.24E+07 1.15E-07 7.10E+04 3.58E-02 3.63E+01
1.88E-05 5.18E+07 1.20E-07 7.69E+04 3.82E-02 3.46E+01
1.90E-05 5.21E+07 1.13E-07 7.53E+04 3.92E-02 3.59E+01
1.92E-05 4.89E+07 1.20E-07 7.77E+04 3.84E-02 3.48E+01
2.00E-05 5.23E+07 1.08E-07 7.61E+04 3.91E-02 3.49E+01
1.90E-05 5.18E+07 1.10E-07 7.32E+04 3.77E-02 3.61E+01
1.89E-05 5.12E+07 1.15E-07 7.41E+04 3.69E-02 3.54E+01
1.88E-05 4.95E+07 1.19E-07 7.26E+04 3.69E-02 3.59E+01
1.95E-05 5.48E+07 1.16E-07 7.89E+04 4.31E-02 3.52E+01
1.91E-05 5.07E+07 1.12E-07 7.31E+04 3.77E-02 3.53E+01
1.87E-05 5.89E+08 1.26E-07 7.19E+04 1.65E-01 3.96E+01
1.80E-05 6.16E+08 1.21E-07 7.77E+04 1.81E-01 3.78E+01
1.91E-05 5.60E+08 1.17E-07 7.15E+04 1.68E-01 3.79E+01
1.82E-05 5.68E+08 1.16E-07 7.42E+04 1.70E-01 3.73E+01
1.80E-05 6.04E+08 1.15E-07 7.06E+04 1.66E-01 3.78E+01
1.84E-05 5.54E+08 1.24E-07 6.80E+04 1.79E-01 3.99E+01
1.98E-05 5.23E+08 1.31E-07 7.89E+04 1.80E-01 3.44E+01
1.91E-05 6.01E+08 1.17E-07 7.17E+04 1.53E-01 3.73E+01
1.92E-05 5.80E+08 1.22E-07 7.79E+04 1.55E-01 3.66E+01
2.00E-05 6.32E+08 1.33E-07 6.97E+04 1.81E-01 3.83E+01
122
Table D.2: Powerex IGBT #1 raw data table (continued)
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
4.5 Ω 3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
206
206
206
206
206
206
206
206
206
206
412
412
412
412
412
412
412
412
412
412
600
600
600
600
600
600
600
600
600
600
704
704
704
704
704
704
704
704
704
704
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
AGD
MUN
CGS
MUP
COXD
KP
7.01E-06 3.48E+07 2.27E-08 1.64E+04 8.07E-03 1.55E+01
7.48E-06 3.74E+07 2.17E-08 1.54E+04 7.15E-03 1.57E+01
4.70E-06 3.84E+07 2.41E-08 1.51E+04 7.97E-03 1.57E+01
6.54E-06 3.41E+07 2.39E-08 1.59E+04 8.56E-03 1.60E+01
5.97E-06 3.85E+07 2.15E-08 1.45E+04 7.42E-03 1.62E+01
8.06E-06 3.49E+07 2.27E-08 1.42E+04 8.25E-03 1.60E+01
4.11E-06 3.71E+07 2.28E-08 1.46E+04 7.95E-03 1.63E+01
4.15E-06 3.72E+07 2.33E-08 1.65E+04 7.27E-03 1.54E+01
4.15E-06 3.76E+07 2.35E-08 1.44E+04 7.27E-03 1.60E+01
5.64E-06 3.71E+07 2.26E-08 1.46E+04 7.98E-03 1.66E+01
9.54E-06 3.06E+07 3.48E-08 4.94E+04 5.92E-04 2.46E+01
1.03E-05 3.25E+07 3.13E-08 5.23E+04 6.41E-04 2.38E+01
1.13E-05 3.02E+07 3.29E-08 4.71E+04 5.53E-04 2.52E+01
1.04E-05 3.22E+07 3.41E-08 5.11E+04 5.77E-04 2.38E+01
1.40E-05 3.12E+07 3.25E-08 5.22E+04 6.27E-04 2.36E+01
1.48E-05 3.28E+07 3.50E-08 5.08E+04 6.10E-04 2.44E+01
1.24E-05 3.39E+07 3.34E-08 4.94E+04 6.15E-04 2.45E+01
8.40E-06 3.35E+07 3.37E-08 5.34E+04 6.32E-04 2.36E+01
9.26E-06 3.11E+07 3.42E-08 4.98E+04 6.19E-04 2.46E+01
1.24E-05 2.96E+07 3.68E-08 5.17E+04 6.11E-04 2.42E+01
1.87E-05 2.80E+07 5.63E-08 6.69E+04 1.01E-03 3.48E+01
1.80E-05 3.06E+07 5.64E-08 7.32E+04 1.05E-03 3.32E+01
1.98E-05 2.97E+07 5.49E-08 7.05E+04 1.00E-03 3.37E+01
1.99E-05 2.74E+07 5.85E-08 7.06E+04 1.03E-03 3.41E+01
1.97E-05 3.08E+07 5.44E-08 6.21E+04 1.14E-03 3.65E+01
1.84E-05 2.97E+07 5.45E-08 7.04E+04 1.06E-03 3.36E+01
1.86E-05 2.69E+07 5.85E-08 6.75E+04 1.05E-03 3.49E+01
2.00E-05 2.74E+07 6.07E-08 6.31E+04 1.08E-03 3.64E+01
1.88E-05 2.71E+07 5.65E-08 6.86E+04 1.08E-03 3.41E+01
1.95E-05 2.62E+07 5.72E-08 6.86E+04 1.08E-03 3.46E+01
1.92E-05 1.61E+07 9.43E-08 7.07E+04 1.56E-03 4.05E+01
1.91E-05 1.56E+07 9.56E-08 7.53E+04 1.79E-03 3.91E+01
1.86E-05 1.52E+07 9.37E-08 7.61E+04 1.72E-03 3.87E+01
1.95E-05 1.51E+07 1.02E-07 7.39E+04 1.74E-03 3.98E+01
1.92E-05 1.54E+07 9.56E-08 7.77E+04 1.85E-03 3.81E+01
1.92E-05 1.45E+07 9.90E-08 7.14E+04 1.69E-03 4.07E+01
2.00E-05 1.56E+07 9.78E-08 7.98E+04 1.75E-03 3.74E+01
1.87E-05 1.54E+07 9.53E-08 7.83E+04 1.57E-03 3.77E+01
1.99E-05 1.59E+07 1.01E-07 7.14E+04 1.67E-03 4.08E+01
1.92E-05 1.43E+07 9.84E-08 7.52E+04 1.61E-03 3.92E+01
123
Table D.3: Powerex IGBT #1 raw data table (continued)
9Ω
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
105
105
105
105
105
105
105
105
105
105
212
212
212
212
212
212
212
212
212
212
312
312
312
312
312
312
312
312
312
312
384
384
384
384
384
384
384
384
384
384
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
AGD
MUN
CGS
MUP
COXD
KP
1.57E-05 1.58E+06 2.02E-08 2.98E+02 8.40E-04 1.04E+01
1.33E-05 1.58E+06 1.95E-08 2.81E+02 8.40E-04 1.01E+01
1.20E-05 1.54E+06 2.04E-08 2.93E+02 8.46E-04 1.04E+01
1.51E-05 1.61E+06 1.95E-08 3.03E+02 8.34E-04 1.04E+01
1.52E-05 1.51E+06 1.96E-08 3.00E+02 8.03E-04 1.02E+01
1.62E-05 1.51E+06 2.00E-08 3.00E+02 8.30E-04 1.06E+01
1.73E-05 1.56E+06 1.91E-08 2.92E+02 8.04E-04 1.04E+01
1.53E-05 1.51E+06 1.99E-08 2.92E+02 8.17E-04 1.03E+01
1.21E-05 1.59E+06 1.90E-08 3.02E+02 7.98E-04 1.03E+01
1.61E-05 1.60E+06 2.02E-08 2.97E+02 8.20E-04 1.05E+01
1.03E-05 8.36E+06 2.16E-08 1.00E+03 3.01E-04 2.11E+01
5.22E-06 9.08E+06 2.24E-08 1.12E+03 2.88E-04 2.07E+01
1.60E-05 8.99E+06 2.36E-08 1.13E+03 2.84E-04 2.10E+01
5.60E-06 8.33E+06 2.18E-08 1.15E+03 2.77E-04 2.08E+01
9.05E-06 9.25E+06 2.26E-08 1.15E+03 2.62E-04 2.11E+01
8.40E-06 8.95E+06 2.20E-08 1.05E+03 2.96E-04 2.12E+01
1.30E-05 8.29E+06 2.44E-08 1.14E+03 3.19E-04 2.09E+01
5.38E-06 8.53E+06 2.46E-08 1.02E+03 2.98E-04 2.07E+01
1.71E-05 7.91E+06 2.37E-08 1.13E+03 2.86E-04 2.12E+01
8.69E-06 8.90E+06 2.20E-08 1.10E+03 3.09E-04 2.12E+01
7.72E-06 5.17E+06 2.70E-08 5.96E+04 1.70E-04 1.54E+01
8.07E-06 5.60E+06 2.71E-08 6.44E+04 1.75E-04 1.47E+01
7.44E-06 5.41E+06 2.52E-08 6.16E+04 1.72E-04 1.49E+01
7.50E-06 5.74E+06 2.40E-08 6.60E+04 1.86E-04 1.43E+01
6.82E-06 5.10E+06 2.50E-08 6.70E+04 1.80E-04 1.41E+01
6.00E-06 5.89E+06 2.50E-08 6.51E+04 1.84E-04 1.40E+01
5.02E-06 5.34E+06 2.61E-08 6.55E+04 1.69E-04 1.38E+01
6.07E-06 4.91E+06 2.90E-08 6.55E+04 1.86E-04 1.40E+01
9.43E-06 5.54E+06 2.52E-08 6.34E+04 1.91E-04 1.47E+01
1.09E-05 5.34E+06 2.58E-08 6.52E+04 1.71E-04 1.48E+01
6.82E-06 4.49E+07 8.85E-09 6.87E+04 1.12E-03 1.74E+01
7.59E-06 5.00E+07 9.16E-09 6.81E+04 1.22E-03 1.71E+01
8.65E-06 4.25E+07 9.27E-09 6.82E+04 1.14E-03 1.75E+01
7.95E-06 4.17E+07 9.85E-09 6.54E+04 1.14E-03 1.81E+01
8.39E-06 4.59E+07 9.32E-09 7.50E+04 1.17E-03 1.63E+01
9.18E-06 4.77E+07 9.64E-09 6.89E+04 1.10E-03 1.72E+01
8.25E-06 4.52E+07 9.94E-09 6.81E+04 1.09E-03 1.72E+01
7.90E-06 4.79E+07 1.02E-08 7.01E+04 1.16E-03 1.71E+01
7.84E-06 4.28E+07 9.30E-09 7.74E+04 1.22E-03 1.56E+01
7.09E-06 4.57E+07 1.01E-08 7.85E+04 1.23E-03 1.57E+01
124
Table D.4: Powerex IGBT #1 raw data table (continued)
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
15 Ω 3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic Shot #
68.8
1
68.8
2
68.8
3
68.8
4
68.8
5
68.8
6
68.8
7
68.8
8
68.8
9
68.8
10
140
1
140
2
140
3
140
4
140
5
140
6
140
7
140
8
140
9
140
10
208
1
208
2
208
3
208
4
208
5
208
6
208
7
208
8
208
9
208
10
235
1
235
2
235
3
235
4
235
5
235
6
235
7
235
8
235
9
235
10
AGD
MUN
CGS
MUP
COXD
KP
1.02E-05 3.19E+06 2.10E-08 1.93E+04 1.57E-03 4.27E+00
2.00E-05 2.98E+06 2.20E-08 2.04E+04 1.42E-03 3.83E+00
1.28E-05 3.53E+06 1.88E-08 1.91E+04 1.36E-03 4.18E+00
6.26E-06 3.12E+06 1.97E-08 1.88E+04 1.46E-03 4.14E+00
6.37E-06 3.30E+06 1.99E-08 1.76E+04 1.42E-03 4.23E+00
1.33E-05 3.18E+06 1.96E-08 1.86E+04 1.53E-03 4.11E+00
1.45E-05 3.17E+06 2.16E-08 1.70E+04 1.57E-03 4.43E+00
6.60E-06 3.15E+06 1.80E-08 1.80E+04 1.53E-03 4.39E+00
1.69E-05 3.25E+06 2.19E-08 1.80E+04 1.44E-03 4.33E+00
1.06E-05 3.10E+06 2.20E-08 1.78E+04 1.29E-03 4.40E+00
8.73E-06 1.04E+06 1.85E-08 3.11E+04 4.00E-04 6.99E+00
8.73E-07 9.18E+05 2.01E-08 2.85E+04 4.44E-04 8.15E+00
1.02E-05 1.03E+06 1.91E-08 3.12E+04 3.76E-04 7.31E+00
1.57E-05 9.19E+05 2.04E-08 2.98E+04 3.96E-04 8.13E+00
8.95E-06 1.11E+06 2.04E-08 2.85E+04 4.29E-04 8.32E+00
1.13E-05 1.10E+06 1.79E-08 2.95E+04 3.85E-04 7.38E+00
6.58E-06 9.18E+05 1.67E-08 3.23E+04 4.44E-04 7.46E+00
1.27E-05 9.20E+05 1.95E-08 3.01E+04 4.22E-04 7.51E+00
9.66E-07 1.12E+06 2.04E-08 2.86E+04 4.46E-04 7.71E+00
5.51E-06 9.18E+05 1.86E-08 3.00E+04 4.14E-04 7.75E+00
9.78E-06 1.09E+06 1.20E-08 5.13E+04 7.13E-04 9.66E+00
8.85E-06 1.01E+06 1.14E-08 5.58E+04 7.57E-04 8.59E+00
6.21E-06 1.23E+06 1.14E-08 4.56E+04 8.02E-04 1.01E+01
4.39E-06 1.14E+06 1.23E-08 5.10E+04 6.64E-04 9.26E+00
3.95E-06 1.01E+06 1.14E-08 4.56E+04 7.99E-04 9.63E+00
2.44E-06 1.01E+06 1.14E-08 5.58E+04 8.02E-04 8.29E+00
5.98E-06 1.13E+06 1.14E-08 4.56E+04 6.56E-04 1.01E+01
5.80E-06 1.14E+06 1.34E-08 5.53E+04 7.40E-04 8.45E+00
5.67E-06 1.23E+06 1.14E-08 5.56E+04 6.91E-04 8.95E+00
4.65E-06 1.01E+06 1.14E-08 5.58E+04 7.38E-04 8.62E+00
5.77E-06 1.01E+07 2.16E-08 6.73E+04 5.51E-04 9.13E+00
5.49E-06 9.12E+06 2.13E-08 6.95E+04 5.55E-04 9.30E+00
6.35E-06 8.52E+06 2.03E-08 6.28E+04 5.97E-04 9.74E+00
6.58E-06 8.92E+06 2.01E-08 6.18E+04 5.26E-04 1.04E+01
6.65E-06 1.03E+07 1.96E-08 7.19E+04 5.56E-04 9.45E+00
5.64E-06 9.28E+06 1.99E-08 6.41E+04 5.94E-04 1.01E+01
7.34E-06 1.03E+07 2.13E-08 6.32E+04 5.11E-04 1.01E+01
2.98E-06 8.44E+06 1.84E-08 6.62E+04 5.69E-04 9.74E+00
7.34E-06 9.93E+06 1.95E-08 6.09E+04 5.17E-04 1.01E+01
6.61E-06 9.43E+06 1.98E-08 6.69E+04 5.39E-04 9.70E+00
125
Table D.5: Powerex IGBT #2 raw data table
3Ω
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
304
304
304
304
304
304
304
304
304
304
596
596
596
596
596
596
596
596
596
596
720
720
720
720
720
720
720
720
720
720
736
736
736
736
736
736
736
736
736
736
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
AGD
MUN
CGS
MUP
COXD
KP
6.76E-06 5.61E+07 2.82E-08 4.71E+04 2.02E-02 1.55E+01
9.90E-06 5.87E+07 3.09E-08 3.86E+04 1.87E-02 1.68E+01
1.32E-05 5.79E+07 3.13E-08 4.48E+04 1.77E-02 1.50E+01
1.06E-05 5.66E+07 3.09E-08 4.25E+04 1.75E-02 1.62E+01
1.59E-05 5.90E+07 3.07E-08 4.21E+04 1.88E-02 1.61E+01
1.40E-05 5.95E+07 3.25E-08 4.12E+04 1.79E-02 1.59E+01
6.35E-06 5.56E+07 2.89E-08 4.23E+04 1.92E-02 1.55E+01
1.40E-05 6.00E+07 3.36E-08 4.07E+04 1.92E-02 1.60E+01
6.89E-06 5.56E+07 3.36E-08 4.19E+04 1.90E-02 1.62E+01
1.30E-05 5.94E+07 3.12E-08 4.47E+04 1.71E-02 1.54E+01
1.85E-05 6.49E+07 1.01E-07 8.66E+04 3.05E-02 2.64E+01
1.88E-05 6.94E+07 9.21E-08 8.95E+04 3.25E-02 2.59E+01
1.95E-05 6.26E+07 9.19E-08 8.90E+04 3.32E-02 2.63E+01
1.90E-05 6.58E+07 9.22E-08 8.74E+04 3.36E-02 2.59E+01
1.91E-05 6.30E+07 8.78E-08 8.71E+04 3.63E-02 2.60E+01
1.80E-05 6.74E+07 9.56E-08 8.89E+04 3.19E-02 2.64E+01
1.87E-05 6.35E+07 8.57E-08 8.77E+04 3.27E-02 2.65E+01
1.89E-05 6.14E+07 9.34E-08 9.02E+04 3.25E-02 2.55E+01
1.93E-05 5.77E+07 8.79E-08 8.84E+04 3.64E-02 2.64E+01
1.94E-05 6.21E+07 9.42E-08 9.13E+04 3.26E-02 2.52E+01
1.97E-05 6.15E+07 1.31E-07 6.96E+04 4.20E-02 3.80E+01
1.95E-05 5.62E+07 1.28E-07 7.92E+04 4.18E-02 3.48E+01
1.98E-05 5.15E+07 1.29E-07 8.09E+04 4.12E-02 3.41E+01
1.93E-05 5.65E+07 1.32E-07 7.44E+04 4.52E-02 3.63E+01
1.87E-05 5.54E+07 1.24E-07 7.46E+04 4.06E-02 3.67E+01
1.73E-05 5.61E+07 1.21E-07 7.02E+04 4.37E-02 3.72E+01
1.83E-05 5.53E+07 1.23E-07 7.92E+04 4.09E-02 3.48E+01
2.00E-05 5.45E+07 1.22E-07 7.89E+04 3.99E-02 3.43E+01
1.91E-05 5.89E+07 1.17E-07 7.22E+04 4.41E-02 3.63E+01
1.82E-05 5.56E+07 1.26E-07 7.87E+04 4.15E-02 3.49E+01
1.99E-05 5.64E+08 1.24E-07 6.49E+04 1.66E-01 3.98E+01
1.99E-05 5.47E+08 1.28E-07 7.69E+04 1.77E-01 3.52E+01
1.78E-05 5.94E+08 1.23E-07 6.83E+04 1.60E-01 4.01E+01
1.83E-05 5.38E+08 1.27E-07 7.26E+04 1.74E-01 3.67E+01
1.91E-05 5.57E+08 1.30E-07 7.10E+04 1.77E-01 3.73E+01
1.91E-05 5.89E+08 1.18E-07 7.04E+04 1.66E-01 3.82E+01
1.88E-05 5.92E+08 1.21E-07 7.12E+04 1.74E-01 3.77E+01
1.99E-05 5.87E+08 1.35E-07 6.39E+04 1.84E-01 4.04E+01
1.89E-05 5.53E+08 1.27E-07 7.36E+04 1.84E-01 3.67E+01
1.86E-05 6.33E+08 1.33E-07 7.68E+04 1.84E-01 3.62E+01
126
Table D.6: Powerex IGBT #2 raw data table (continued)
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
4.5 Ω 3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
212
212
212
212
212
212
212
212
212
212
424
424
424
424
424
424
424
424
424
424
624
624
624
624
624
624
624
624
624
624
688
688
688
688
688
688
688
688
688
688
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
AGD
MUN
CGS
MUP
COXD
KP
9.45E-06 3.44E+07 2.24E-08 1.28E+04 7.16E-03 1.44E+01
1.16E-05 3.60E+07 2.46E-08 1.45E+04 8.17E-03 1.49E+01
1.10E-05 3.68E+07 2.26E-08 1.52E+04 7.84E-03 1.55E+01
1.46E-05 3.47E+07 2.50E-08 1.30E+04 7.21E-03 1.60E+01
1.36E-05 3.46E+07 2.33E-08 1.27E+04 8.15E-03 1.64E+01
1.45E-05 3.39E+07 2.42E-08 1.37E+04 7.59E-03 1.55E+01
1.33E-05 3.31E+07 2.55E-08 1.36E+04 7.80E-03 1.52E+01
9.75E-06 3.63E+07 2.54E-08 1.32E+04 8.44E-03 1.56E+01
8.90E-06 3.83E+07 2.66E-08 1.36E+04 7.07E-03 1.56E+01
9.72E-06 3.26E+07 2.39E-08 1.37E+04 6.94E-03 1.57E+01
1.87E-05 3.46E+07 3.79E-08 4.68E+04 5.90E-04 2.44E+01
1.98E-05 3.20E+07 3.38E-08 4.90E+04 5.29E-04 2.57E+01
1.82E-05 3.62E+07 3.63E-08 4.84E+04 5.94E-04 2.43E+01
1.67E-05 3.56E+07 3.68E-08 4.63E+04 5.35E-04 2.28E+01
1.75E-05 3.25E+07 3.88E-08 4.63E+04 5.92E-04 2.41E+01
1.88E-05 3.03E+07 3.75E-08 5.13E+04 5.26E-04 2.38E+01
1.83E-05 3.58E+07 3.71E-08 4.65E+04 5.36E-04 2.44E+01
1.80E-05 3.68E+07 4.06E-08 4.78E+04 5.29E-04 2.38E+01
1.58E-05 3.41E+07 3.70E-08 4.45E+04 6.12E-04 2.42E+01
1.51E-05 3.58E+07 3.76E-08 4.29E+04 5.98E-04 2.43E+01
1.81E-05 2.96E+07 6.99E-08 6.31E+04 1.20E-03 3.27E+01
1.70E-05 2.64E+07 6.87E-08 5.99E+04 1.20E-03 3.46E+01
1.95E-05 2.87E+07 6.99E-08 6.55E+04 1.02E-03 3.27E+01
1.70E-05 2.98E+07 6.56E-08 6.97E+04 1.16E-03 3.11E+01
1.79E-05 2.63E+07 7.10E-08 6.06E+04 1.09E-03 3.46E+01
1.90E-05 2.46E+07 7.13E-08 6.28E+04 1.10E-03 3.48E+01
1.85E-05 2.76E+07 7.03E-08 5.99E+04 1.14E-03 3.48E+01
1.81E-05 2.65E+07 6.89E-08 6.22E+04 1.04E-03 3.34E+01
1.94E-05 2.70E+07 6.50E-08 6.28E+04 1.20E-03 3.33E+01
1.70E-05 2.73E+07 6.71E-08 6.97E+04 1.01E-03 3.05E+01
1.82E-05 1.59E+07 1.09E-07 6.38E+04 1.53E-03 3.82E+01
1.85E-05 1.44E+07 1.03E-07 6.38E+04 1.58E-03 3.74E+01
1.90E-05 1.68E+07 1.13E-07 6.76E+04 1.63E-03 3.62E+01
1.88E-05 1.60E+07 1.00E-07 6.94E+04 1.64E-03 3.54E+01
1.83E-05 1.39E+07 1.08E-07 7.21E+04 1.78E-03 3.45E+01
1.85E-05 1.50E+07 1.06E-07 6.67E+04 1.57E-03 3.63E+01
1.89E-05 1.57E+07 1.18E-07 6.97E+04 1.80E-03 3.56E+01
2.00E-05 1.46E+07 1.14E-07 6.35E+04 1.69E-03 3.75E+01
1.93E-05 1.39E+07 9.63E-08 6.38E+04 1.59E-03 3.72E+01
1.97E-05 1.59E+07 1.07E-07 6.99E+04 1.61E-03 3.62E+01
127
Table D.7: Powerex IGBT #2 raw data table (continued)
9Ω
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
Shot #
106.4
1
106.4
2
106.4
3
106.4
4
106.4
5
106.4
6
106.4
7
106.4
8
106.4
9
106.4
10
212
1
212
2
212
3
212
4
212
5
212
6
212
7
212
8
212
9
212
10
320
1
320
2
320
3
320
4
320
5
320
6
320
7
320
8
320
9
320
10
372
1
372
2
372
3
372
4
372
5
372
6
372
7
372
8
372
9
372
10
AGD
MUN
CGS
MUP
COXD
KP
9.38E-06 1.61E+06 1.89E-08 3.59E+02 8.40E-04 9.76E+00
1.44E-05 1.57E+06 1.99E-08 3.59E+02 7.76E-04 9.76E+00
1.13E-05 1.53E+06 2.14E-08 3.14E+02 8.43E-04 9.33E+00
1.14E-05 1.73E+06 1.92E-08 3.34E+02 7.72E-04 9.74E+00
8.91E-06 1.52E+06 1.87E-08 2.93E+02 7.64E-04 9.51E+00
8.92E-06 1.68E+06 2.00E-08 3.34E+02 7.72E-04 9.50E+00
1.19E-05 1.53E+06 2.03E-08 3.56E+02 7.23E-04 9.67E+00
1.32E-05 1.69E+06 2.13E-08 3.03E+02 7.75E-04 9.84E+00
8.91E-06 1.49E+06 2.00E-08 3.59E+02 7.67E-04 9.21E+00
9.80E-06 1.60E+06 1.92E-08 3.45E+02 8.44E-04 9.76E+00
1.32E-05 9.67E+06 2.38E-08 1.15E+03 3.07E-04 1.97E+01
9.51E-06 9.11E+06 2.39E-08 1.23E+03 3.23E-04 1.88E+01
1.01E-05 9.47E+06 2.36E-08 1.20E+03 2.98E-04 1.95E+01
1.52E-05 9.66E+06 2.24E-08 1.27E+03 3.02E-04 1.94E+01
9.35E-06 9.18E+06 2.18E-08 1.21E+03 2.80E-04 1.90E+01
9.33E-06 1.03E+07 2.55E-08 1.30E+03 3.03E-04 1.92E+01
1.07E-05 1.01E+07 2.39E-08 1.26E+03 2.98E-04 1.92E+01
8.42E-06 9.58E+06 2.42E-08 1.30E+03 3.00E-04 1.91E+01
1.24E-05 9.33E+06 2.37E-08 1.24E+03 3.07E-04 1.88E+01
8.94E-06 9.66E+06 2.62E-08 1.33E+03 2.83E-04 1.99E+01
7.41E-06 4.51E+06 2.80E-08 5.30E+04 1.82E-04 1.52E+01
6.50E-06 5.24E+06 2.85E-08 5.19E+04 1.67E-04 1.44E+01
9.79E-06 4.90E+06 2.99E-08 4.87E+04 1.77E-04 1.62E+01
6.98E-06 4.96E+06 2.77E-08 5.17E+04 1.56E-04 1.57E+01
1.35E-05 4.69E+06 2.80E-08 5.59E+04 1.64E-04 1.55E+01
4.56E-06 5.13E+06 2.79E-08 5.29E+04 1.56E-04 1.45E+01
5.39E-06 4.88E+06 3.02E-08 4.78E+04 1.61E-04 1.59E+01
9.71E-06 4.81E+06 2.82E-08 5.26E+04 1.73E-04 1.54E+01
4.86E-06 4.66E+06 2.91E-08 4.89E+04 1.77E-04 1.58E+01
5.37E-06 5.14E+06 3.16E-08 5.11E+04 1.71E-04 1.46E+01
1.95E-05 4.70E+07 9.83E-09 7.80E+04 9.35E-04 1.46E+01
1.85E-05 4.79E+07 1.04E-08 7.56E+04 1.04E-03 1.49E+01
1.91E-05 4.40E+07 1.02E-08 7.99E+04 9.67E-04 1.40E+01
1.88E-05 5.07E+07 1.08E-08 6.76E+04 9.56E-04 1.61E+01
1.94E-05 4.69E+07 9.94E-09 7.93E+04 1.05E-03 1.40E+01
1.93E-05 4.87E+07 9.58E-09 7.20E+04 9.98E-04 1.51E+01
1.94E-05 4.92E+07 9.75E-09 7.79E+04 9.82E-04 1.45E+01
1.80E-05 4.90E+07 1.02E-08 7.97E+04 9.63E-04 1.41E+01
1.88E-05 4.80E+07 1.01E-08 7.08E+04 1.04E-03 1.53E+01
1.92E-05 4.67E+07 9.64E-09 7.40E+04 9.41E-04 1.45E+01
128
Table D.8: Powerex IGBT #2 raw data table (continued)
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
15 Ω
3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
67.2
67.2
67.2
67.2
67.2
67.2
67.2
67.2
67.2
67.2
132
132
132
132
132
132
132
132
132
132
200
200
200
200
200
200
200
200
200
200
234
234
234
234
234
234
234
234
234
234
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
AGD
MUN
CGS
MUP
COXD
KP
1.20E-05 1.28E+06 1.30E-08 1.25E+04 1.26E-03 4.26E+00
1.43E-05 1.37E+06 1.33E-08 1.21E+04 1.33E-03 4.29E+00
1.32E-05 1.38E+06 1.32E-08 1.09E+04 1.31E-03 4.52E+00
1.60E-05 1.41E+06 1.28E-08 1.21E+04 1.36E-03 4.27E+00
1.91E-05 1.47E+06 1.19E-08 1.22E+04 1.49E-03 4.29E+00
1.12E-05 1.48E+06 1.36E-08 1.23E+04 1.26E-03 4.28E+00
1.13E-05 1.44E+06 1.31E-08 1.22E+04 1.38E-03 4.26E+00
1.69E-05 1.55E+06 1.26E-08 1.30E+04 1.39E-03 4.18E+00
1.60E-05 1.43E+06 1.25E-08 1.22E+04 1.37E-03 4.36E+00
1.11E-05 1.37E+06 1.37E-08 1.21E+04 1.47E-03 4.28E+00
7.91E-06 8.91E+05 1.68E-08 2.34E+04 4.54E-04 8.15E+00
8.30E-06 9.85E+05 1.75E-08 2.54E+04 4.64E-04 7.14E+00
4.56E-06 8.96E+05 1.74E-08 2.66E+04 4.11E-04 7.08E+00
8.88E-06 1.00E+06 1.75E-08 2.55E+04 4.21E-04 7.30E+00
4.56E-06 9.29E+05 1.61E-08 2.38E+04 4.07E-04 6.99E+00
5.36E-06 1.01E+06 1.70E-08 2.30E+04 4.51E-04 7.10E+00
6.02E-06 9.20E+05 1.87E-08 2.59E+04 3.80E-04 7.41E+00
4.56E-06 1.00E+06 1.82E-08 2.81E+04 3.80E-04 6.67E+00
7.83E-06 9.67E+05 1.70E-08 2.48E+04 4.39E-04 7.47E+00
4.56E-06 8.86E+05 1.62E-08 2.36E+04 3.84E-04 7.38E+00
8.67E-06 1.02E+06 1.35E-08 4.56E+04 5.90E-04 8.69E+00
8.71E-06 1.04E+06 1.38E-08 5.01E+04 6.87E-04 8.12E+00
7.60E-06 1.08E+06 1.32E-08 4.25E+04 7.12E-04 9.30E+00
8.24E-06 9.49E+05 1.44E-08 4.17E+04 6.25E-04 9.92E+00
7.23E-06 1.02E+06 1.25E-08 4.92E+04 5.90E-04 8.24E+00
7.23E-06 9.70E+05 1.37E-08 4.20E+04 6.94E-04 9.64E+00
1.43E-05 1.02E+06 1.46E-08 4.22E+04 6.77E-04 9.43E+00
7.23E-06 1.03E+06 1.48E-08 4.86E+04 6.58E-04 8.83E+00
7.35E-06 1.04E+06 1.31E-08 4.61E+04 6.27E-04 8.69E+00
7.23E-06 9.83E+05 1.31E-08 5.00E+04 5.90E-04 8.12E+00
6.07E-06 8.35E+06 1.84E-08 6.17E+04 6.16E-04 9.50E+00
3.47E-06 9.19E+06 2.17E-08 5.93E+04 5.74E-04 9.45E+00
3.82E-06 8.46E+06 2.12E-08 6.54E+04 5.97E-04 8.17E+00
4.91E-06 8.09E+06 1.82E-08 6.46E+04 6.17E-04 8.24E+00
3.47E-06 9.72E+06 2.22E-08 6.00E+04 5.61E-04 9.26E+00
3.77E-06 8.60E+06 2.22E-08 6.30E+04 6.20E-04 8.73E+00
7.21E-06 8.92E+06 2.22E-08 5.45E+04 6.20E-04 9.02E+00
5.94E-06 8.93E+06 1.98E-08 6.50E+04 5.75E-04 8.31E+00
5.34E-06 8.95E+06 2.04E-08 6.34E+04 5.54E-04 8.65E+00
4.81E-06 9.60E+06 1.86E-08 5.53E+04 5.11E-04 9.33E+00
129
Table D.9: Powerex IGBT #3 raw data table
3Ω
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
304
304
304
304
304
304
304
304
304
304
592
592
592
592
592
592
592
592
592
592
720
720
720
720
720
720
720
720
720
720
736
736
736
736
736
736
736
736
736
736
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
AGD
MUN
CGS
MUP
COXD
KP
1.34E-05 5.02E+07 3.06E-08 4.26E+04 2.02E-02 1.66E+01
8.75E-06 4.97E+07 2.70E-08 3.84E+04 1.93E-02 1.67E+01
1.23E-05 4.87E+07 2.92E-08 4.19E+04 1.96E-02 1.62E+01
1.03E-05 4.75E+07 2.99E-08 3.90E+04 1.94E-02 1.62E+01
1.32E-05 5.04E+07 2.94E-08 3.94E+04 1.83E-02 1.66E+01
9.10E-06 4.89E+07 3.15E-08 4.18E+04 1.91E-02 1.62E+01
1.21E-05 5.03E+07 3.00E-08 3.97E+04 1.89E-02 1.65E+01
9.44E-06 5.42E+07 3.12E-08 4.16E+04 1.96E-02 1.57E+01
1.53E-05 4.91E+07 3.02E-08 4.09E+04 1.90E-02 1.59E+01
1.48E-05 5.18E+07 3.10E-08 3.89E+04 1.95E-02 1.65E+01
1.86E-05 5.24E+07 9.20E-08 8.39E+04 1.55E-02 2.68E+01
1.82E-05 5.40E+07 8.58E-08 8.41E+04 1.57E-02 2.74E+01
1.84E-05 6.29E+07 9.29E-08 8.51E+04 1.55E-02 2.67E+01
1.99E-05 6.30E+07 8.02E-08 9.72E+04 1.64E-02 2.42E+01
1.82E-05 5.57E+07 9.02E-08 9.01E+04 1.64E-02 2.56E+01
1.94E-05 5.44E+07 8.50E-08 8.76E+04 1.58E-02 2.51E+01
1.82E-05 6.22E+07 7.85E-08 9.33E+04 1.60E-02 2.43E+01
1.85E-05 5.71E+07 9.06E-08 9.31E+04 1.63E-02 2.52E+01
1.86E-05 5.65E+07 9.20E-08 8.73E+04 1.46E-02 2.63E+01
1.89E-05 5.68E+07 8.66E-08 9.53E+04 1.55E-02 2.43E+01
1.98E-05 5.24E+07 1.15E-07 7.10E+04 3.58E-02 3.63E+01
1.88E-05 5.18E+07 1.20E-07 7.69E+04 3.82E-02 3.46E+01
1.90E-05 5.21E+07 1.13E-07 7.53E+04 3.92E-02 3.59E+01
1.92E-05 4.89E+07 1.20E-07 7.77E+04 3.84E-02 3.48E+01
2.00E-05 5.23E+07 1.08E-07 7.61E+04 3.91E-02 3.49E+01
1.90E-05 5.18E+07 1.10E-07 7.32E+04 3.77E-02 3.61E+01
1.89E-05 5.12E+07 1.15E-07 7.41E+04 3.69E-02 3.54E+01
1.88E-05 4.95E+07 1.19E-07 7.26E+04 3.69E-02 3.59E+01
1.95E-05 5.48E+07 1.16E-07 7.89E+04 4.31E-02 3.52E+01
1.91E-05 5.07E+07 1.12E-07 7.31E+04 3.77E-02 3.53E+01
1.87E-05 5.89E+08 1.26E-07 7.19E+04 1.65E-01 3.96E+01
1.80E-05 6.16E+08 1.21E-07 7.77E+04 1.81E-01 3.78E+01
1.91E-05 5.60E+08 1.17E-07 7.15E+04 1.68E-01 3.79E+01
1.82E-05 5.68E+08 1.16E-07 7.42E+04 1.70E-01 3.73E+01
1.80E-05 6.04E+08 1.15E-07 7.06E+04 1.66E-01 3.78E+01
1.84E-05 5.54E+08 1.24E-07 6.80E+04 1.79E-01 3.99E+01
1.98E-05 5.23E+08 1.31E-07 7.89E+04 1.80E-01 3.44E+01
1.91E-05 6.01E+08 1.17E-07 7.17E+04 1.53E-01 3.73E+01
1.92E-05 5.80E+08 1.22E-07 7.79E+04 1.55E-01 3.66E+01
2.00E-05 6.32E+08 1.33E-07 6.97E+04 1.81E-01 3.83E+01
130
Table D.10: Powerex IGBT #3 raw data table (continued)
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
4.5 Ω 3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
210
210
210
210
210
210
210
210
210
210
424
424
424
424
424
424
424
424
424
424
616
616
616
616
616
616
616
616
616
616
696
696
696
696
696
696
696
696
696
696
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
AGD
MUN
CGS
MUP
COXD
KP
1.57E-05 3.26E+07 2.29E-08 1.31E+04 7.44E-03 1.57E+01
1.40E-05 3.19E+07 2.01E-08 1.45E+04 8.69E-03 1.43E+01
9.10E-06 3.20E+07 2.30E-08 1.38E+04 8.66E-03 1.54E+01
1.59E-05 3.29E+07 2.08E-08 1.36E+04 7.87E-03 1.51E+01
1.29E-05 3.27E+07 1.99E-08 1.50E+04 7.85E-03 1.54E+01
1.03E-05 3.24E+07 2.03E-08 1.32E+04 8.37E-03 1.54E+01
1.16E-05 3.29E+07 2.13E-08 1.37E+04 8.52E-03 1.47E+01
1.01E-05 3.00E+07 2.22E-08 1.43E+04 8.72E-03 1.48E+01
1.89E-05 3.03E+07 1.89E-08 1.44E+04 7.24E-03 1.50E+01
1.55E-05 3.49E+07 1.96E-08 1.45E+04 7.55E-03 1.52E+01
1.85E-05 3.62E+07 3.81E-08 4.51E+04 5.22E-04 2.39E+01
1.82E-05 3.72E+07 4.20E-08 4.86E+04 5.00E-04 2.31E+01
1.90E-05 3.42E+07 3.85E-08 5.03E+04 5.14E-04 2.31E+01
1.89E-05 3.47E+07 4.23E-08 5.14E+04 5.14E-04 2.22E+01
1.83E-05 3.66E+07 4.19E-08 4.53E+04 5.56E-04 2.49E+01
1.81E-05 3.52E+07 4.02E-08 4.81E+04 5.37E-04 2.37E+01
1.79E-05 3.68E+07 3.81E-08 4.73E+04 4.95E-04 2.38E+01
1.83E-05 3.64E+07 4.15E-08 4.93E+04 5.11E-04 2.43E+01
1.81E-05 3.84E+07 4.01E-08 4.93E+04 5.48E-04 2.37E+01
1.79E-05 3.72E+07 4.30E-08 4.47E+04 5.22E-04 2.44E+01
2.00E-05 2.90E+07 8.12E-08 6.62E+04 9.91E-04 3.26E+01
1.85E-05 2.69E+07 7.73E-08 6.58E+04 9.81E-04 3.39E+01
1.83E-05 2.78E+07 7.76E-08 6.46E+04 1.01E-03 3.33E+01
1.91E-05 2.90E+07 7.58E-08 6.73E+04 1.06E-03 3.21E+01
1.77E-05 2.64E+07 7.52E-08 6.16E+04 9.99E-04 3.47E+01
1.91E-05 2.63E+07 7.92E-08 7.01E+04 1.03E-03 3.25E+01
1.86E-05 2.72E+07 7.75E-08 6.66E+04 1.03E-03 3.32E+01
1.93E-05 2.85E+07 8.12E-08 6.50E+04 9.54E-04 3.35E+01
1.88E-05 2.57E+07 7.26E-08 6.63E+04 9.79E-04 3.16E+01
1.74E-05 2.45E+07 7.59E-08 6.35E+04 9.46E-04 3.37E+01
1.88E-05 1.86E+07 1.17E-07 5.37E+04 1.61E-03 4.52E+01
1.82E-05 1.68E+07 1.28E-07 5.63E+04 1.95E-03 4.11E+01
2.00E-05 1.62E+07 1.26E-07 5.70E+04 1.79E-03 4.27E+01
1.73E-05 1.63E+07 1.18E-07 5.59E+04 1.85E-03 4.08E+01
1.96E-05 1.90E+07 1.17E-07 5.25E+04 1.72E-03 4.40E+01
1.89E-05 1.91E+07 1.16E-07 6.41E+04 1.82E-03 3.87E+01
1.91E-05 1.74E+07 1.34E-07 5.72E+04 1.77E-03 4.10E+01
2.00E-05 1.63E+07 1.28E-07 5.96E+04 1.72E-03 4.08E+01
1.97E-05 1.70E+07 1.38E-07 5.46E+04 1.95E-03 4.05E+01
1.99E-05 1.58E+07 1.21E-07 5.45E+04 1.64E-03 4.19E+01
131
Table D.11: Powerex IGBT #3 raw data table (continued)
9Ω
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
Shot #
106.4
1
106.4
2
106.4
3
106.4
4
106.4
5
106.4
6
106.4
7
106.4
8
106.4
9
106.4
10
212
1
212
2
212
3
212
4
212
5
212
6
212
7
212
8
212
9
212
10
320
1
320
2
320
3
320
4
320
5
320
6
320
7
320
8
320
9
320
10
372
1
372
2
372
3
372
4
372
5
372
6
372
7
372
8
372
9
372
10
AGD
MUN
CGS
MUP
COXD
KP
9.55E-06 1.25E+06 1.93E-08 3.82E+02 7.98E-04 8.92E+00
1.47E-05 1.25E+06 1.81E-08 3.88E+02 9.61E-04 9.39E+00
1.05E-05 1.42E+06 1.94E-08 4.16E+02 8.86E-04 9.75E+00
1.39E-05 1.36E+06 1.83E-08 3.98E+02 8.88E-04 9.80E+00
1.20E-05 1.34E+06 2.05E-08 4.10E+02 8.82E-04 9.57E+00
1.97E-05 1.33E+06 1.81E-08 3.93E+02 9.47E-04 9.71E+00
1.33E-05 1.34E+06 2.07E-08 3.99E+02 8.49E-04 9.43E+00
1.45E-05 1.42E+06 1.97E-08 3.90E+02 8.77E-04 9.39E+00
1.33E-05 1.41E+06 1.94E-08 3.83E+02 8.81E-04 9.63E+00
1.30E-05 1.39E+06 1.96E-08 4.08E+02 8.84E-04 9.77E+00
1.15E-05 1.09E+07 2.35E-08 1.35E+03 2.50E-04 1.84E+01
1.16E-05 1.00E+07 2.32E-08 1.39E+03 2.82E-04 1.86E+01
1.14E-05 9.52E+06 2.45E-08 1.29E+03 2.56E-04 1.94E+01
1.35E-05 1.04E+07 2.60E-08 1.45E+03 2.73E-04 1.93E+01
8.96E-06 1.06E+07 2.55E-08 1.31E+03 2.90E-04 1.87E+01
1.35E-05 9.71E+06 2.40E-08 1.39E+03 2.73E-04 1.86E+01
1.86E-05 9.09E+06 2.14E-08 1.53E+03 2.47E-04 1.83E+01
1.82E-05 1.04E+07 2.30E-08 1.42E+03 2.70E-04 1.90E+01
1.31E-05 9.81E+06 2.34E-08 1.35E+03 2.74E-04 1.89E+01
1.45E-05 1.07E+07 2.26E-08 1.37E+03 2.68E-04 1.91E+01
1.50E-05 4.46E+07 2.60E-08 4.75E+04 1.50E-04 1.70E+01
9.57E-06 5.18E+07 2.94E-08 4.82E+04 1.56E-04 1.64E+01
9.19E-06 4.67E+07 2.56E-08 4.68E+04 1.57E-04 1.69E+01
8.49E-06 4.75E+07 2.72E-08 4.99E+04 1.69E-04 1.64E+01
8.66E-06 4.80E+07 2.59E-08 5.02E+04 1.72E-04 1.61E+01
1.69E-05 4.85E+07 2.95E-08 5.04E+04 1.62E-04 1.61E+01
7.45E-06 5.12E+07 2.65E-08 4.63E+04 1.69E-04 1.62E+01
8.41E-06 4.73E+07 2.59E-08 5.01E+04 1.77E-04 1.62E+01
9.55E-06 5.18E+07 2.47E-08 4.75E+04 1.59E-04 1.67E+01
7.59E-06 4.75E+07 2.70E-08 5.11E+04 1.58E-04 1.60E+01
1.94E-05 5.31E+07 9.35E-09 8.05E+04 1.05E-03 1.41E+01
1.83E-05 4.77E+07 1.03E-08 7.96E+04 9.83E-04 1.40E+01
1.79E-05 4.99E+07 1.01E-08 7.97E+04 1.05E-03 1.38E+01
1.93E-05 4.76E+07 9.81E-09 8.45E+04 9.84E-04 1.34E+01
1.78E-05 4.67E+07 1.01E-08 7.91E+04 1.02E-03 1.40E+01
1.83E-05 4.55E+07 9.70E-09 8.43E+04 1.09E-03 1.31E+01
1.83E-05 4.67E+07 1.03E-08 8.08E+04 1.03E-03 1.39E+01
1.83E-05 4.57E+07 1.04E-08 8.28E+04 1.01E-03 1.34E+01
1.72E-05 4.82E+07 9.85E-09 8.25E+04 9.91E-04 1.35E+01
1.80E-05 4.90E+07 1.09E-08 7.97E+04 1.03E-03 1.39E+01
132
Table D.12: Powerex IGBT #3 raw data table (continued)
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
15 Ω 3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
67.2
67.2
67.2
67.2
67.2
67.2
67.2
67.2
67.2
67.2
132
132
132
132
132
132
132
132
132
132
200
200
200
200
200
200
200
200
200
200
234
234
234
234
234
234
234
234
234
234
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
AGD
MUN
CGS
MUP
COXD
KP
1.47E-05 1.42E+06 1.25E-08 1.17E+04 1.36E-03 4.45E+00
1.38E-05 1.45E+06 1.24E-08 1.21E+04 1.35E-03 4.03E+00
1.86E-05 1.64E+06 1.37E-08 1.11E+04 1.62E-03 4.08E+00
1.90E-05 1.46E+06 1.32E-08 1.10E+04 1.40E-03 3.92E+00
1.71E-05 1.50E+06 1.46E-08 1.09E+04 1.55E-03 4.25E+00
1.64E-05 1.38E+06 1.46E-08 1.15E+04 1.62E-03 4.40E+00
1.27E-05 1.48E+06 1.22E-08 1.05E+04 1.63E-03 4.51E+00
1.24E-05 1.55E+06 1.33E-08 1.15E+04 1.55E-03 4.31E+00
2.00E-05 1.64E+06 1.42E-08 1.24E+04 1.48E-03 4.00E+00
1.56E-05 1.42E+06 1.30E-08 1.29E+04 1.44E-03 4.29E+00
1.27E-05 8.86E+05 1.57E-08 2.22E+04 4.16E-04 7.84E+00
8.86E-06 8.57E+05 1.62E-08 1.96E+04 4.14E-04 7.75E+00
4.23E-06 8.03E+05 1.33E-08 1.91E+04 3.69E-04 7.95E+00
9.99E-06 8.77E+05 1.44E-08 2.08E+04 4.00E-04 7.50E+00
8.86E-06 8.83E+05 1.45E-08 2.13E+04 4.09E-04 7.29E+00
6.31E-06 8.00E+05 1.62E-08 1.98E+04 3.92E-04 8.20E+00
6.56E-06 9.09E+05 1.54E-08 2.33E+04 4.35E-04 7.11E+00
1.38E-05 8.73E+05 1.40E-08 2.16E+04 4.09E-04 7.48E+00
9.22E-06 8.99E+05 1.46E-08 2.11E+04 4.10E-04 7.54E+00
4.31E-06 9.46E+05 1.53E-08 2.14E+04 4.24E-04 8.13E+00
8.87E-06 1.10E+06 1.57E-08 4.73E+04 7.05E-04 8.15E+00
8.47E-06 9.41E+05 1.61E-08 4.79E+04 6.21E-04 8.34E+00
7.51E-06 1.03E+06 1.61E-08 5.00E+04 6.12E-04 8.50E+00
8.05E-06 9.39E+05 1.44E-08 4.96E+04 7.00E-04 8.66E+00
7.36E-06 9.64E+05 1.53E-08 4.37E+04 7.24E-04 9.47E+00
8.75E-06 1.06E+06 1.51E-08 4.85E+04 6.58E-04 9.05E+00
6.67E-06 1.06E+06 1.61E-08 4.38E+04 6.43E-04 8.65E+00
6.51E-06 1.04E+06 1.33E-08 5.35E+04 7.24E-04 7.97E+00
8.06E-06 1.05E+06 1.54E-08 5.03E+04 6.67E-04 7.95E+00
9.44E-06 1.03E+06 1.57E-08 4.71E+04 7.02E-04 8.61E+00
9.75E-06 7.78E+06 2.10E-08 5.56E+04 6.02E-04 9.68E+00
5.68E-06 8.03E+06 1.87E-08 6.17E+04 6.23E-04 8.86E+00
1.22E-05 8.16E+06 2.01E-08 5.66E+04 5.43E-04 9.02E+00
7.45E-06 8.26E+06 2.17E-08 5.25E+04 6.38E-04 9.44E+00
5.68E-06 7.68E+06 1.96E-08 5.37E+04 5.82E-04 9.32E+00
5.68E-06 8.06E+06 2.06E-08 6.33E+04 6.18E-04 9.06E+00
1.36E-05 8.88E+06 1.98E-08 5.93E+04 5.82E-04 9.08E+00
8.61E-06 7.95E+06 2.22E-08 5.93E+04 5.94E-04 9.48E+00
5.68E-06 8.65E+06 2.13E-08 6.41E+04 5.30E-04 8.36E+00
7.60E-06 7.77E+06 2.00E-08 6.12E+04 6.32E-04 9.14E+00
133
Table D.13: IXYS IGBT raw data table
3Ω
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
288
288
288
288
288
288
288
288
288
288
352
352
352
352
352
352
352
352
352
352
364
364
364
364
364
364
364
364
364
364
368
368
368
368
368
368
368
368
368
368
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
AGD
MUN
CGS
MUP
COXD
KP
1.95E-05 9.30E+08 9.17E-08 2.06E+04 1.91E+00 2.24E+01
1.97E-05 1.06E+09 8.85E-08 2.08E+04 1.67E+00 2.22E+01
2.00E-05 1.13E+09 1.01E-07 1.74E+04 1.57E+00 2.33E+01
1.81E-05 1.12E+09 8.40E-08 1.74E+04 1.57E+00 2.33E+01
1.84E-05 9.66E+08 9.10E-08 2.08E+04 1.67E+00 2.23E+01
1.92E-05 1.04E+09 8.99E-08 1.82E+04 1.68E+00 2.32E+01
1.88E-05 1.02E+09 9.33E-08 2.01E+04 1.75E+00 2.28E+01
1.82E-05 9.44E+08 8.32E-08 2.08E+04 1.83E+00 2.20E+01
1.93E-05 1.00E+09 9.33E-08 1.97E+04 1.69E+00 2.28E+01
1.96E-05 1.05E+09 8.62E-08 1.95E+04 1.63E+00 2.23E+01
1.47E-05 2.65E+10 9.78E-08 2.10E+04 2.68E+00 2.62E+01
1.46E-05 2.68E+10 9.06E-08 2.11E+04 2.55E+00 2.62E+01
1.46E-05 2.68E+10 9.13E-08 2.08E+04 2.59E+00 2.64E+01
1.42E-05 2.73E+10 9.17E-08 2.18E+04 2.46E+00 2.62E+01
1.37E-05 2.58E+10 9.07E-08 2.16E+04 2.84E+00 2.64E+01
1.24E-05 2.83E+10 9.55E-08 2.10E+04 2.46E+00 2.69E+01
1.47E-05 2.87E+10 9.63E-08 2.17E+04 2.69E+00 2.61E+01
1.52E-05 2.82E+10 9.78E-08 2.32E+04 2.58E+00 2.55E+01
1.42E-05 2.73E+10 9.32E-08 2.21E+04 2.56E+00 2.60E+01
1.33E-05 2.95E+10 8.78E-08 2.23E+04 2.59E+00 2.55E+01
1.51E-05 3.63E+11 1.06E-07 2.64E+04 1.02E+01 2.61E+01
1.59E-05 3.76E+11 1.06E-07 2.65E+04 9.93E+00 2.57E+01
1.59E-05 3.83E+11 1.05E-07 2.57E+04 1.12E+01 2.68E+01
1.36E-05 3.52E+11 1.02E-07 2.59E+04 1.04E+01 2.67E+01
1.40E-05 3.81E+11 1.08E-07 2.63E+04 1.02E+01 2.67E+01
1.54E-05 3.53E+11 1.08E-07 2.64E+04 1.08E+01 2.66E+01
1.39E-05 3.97E+11 9.84E-08 2.49E+04 1.06E+01 2.65E+01
1.66E-05 3.46E+11 1.09E-07 2.51E+04 1.01E+01 2.68E+01
1.47E-05 3.41E+11 1.03E-07 2.82E+04 1.08E+01 2.55E+01
1.55E-05 3.43E+11 1.01E-07 2.69E+04 1.05E+01 2.55E+01
2.00E-05 1.56E+12 1.38E-07 2.12E+04 5.35E+01 2.83E+01
1.92E-05 1.53E+12 1.33E-07 2.36E+04 5.23E+01 2.74E+01
1.84E-05 1.72E+12 1.27E-07 2.27E+04 5.97E+01 2.77E+01
1.93E-05 1.61E+12 1.26E-07 2.47E+04 5.09E+01 2.67E+01
1.89E-05 1.60E+12 1.20E-07 2.39E+04 5.17E+01 2.69E+01
1.98E-05 1.51E+12 1.28E-07 2.42E+04 5.78E+01 2.67E+01
1.82E-05 1.74E+12 1.21E-07 2.17E+04 5.76E+01 2.74E+01
1.84E-05 1.77E+12 1.27E-07 2.34E+04 5.04E+01 2.73E+01
1.83E-05 1.61E+12 1.27E-07 2.41E+04 5.81E+01 2.68E+01
1.90E-05 1.79E+12 1.27E-07 2.31E+04 5.63E+01 2.75E+01
134
Table D.14: IXYS IGBT raw data table (continued)
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
4.5 Ω 3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
202
202
202
202
202
202
202
202
202
202
332
332
332
332
332
332
332
332
332
332
356
356
356
356
356
356
356
356
356
356
364
364
364
364
364
364
364
364
364
364
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
AGD
MUN
CGS
MUP
COXD
KP
1.67E-05 2.39E+08 3.61E-08 1.95E+04 4.32E-01 1.50E+01
1.29E-05 2.37E+08 3.78E-08 2.15E+04 4.59E-01 1.47E+01
9.84E-06 2.32E+08 3.98E-08 2.05E+04 4.08E-01 1.49E+01
1.12E-05 2.35E+08 3.91E-08 2.03E+04 4.20E-01 1.48E+01
1.49E-05 2.51E+08 3.63E-08 1.81E+04 4.62E-01 1.54E+01
1.38E-05 2.17E+08 3.64E-08 1.90E+04 3.92E-01 1.44E+01
1.21E-05 2.40E+08 3.49E-08 2.14E+04 4.44E-01 1.43E+01
9.21E-06 2.28E+08 3.37E-08 1.95E+04 4.40E-01 1.44E+01
1.41E-05 2.24E+08 3.46E-08 2.07E+04 4.33E-01 1.46E+01
1.40E-05 2.42E+08 3.72E-08 2.06E+04 4.48E-01 1.48E+01
1.99E-05 9.80E+08 9.97E-08 9.39E+03 1.58E+00 2.92E+01
1.90E-05 9.89E+08 9.92E-08 9.08E+03 1.85E+00 3.00E+01
1.94E-05 9.87E+08 1.04E-07 8.63E+03 1.69E+00 3.05E+01
2.00E-05 9.00E+08 1.08E-07 8.38E+03 1.60E+00 3.00E+01
1.96E-05 1.03E+09 9.67E-08 8.42E+03 1.74E+00 3.02E+01
1.93E-05 9.94E+08 9.38E-08 9.30E+03 1.80E+00 2.93E+01
1.95E-05 1.09E+09 1.03E-07 9.36E+03 1.90E+00 2.94E+01
1.92E-05 9.83E+08 9.96E-08 8.62E+03 1.65E+00 3.03E+01
1.80E-05 9.00E+08 9.89E-08 9.50E+03 1.76E+00 3.03E+01
2.00E-05 9.56E+08 1.04E-07 9.50E+03 1.90E+00 2.90E+01
1.95E-05 3.29E+09 1.22E-07 1.11E+04 3.55E+00 3.13E+01
1.84E-05 3.12E+09 1.11E-07 1.09E+04 3.64E+00 3.08E+01
1.91E-05 3.50E+09 1.15E-07 1.07E+04 3.69E+00 3.10E+01
2.00E-05 2.94E+09 1.25E-07 1.03E+04 3.94E+00 3.12E+01
1.92E-05 3.12E+09 1.24E-07 1.12E+04 3.82E+00 3.09E+01
1.83E-05 3.00E+09 1.14E-07 1.18E+04 3.73E+00 3.03E+01
1.98E-05 2.95E+09 1.18E-07 1.18E+04 3.96E+00 3.06E+01
1.84E-05 3.36E+09 1.17E-07 1.11E+04 3.90E+00 3.13E+01
1.89E-05 2.88E+09 1.19E-07 1.14E+04 3.43E+00 3.10E+01
1.90E-05 2.94E+09 1.24E-07 1.10E+04 3.58E+00 3.23E+01
1.84E-05 2.34E+10 1.15E-07 7.44E+03 8.00E+00 3.26E+01
1.98E-05 2.22E+10 1.26E-07 7.32E+03 8.11E+00 3.32E+01
2.00E-05 1.94E+10 1.13E-07 7.28E+03 8.65E+00 3.26E+01
1.91E-05 2.02E+10 1.20E-07 7.30E+03 8.49E+00 3.34E+01
1.88E-05 2.27E+10 1.14E-07 7.03E+03 9.02E+00 3.30E+01
2.00E-05 2.26E+10 1.19E-07 6.81E+03 8.80E+00 3.29E+01
1.95E-05 2.23E+10 1.22E-07 7.43E+03 9.45E+00 3.38E+01
1.93E-05 2.27E+10 1.22E-07 7.53E+03 8.65E+00 3.25E+01
1.87E-05 2.12E+10 1.20E-07 7.74E+03 9.18E+00 3.35E+01
1.90E-05 2.23E+10 1.20E-07 7.32E+03 8.84E+00 3.34E+01
135
Table D.15: IXYS IGBT raw data table (continued)
9Ω
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
104
104
104
104
104
104
104
104
104
104
210
210
210
210
210
210
210
210
210
210
304
304
304
304
304
304
304
304
304
304
328
328
328
328
328
328
328
328
328
328
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
AGD
MUN
CGS
MUP
COXD
KP
4.83E-06 1.15E+07 2.21E-08 1.12E+04 3.20E-02 8.07E+00
1.31E-05 1.17E+07 2.06E-08 9.62E+03 2.92E-02 8.43E+00
9.20E-06 1.15E+07 2.00E-08 9.50E+03 2.81E-02 8.33E+00
1.62E-05 1.19E+07 2.18E-08 1.04E+04 3.03E-02 8.40E+00
8.85E-06 1.25E+07 2.08E-08 1.06E+04 3.17E-02 7.74E+00
1.11E-05 1.20E+07 2.43E-08 1.03E+04 3.06E-02 8.18E+00
7.62E-06 1.12E+07 2.28E-08 9.36E+03 3.18E-02 7.61E+00
3.96E-06 1.09E+07 2.11E-08 1.13E+04 3.08E-02 8.55E+00
1.23E-05 1.13E+07 2.12E-08 9.88E+03 3.15E-02 7.95E+00
7.39E-06 1.04E+07 2.20E-08 9.70E+03 2.98E-02 8.40E+00
1.09E-05 5.42E+07 3.79E-08 2.44E+04 4.59E-02 1.44E+01
1.89E-05 5.59E+07 3.91E-08 2.12E+04 5.20E-02 1.52E+01
1.16E-05 5.81E+07 3.38E-08 2.32E+04 5.28E-02 1.45E+01
1.16E-05 6.02E+07 3.28E-08 2.37E+04 5.23E-02 1.42E+01
9.98E-06 5.73E+07 3.54E-08 2.20E+04 5.51E-02 1.51E+01
7.70E-06 6.09E+07 3.58E-08 2.24E+04 5.44E-02 1.47E+01
9.94E-06 5.24E+07 3.32E-08 2.52E+04 5.06E-02 1.33E+01
1.13E-05 5.89E+07 3.90E-08 2.06E+04 4.53E-02 1.57E+01
1.45E-05 5.51E+07 3.58E-08 2.49E+04 4.81E-02 1.50E+01
7.70E-06 5.54E+07 3.86E-08 2.17E+04 5.20E-02 1.47E+01
1.71E-05 1.52E+08 6.45E-08 9.29E+03 9.05E-02 2.70E+01
1.83E-05 1.55E+08 6.85E-08 8.77E+03 8.91E-02 2.76E+01
1.95E-05 1.56E+08 6.32E-08 8.85E+03 9.01E-02 2.81E+01
1.79E-05 1.41E+08 6.30E-08 8.92E+03 9.41E-02 2.75E+01
1.99E-05 1.54E+08 6.87E-08 9.10E+03 9.69E-02 2.76E+01
1.84E-05 1.55E+08 6.72E-08 8.97E+03 9.22E-02 2.78E+01
1.72E-05 1.41E+08 6.59E-08 9.31E+03 9.63E-02 2.80E+01
1.74E-05 1.62E+08 6.29E-08 9.62E+03 9.32E-02 2.64E+01
1.73E-05 1.66E+08 6.53E-08 9.39E+03 8.09E-02 2.79E+01
1.85E-05 1.44E+08 6.54E-08 9.14E+03 8.91E-02 2.70E+01
1.99E-05 6.27E+08 7.46E-08 1.11E+04 7.61E-01 2.86E+01
1.89E-05 6.49E+08 8.04E-08 1.20E+04 7.42E-01 2.93E+01
1.91E-05 6.71E+08 7.75E-08 1.19E+04 7.78E-01 2.86E+01
1.88E-05 6.47E+08 7.71E-08 1.07E+04 7.26E-01 2.83E+01
1.92E-05 6.68E+08 7.80E-08 1.28E+04 7.08E-01 2.84E+01
1.94E-05 6.40E+08 7.97E-08 1.10E+04 6.56E-01 2.84E+01
1.95E-05 6.44E+08 8.07E-08 1.20E+04 7.78E-01 2.82E+01
1.92E-05 6.63E+08 7.94E-08 1.15E+04 7.30E-01 2.92E+01
1.89E-05 6.47E+08 7.52E-08 1.06E+04 7.92E-01 2.93E+01
1.82E-05 6.22E+08 7.81E-08 1.10E+04 7.17E-01 2.96E+01
136
Table D.16: IXYS IGBT raw data table (continued)
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
15 Ω 3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic Shot #
63.2
1
63.2
2
63.2
3
63.2
4
63.2
5
63.2
6
63.2
7
63.2
8
63.2
9
63.2
10
126
1
126
2
126
3
126
4
126
5
126
6
126
7
126
8
126
9
126
10
190
1
190
2
190
3
190
4
190
5
190
6
190
7
190
8
190
9
190
10
222
1
222
2
222
3
222
4
222
5
222
6
222
7
222
8
222
9
222
10
AGD
MUN
CGS
MUP
COXD
KP
1.42E-05 3.38E+07 1.75E-08 9.35E+03
2.71E-02 4.84E+00
1.43E-05 3.77E+07 1.64E-08 9.36E+03
2.49E-02 5.02E+00
6.13E-06 3.40E+07 1.80E-08 8.90E+03
2.58E-02 4.91E+00
9.50E-06 3.77E+07 1.68E-08 9.92E+03
2.53E-02 4.65E+00
7.15E-06 3.57E+07 1.71E-08 9.67E+03
2.71E-02 4.96E+00
6.09E-06 3.55E+07 1.67E-08 9.24E+03
2.53E-02 4.76E+00
5.98E-06 3.72E+07 1.77E-08 9.68E+03
2.46E-02 4.84E+00
6.30E-06 3.42E+07 1.72E-08 9.44E+03
2.74E-02 4.95E+00
1.20E-05 3.57E+07 1.76E-08 9.89E+03
2.56E-02 4.68E+00
2.64E-06 3.43E+07 1.66E-08 9.55E+03
2.85E-02 4.79E+00
8.70E-06 4.69E+06 2.27E-08 1.67E+04
1.48E-03 9.24E+00
9.43E-06 4.78E+06 2.37E-08 1.70E+04
1.41E-03 9.02E+00
6.13E-06 4.58E+06 2.18E-08 1.64E+04
1.49E-03 8.98E+00
1.13E-05 4.50E+06 2.06E-08 1.81E+04
1.52E-03 8.85E+00
4.25E-06 4.10E+06 2.37E-08 1.70E+04
1.65E-03 9.11E+00
6.10E-06 4.21E+06 2.18E-08 1.67E+04
1.49E-03 8.88E+00
7.10E-06 4.46E+06 2.10E-08 1.78E+04
1.70E-03 8.86E+00
7.80E-06 4.34E+06 2.21E-08 1.73E+04
1.60E-03 9.01E+00
9.37E-06 4.49E+06 2.14E-08 1.66E+04
1.49E-03 9.03E+00
8.89E-06 4.43E+06 1.96E-08 1.87E+04
1.71E-03 8.64E+00
8.34E-06 1.58E+09 3.13E-08 2.44E+04
1.99E-01 1.32E+01
1.12E-05 1.60E+09 3.13E-08 2.08E+04
1.69E-01 1.46E+01
1.10E-05 1.50E+09 3.15E-08 2.32E+04
1.71E-01 1.38E+01
1.76E-05 1.60E+09 3.01E-08 2.14E+04
1.94E-01 1.41E+01
7.80E-06 1.51E+09 3.37E-08 2.14E+04
1.69E-01 1.46E+01
8.43E-06 1.55E+09 3.07E-08 2.20E+04
1.79E-01 1.46E+01
7.28E-06 1.68E+09 3.30E-08 2.24E+04
1.87E-01 1.36E+01
9.33E-06 1.69E+09 2.94E-08 2.08E+04
1.78E-01 1.36E+01
1.43E-05 1.54E+09 3.35E-08 2.24E+04
1.80E-01 1.45E+01
1.08E-05 1.41E+09 2.93E-08 2.39E+04
1.97E-01 1.32E+01
1.28E-05 3.06E+07 3.30E-08 2.33E+04
1.26E-02 1.57E+01
1.28E-05 3.04E+07 3.45E-08 2.30E+04
1.25E-02 1.63E+01
1.47E-05 3.05E+07 3.68E-08 2.21E+04
1.32E-02 1.65E+01
1.31E-05 3.13E+07 3.62E-08 2.41E+04
1.27E-02 1.59E+01
1.14E-05 3.22E+07 3.53E-08 2.29E+04
1.25E-02 1.61E+01
1.46E-05 3.21E+07 3.57E-08 2.52E+04
1.17E-02 1.56E+01
1.25E-05 3.25E+07 3.49E-08 2.23E+04
1.21E-02 1.62E+01
9.54E-06 3.29E+07 3.20E-08 2.61E+04
1.31E-02 1.52E+01
1.27E-05 3.23E+07 3.55E-08 2.49E+04
1.14E-02 1.56E+01
1.58E-05 3.24E+07 3.47E-08 2.35E+04
1.26E-02 1.60E+01
137
E. Coefficient of Determination, R2, Values for
Simulated Waveforms
138
Table E.1: Powerex IGBT #1 R2 values
3Ω
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
300
300
300
300
300
300
300
300
300
300
588
588
588
588
588
588
588
588
588
588
736
736
736
736
736
736
736
736
736
736
752
752
752
752
752
752
752
752
752
752
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
139
VCE R2
I C R2
VGE R2
0.939491 0.964738 0.440901
0.944411 0.975885
0.43884
0.9512 0.973675 0.438499
0.943478 0.964897 0.447221
0.939176 0.958505 0.421586
0.944438 0.977774 0.436861
0.946519 0.977308 0.450451
0.941515 0.964458 0.439283
0.937499 0.960595 0.443442
0.936304 0.963401 0.434517
0.938669 0.964723 0.292279
0.936419 0.959065 0.304923
0.955524 0.975246
0.940184
0.29761
0.96101 0.285009
0.963759 0.976944 0.271842
0.941101 0.963764 0.315401
0.937685 0.964087 0.280267
0.968299 0.981057
0.23924
0.941233 0.965414 0.304866
0.963969
0.97537 0.266372
0.969881 0.978403 0.545271
0.95747 0.980983 0.566017
0.964611 0.981348
0.5646
0.968985 0.978469 0.551219
0.970091 0.980717 0.541377
0.965666 0.981304 0.549533
0.97764 0.977469 0.547753
0.973467
0.97716 0.541587
0.960476 0.978598 0.574722
0.962173 0.977149 0.545674
0.966532 0.979441 0.616968
0.960782 0.979154 0.624444
0.96115
0.97911 0.631263
0.966276 0.977068 0.616629
0.960946 0.976094 0.627608
0.966716 0.977952 0.625541
0.964347 0.978947 0.616366
0.973332
0.97576 0.629108
0.972314 0.977837 0.618756
0.966511 0.978552 0.620793
Table E.2: Powerex IGBT #1 R2 values (continued)
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
4.5 Ω 3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
206
206
206
206
206
206
206
206
206
206
412
412
412
412
412
412
412
412
412
412
600
600
600
600
600
600
600
600
600
600
704
704
704
704
704
704
704
704
704
704
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
140
VCE R2
I C R2
VGE R2
0.949257 0.964287 0.679371
0.928711 0.944241 0.665698
0.949738 0.970529 0.682835
0.94806 0.963526 0.681413
0.926054 0.939345 0.663092
0.947559 0.967327 0.682814
0.950841 0.964954
0.68386
0.926967 0.945476 0.667146
0.927343 0.947122 0.669054
0.952155 0.971349 0.668848
0.137306 0.183549 0.770267
0.147181 0.170726 0.775516
0.138064 0.173569 0.774766
0.132867 0.184267 0.771016
-1.58634 -1.558244 -0.779565
0.136998 0.145083 0.766248
0.135323 0.176526 0.767449
0.134895 0.181387 0.771565
0.137211 0.175808 0.765402
0.123932 0.151884 0.758613
0.962856 0.971517 0.307936
0.957549 0.969975 0.324222
0.975167 0.981361 0.337366
0.975891 0.980178 0.342563
0.96473 0.972815 0.338842
0.977314 0.982004 0.327518
0.971338 0.977276 0.315978
0.963853 0.970448 0.347106
0.976625 0.982412 0.313248
0.967255 0.974141 0.326987
0.975719 0.981089 0.490566
0.972476 0.981204 0.490164
0.971586 0.981924 0.478856
0.974693 0.981892 0.491585
0.975474 0.981548 0.481744
0.972521 0.981438 0.499059
0.973712 0.982225 0.487764
0.976075 0.982514 0.488149
0.971704 0.981578 0.491566
0.972258 0.981834 0.502501
Table E.3: Powerex IGBT #1 R2 values (continued)
9Ω
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
105
105
105
105
105
105
105
105
105
105
212
212
212
212
212
212
212
212
212
212
312
312
312
312
312
312
312
312
312
312
384
384
384
384
384
384
384
384
384
384
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
141
VCE R2
I C R2
VGE R2
0.938807 0.940748 0.758929
0.941393 0.950398 0.764566
0.945502 0.950997 0.770603
0.939858
0.94289 0.759158
0.941758 0.947235 0.764849
0.939394 0.936438 0.764583
0.938011 0.941094 0.757204
0.940044 0.944067 0.765463
0.942109 0.950013 0.763112
0.938937 0.935817 0.759286
0.963929 0.963247 0.767292
0.944786 0.956983 0.746536
0.955557 0.953034
0.76447
0.94436 0.955549 0.751598
0.974365 0.966505 0.772108
0.959822 0.961553 0.759944
0.933717 0.942507 0.744291
0.944214 0.956091 0.751581
0.966913 0.956699 0.771944
0.952514 0.958777 0.760763
0.954945 0.959768 0.603528
0.976499
0.97595 0.616992
0.97724 0.981317 0.610262
0.987322 0.984409 0.616655
0.975721 0.976225 0.595345
0.978435 0.984258 0.600643
0.978091 0.984494 0.594546
-0.64201 -0.583215 -1.292746
0.974766 0.976596
0.972058
0.60437
0.97068 0.611055
0.974385 0.975968 0.707183
0.978138 0.978878 0.701338
0.972484 0.971559 0.716485
0.973541 0.975165 0.714111
0.974458 0.973143 0.711768
0.717177 0.707684 0.282308
0.946091 0.947064 0.692169
0.973443 0.974936 0.723641
0.973149 0.973175 0.707109
0.981624 0.978238 0.700369
Table E.4: Powerex IGBT #1 R2 values (continued)
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
15 Ω 3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic Shot #
68.8
1
68.8
2
68.8
3
68.8
4
68.8
5
68.8
6
68.8
7
68.8
8
68.8
9
68.8
10
140
1
140
2
140
3
140
4
140
5
140
6
140
7
140
8
140
9
140
10
208
1
208
2
208
3
208
4
208
5
208
6
208
7
208
8
208
9
208
10
235
1
235
2
235
3
235
4
235
5
235
6
235
7
235
8
235
9
235
10
142
VCE R2
I C R2
VGE R2
0.972658 0.971605 0.739478
0.959948 0.962333
0.72473
0.970493 0.971609 0.739018
0.965911 0.968756 0.723762
0.975444 0.978809
0.73488
0.96866 0.971959 0.731802
0.968517 0.969613
0.73724
0.975602 0.974613 0.735576
0.965535 0.963522 0.735417
0.971871 0.969564
0.73733
0.960309 0.965185 0.703033
0.968574 0.972457 0.710319
-0.54967
-0.46635
-1.10976
0.948598 0.942651
0.7077
0.961417 0.958455 0.706924
0.957852
0.96242 0.703335
0.962853 0.968067 0.711557
0.953242 0.957832 0.700946
0.96265 0.969675 0.710019
0.965668 0.968771 0.711352
0.955569 0.950266 0.722825
-0.59201 -0.612152 -1.236781
0.976426 0.972017 0.726983
0.976531 0.973257 0.720495
0.966996 0.973121 0.714116
-0.58698 -0.536492 -1.252974
0.975842 0.968559 0.721489
-0.59019 -0.563149
-1.21822
-0.58908 -0.723765 -1.252402
-0.59179 -0.803346 -1.300433
0.980667
0.98616
0.72017
0.959618 0.962823 0.695756
0.959993 0.967143 0.687899
0.980647 0.980193 0.717783
0.979438 0.972394 0.717448
0.962575 0.965504
0.69027
-0.5997 -0.667981 -1.276694
0.964491 0.968249 0.689969
0.97932 0.983562 0.717535
0.980343
0.9809 0.718613
Table E.5: Powerex IGBT #2 R2 values
3Ω
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
304
304
304
304
304
304
304
304
304
304
596
596
596
596
596
596
596
596
596
596
720
720
720
720
720
720
720
720
720
720
736
736
736
736
736
736
736
736
736
736
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
143
VCE R2
IC R2
VGE R2
0.957072 0.991292 0.647956
0.956389 0.991053 0.741522
0.960689 0.990376 0.731128
0.961262 0.990101 0.676077
0.955855 0.990349 0.723979
0.96097 0.990511
0.69053
0.965096 0.991652 0.666093
0.960724 0.990304 0.722239
0.958904 0.991371 0.737238
0.959321 0.990991
0.73184
0.972045 0.987362 0.587905
0.974608 0.989428 0.593373
0.973118 0.990143 0.573038
0.975019
0.98956
0.61099
0.975289 0.989629 0.483429
0.970714
0.98653 0.505883
0.973511 0.989778 0.608335
0.974898
0.99011 0.722315
0.975492 0.989245
0.56875
0.972918 0.989811
0.54903
0.979012 0.988475 0.804013
0.977677 0.987754 0.813989
0.980161 0.987618 0.831322
0.980086
0.98743 0.845195
0.979995 0.986535 0.821746
0.978914 0.986719 0.820237
0.98016 0.987005 0.793965
0.981632 0.987453 0.826734
0.9812 0.987302
0.81764
0.980194 0.987162 0.810678
0.978073 0.985998 0.793877
0.980574 0.985491 0.806174
0.979596 0.984753 0.819843
0.976213
0.98479 0.795485
0.980824 0.985513 0.787127
0.980528 0.984755 0.792087
0.979477 0.985775 0.825032
0.979197 0.985732 0.770611
0.979176
0.98519 0.802682
0.976619 0.985713
0.8049
Table E.6: Powerex IGBT #2 R2 values (continued)
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
4.5 Ω 3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
212
212
212
212
212
212
212
212
212
212
424
424
424
424
424
424
424
424
424
424
624
624
624
624
624
624
624
624
624
624
688
688
688
688
688
688
688
688
688
688
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
144
VCE R2
IC R2
VGE R2
0.977052 0.979616 0.778061
0.97633 0.993089 0.803275
0.976903
0.9929 0.795248
0.973083 0.992687 0.828589
0.969789 0.989598 0.809099
0.970614 0.991085 0.806981
0.970902 0.991242 0.819254
0.973323 0.992479 0.826941
0.970711 0.992476 0.815518
0.973444 0.991427 0.791846
0.978147 0.988528 0.771764
0.975256 0.978403 0.694558
0.975602 0.985872 0.748647
0.976494 0.984047 0.759419
0.979589 0.989283 0.752721
0.97919 0.987243
0.69777
0.977344 0.986731 0.662475
0.980315 0.989706 0.745678
0.975801 0.986007
0.65051
0.976285 0.985707 0.686547
0.983532 0.988965
0.60414
0.982577 0.988818 0.574515
0.983724 0.989424 0.633755
0.981453 0.988576 0.586692
0.983071 0.989756 0.583179
0.980749 0.989283 0.654623
0.983248 0.989092 0.650829
0.982691 0.989301 0.724335
0.982812 0.988239 0.590915
0.983438 0.988967 0.575046
0.985554 0.989219 0.826521
0.985667 0.989367 0.816972
0.983952 0.988144 0.813425
0.985523 0.989828 0.816519
0.986051 0.989784
0.80417
0.985255 0.989597 0.820003
0.987192 0.990465 0.800348
0.985172 0.989545 0.819996
0.986647 0.989801 0.779844
0.984088 0.989736
0.82089
Table E.7: Powerex IGBT #2 R2 values (continued)
9Ω
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
Shot #
106.4
1
106.4
2
106.4
3
106.4
4
106.4
5
106.4
6
106.4
7
106.4
8
106.4
9
106.4
10
212
1
212
2
212
3
212
4
212
5
212
6
212
7
212
8
212
9
212
10
320
1
320
2
320
3
320
4
320
5
320
6
320
7
320
8
320
9
320
10
372
1
372
2
372
3
372
4
372
5
372
6
372
7
372
8
372
9
372
10
145
VCE R2
IC R2
VGE R2
0.985874 0.990314 0.839577
0.982858 0.990927 0.824254
0.98296 0.990552 0.839068
0.983696 0.990473 0.827022
0.984007 0.991144 0.824636
0.984034 0.991815 0.832426
0.983172 0.991197 0.832645
0.982439 0.989999 0.831984
0.984138
0.99014 0.835378
0.984176 0.990069 0.822321
0.985746 0.990359 0.830223
0.988013 0.990836
0.83423
0.988095 0.991132 0.828992
0.985775 0.990058 0.829025
0.987652 0.990817
0.82857
0.986618 0.991305 0.826846
0.987401 0.991258 0.835767
0.988339 0.991531
0.986296
0.83226
0.99025 0.827452
0.987679 0.988974 0.833995
0.994802 0.995789 0.785071
0.99403 0.993135 0.806132
0.990972 0.995763 0.822043
0.994908 0.995908 0.805797
0.990513 0.991336 0.811495
0.994839
0.99503 0.788036
0.991729 0.996309 0.800081
0.993155 0.995359 0.803906
0.994962 0.996639 0.796093
0.992122 0.993288 0.833163
0.989884 0.989739 0.786346
0.989714 0.989673 0.761376
0.989733 0.990414
0.77377
0.989438 0.990178 0.739598
0.98924
0.99048 0.750953
0.989866
0.99017 0.784795
0.990161
0.98995 0.753622
0.990059 0.989702 0.729837
0.989846 0.989914 0.721761
0.989776 0.989392 0.737481
Table E.8: Powerex IGBT #2 R2 values (continued)
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
15 Ω 3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
67.2
67.2
67.2
67.2
67.2
67.2
67.2
67.2
67.2
67.2
132
132
132
132
132
132
132
132
132
132
200
200
200
200
200
200
200
200
200
200
234
234
234
234
234
234
234
234
234
234
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
146
VCE R2
IC R2
VGE R2
0.989388 0.992531 0.794234
0.988185
0.99344 0.824297
0.988099 0.992302 0.814809
0.984517 0.988989 0.811736
0.986025 0.991968 0.831308
0.989064 0.992887 0.815421
0.989772 0.992262 0.819834
0.987544
0.9932 0.832115
0.985045 0.990344 0.811351
0.989756
0.99334 0.816583
0.992304 0.983653 0.854731
0.992453 0.991872 0.857226
0.996108 0.992893 0.848396
0.994112 0.992355 0.851906
0.995545 0.991179 0.840707
0.994767 0.990119 0.849648
0.995871 0.991456 0.848681
0.994766 0.992319 0.849303
0.993501 0.990714
0.85572
0.995511 0.993226 0.843109
0.994257 0.993617 0.846126
0.993864 0.993103 0.845438
0.995186 0.994467 0.841978
0.994852 0.990161 0.847394
0.994363
0.99382 0.839036
0.995529 0.992867 0.847088
0.989244
0.98976 0.860391
0.99558 0.991328 0.849481
0.995151 0.994206
0.84344
0.994835 0.993739 0.840472
0.994978 0.990413 0.846879
0.997155 0.995398 0.848435
0.996279 0.995118 0.848775
0.996282 0.994593 0.843164
0.997091 0.995939 0.852542
0.995938 0.996198 0.852968
0.995917 0.989401 0.854163
0.996183 0.994858 0.850226
0.996448 0.995708 0.848278
0.997057 0.995346 0.838323
Table E.9: Powerex IGBT #3 R2 values
3Ω
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
304
304
304
304
304
304
304
304
304
304
592
592
592
592
592
592
592
592
592
592
720
720
720
720
720
720
720
720
720
720
736
736
736
736
736
736
736
736
736
736
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
147
VCE R2
IC R2
VGE R2
0.958183 0.991377 0.724443
0.965489 0.991031 0.677134
0.963889 0.991178
0.73105
0.964971 0.990295 0.720982
0.963049 0.990804 0.689437
0.966865 0.992597 0.671012
0.961958 0.992042 0.731335
0.965627 0.992286 0.736775
0.961146
0.99031 0.693556
0.961884 0.990807 0.688032
0.974541 0.989494 0.709486
0.974633 0.989631 0.528427
0.974728 0.989757 0.607465
0.971367 0.988074 0.531334
0.975189 0.989755 0.561757
0.974224 0.989375 0.657698
0.971895 0.989242 0.556139
0.973054 0.989041 0.475543
0.971829 0.989137 0.579469
0.974095 0.989369 0.558437
0.97958 0.986853 0.775957
0.979702
0.98615
0.80687
0.979905
0.9864
0.819
0.978805 0.985221 0.791305
0.98028 0.986101
0.79568
0.980848 0.986467 0.786348
0.979876 0.985822 0.798318
0.978175 0.985846 0.808831
0.979971 0.986849 0.809154
0.978689 0.986695 0.802247
0.980749
0.98449 0.781859
0.979495 0.983672 0.764922
0.979919 0.986246
0.80321
0.980372 0.985774
0.80219
0.978033 0.983093 0.748843
0.98151 0.986139 0.815119
0.98055 0.985853 0.747223
0.980335 0.986167 0.771376
0.98048 0.985814 0.799845
0.98158 0.986073
0.78753
Table E.10: Powerex IGBT #3 R2 values (continued)
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
4.5 Ω 3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
210
210
210
210
210
210
210
210
210
210
424
424
424
424
424
424
424
424
424
424
616
616
616
616
616
616
616
616
616
616
696
696
696
696
696
696
696
696
696
696
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
148
VCE R2
IC R2
VGE R2
0.971259 0.990882 0.752777
0.973713 0.989824
0.76144
0.973669 0.994613 0.741074
0.969326 0.989791 0.755932
0.972341 0.989452 0.731211
0.976024 0.993586 0.725859
0.972666 0.993047
0.78386
0.974853 0.994151 0.750201
0.970702 0.991143 0.763126
0.9725 0.992243
0.76313
0.980004 0.988815 0.643249
0.977977 0.988413 0.603397
0.982284 0.990997 0.658381
0.981542 0.992314 0.676799
0.979551 0.988825 0.619957
0.982396 0.991731
0.66086
0.981938 0.991308
0.61987
0.980511 0.988589 0.700568
0.980425 0.989904 0.632191
0.980368
0.98995
0.984888
0.99118 0.643987
0.985008 0.990951
0.71976
0.61952
0.98503 0.992103 0.514678
0.984905 0.991611 0.694654
0.98493 0.991282 0.663131
0.985509 0.991589 0.625959
0.98495 0.992253 0.581666
0.984512
0.99177 0.630247
0.985156 0.990922
0.55627
0.983955 0.991425 0.547564
0.983474 0.984558 0.790053
0.981972 0.987647 0.816828
0.985369 0.987454 0.801755
0.984605 0.989008 0.821498
0.982921 0.988019 0.825022
0.983485 0.988575 0.809881
0.982538 0.987469 0.836973
0.981819 0.987696 0.848099
0.983638 0.987271 0.828463
0.982542 0.988833
0.84908
Table E.11: Powerex IGBT #3 R2 values (continued)
9Ω
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
Shot #
106.4
1
106.4
2
106.4
3
106.4
4
106.4
5
106.4
6
106.4
7
106.4
8
106.4
9
106.4
10
212
1
212
2
212
3
212
4
212
5
212
6
212
7
212
8
212
9
212
10
320
1
320
2
320
3
320
4
320
5
320
6
320
7
320
8
320
9
320
10
372
1
372
2
372
3
372
4
372
5
372
6
372
7
372
8
372
9
372
10
149
VCE R2
IC R2
VGE R2
0.980877 0.984406 0.848462
0.976383 0.990085 0.843148
0.979667 0.988775
0.84484
0.978942 0.988642 0.842741
0.978823 0.990367 0.845111
0.976767 0.988432 0.824896
0.976308 0.990312 0.837677
0.978608 0.990002 0.840995
0.978616 0.990123 0.836471
0.979807 0.989168
0.83861
0.986519 0.986388 0.853041
0.982886 0.989415 0.847357
0.983527 0.989929 0.860355
0.981103 0.989949 0.861717
0.985285 0.989689 0.848882
0.985043 0.987665 0.846184
0.976305 0.985977 0.856164
0.981437 0.987067 0.851822
0.983452 0.989862
0.85515
0.983813 0.989032 0.859001
0.990487 0.993354 0.828068
0.992514 0.995361 0.821301
0.99337 0.995433 0.798991
0.992992
0.99502 0.809304
0.992918
0.99558 0.785407
0.985359
0.99381 0.828307
0.994059 0.996267 0.816769
0.99392 0.995396 0.814095
0.990463 0.995018 0.812663
0.992069 0.995767 0.821822
0.989947 0.990779 0.756843
0.990379 0.990768 0.784362
0.990222 0.991341 0.777829
0.990447 0.990956 0.765436
0.990342 0.991084 0.751532
0.990367 0.991123 0.745888
0.990348 0.991306 0.770029
0.989827 0.991636 0.771963
0.989764 0.990996 0.764063
0.990325 0.990732 0.775638
Table E.12: Powerex IGBT #3 R2 values (continued)
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
15 Ω 3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
67.2
67.2
67.2
67.2
67.2
67.2
67.2
67.2
67.2
67.2
132
132
132
132
132
132
132
132
132
132
200
200
200
200
200
200
200
200
200
200
234
234
234
234
234
234
234
234
234
234
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
150
VCE R2
IC R2
VGE R2
0.984259 0.995541 0.847714
0.983685 0.987396 0.853641
0.979193 0.981049 0.850277
0.980362 0.968361 0.847576
0.977651 0.989111 0.852099
0.976902
0.99459 0.852229
0.982427 0.995121 0.848509
0.984097 0.994568 0.854367
0.974335 0.984435 0.850786
0.980412 0.995169 0.854024
0.98957 0.990644 0.846248
0.990884 0.992495 0.850069
0.995058 0.994702 0.841526
0.993006 0.989828 0.846653
0.993547 0.989615 0.846419
0.99306 0.993483 0.846766
0.990276 0.992894 0.843933
0.987542
0.99008 0.847251
0.993736 0.992276 0.845807
0.995343 0.991711 0.844513
0.992482 0.991584 0.846199
0.990019 0.994835 0.829309
0.993565 0.993684 0.839057
0.991102 0.991872 0.815307
0.994107
0.99313 0.827031
0.993924 0.989796 0.830448
0.995025
0.99429 0.843097
0.993674 0.993604 0.834644
0.993306 0.993103 0.839013
0.992728 0.993311 0.843991
0.99168 0.994392 0.851696
0.992025 0.996914 0.839883
0.988253 0.988565 0.855016
0.993904 0.989796
0.85109
0.996225 0.993932 0.836478
0.994045 0.994525
0.83984
0.987091 0.990734
0.85493
0.991896
0.99382 0.827767
0.993505 0.995106
0.83384
0.991509 0.995525 0.845934
Table E.13: IXYS IGBT R2 values
3Ω
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
288
288
288
288
288
288
288
288
288
288
352
352
352
352
352
352
352
352
352
352
364
364
364
364
364
364
364
364
364
364
368
368
368
368
368
368
368
368
368
368
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
151
VCE R2
IC R2
VGE R2
0.954427 0.965185 0.327454
0.921926 0.947574 0.354599
0.923951 0.949378
0.37971
0.925085 0.948335 0.354907
0.926693 0.946957 0.363618
0.923854 0.946015 0.350461
0.927657 0.940534 0.371713
0.92768 0.951601
0.33916
0.925186 0.946013 0.371229
0.925585 0.949252 0.340604
0.966563 0.979373 0.358838
0.970714 0.975316 0.329467
0.962146 0.975344 0.331602
0.961304 0.974849 0.346761
0.973103 0.972159 0.349889
0.970173 0.974018 0.387414
0.973246 0.976744 0.351518
0.957753 0.978422 0.352113
0.976437 0.974546 0.352677
0.967346
0.97816 0.336489
0.974226 0.972681 0.346119
0.96691 0.975661 0.334753
0.970788 0.965793 0.330962
0.968801
0.97001
0.35527
0.963181 0.969603 0.371838
0.97607 0.968873 0.347198
0.966748 0.973019
0.33377
0.968033 0.968045 0.328472
0.971725 0.971291
0.33782
0.964824 0.973504 0.327177
0.974532 0.971085
0.37808
0.967307 0.972142 0.385068
0.962147 0.971588 0.387301
0.955271 0.973149 0.363793
0.964376 0.972637 0.351633
0.970097 0.973457 0.356675
0.963402
0.97661 0.372793
0.968226 0.971713 0.386555
0.968974 0.974597 0.380946
0.966101 0.971479
0.37653
Table E.14: IXYS IGBT R2 values (continued)
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
4.5 Ω 3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
202
202
202
202
202
202
202
202
202
202
332
332
332
332
332
332
332
332
332
332
356
356
356
356
356
356
356
356
356
356
364
364
364
364
364
364
364
364
364
364
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
152
VCE R2
IC R2
0.885886 0.880933
VGE R2
0.32167
0.890328 0.882416 0.342919
0.90221 0.893138 0.345592
0.893098 0.885823 0.344389
0.889735 0.882667 0.326675
0.890289 0.898568 0.329412
0.89156 0.891624 0.329402
0.907347 0.911793 0.307651
0.888648 0.887121 0.320519
0.889837 0.880994 0.328249
0.971863 0.982155 0.339633
0.973524 0.976318 0.353706
0.976973 0.973431
0.35711
0.974755 0.980905 0.948328
0.976101 0.974981 0.331443
0.971656 0.980498 0.333981
0.975161 0.980996 0.356782
0.972705 0.975598 0.351082
0.972637 0.973273 0.366515
0.973458 0.984399 0.353347
0.972203 0.974017 0.376587
0.973239 0.977465 0.364139
0.974409
0.97634 0.364214
0.964752 0.978496 0.374489
0.9741 0.978569 0.389149
0.971556 0.979588 0.373687
0.974263 0.977007 0.355335
0.976018 0.973817 0.384335
0.971833 0.975729 0.380241
0.972103 0.967679 0.388626
0.970268 0.978467 0.380126
0.966205 0.975353 0.389719
0.966935 0.976205 0.344756
0.97288
0.97285 0.382233
0.971727 0.975982 0.372107
0.965331 0.977799 0.370663
0.969196
0.96732 0.387582
0.973446 0.979055 0.386235
0.977469
0.96957 0.393386
0.972966 0.971947 0.387764
Table E.15: IXYS IGBT R2 values (continued)
9Ω
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic
104
104
104
104
104
104
104
104
104
104
210
210
210
210
210
210
210
210
210
210
304
304
304
304
304
304
304
304
304
304
328
328
328
328
328
328
328
328
328
328
Shot #
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
153
VCE R2
IC R2
VGE R2
0.948589 0.949542 0.844192
0.93901 0.939611 0.836964
0.944661 0.947504 0.842848
0.898425
0.89208 0.782059
-0.9571 -0.871198 -1.449647
0.939877 0.938856 0.832472
0.944693 0.951839
0.83989
0.949987 0.938816 0.850118
0.940356 0.948704 0.834623
0.945624 0.947346 0.844192
0.949447 0.949806 0.847696
0.939019 0.938328 0.838874
0.948687 0.950467 0.846482
0.948611 0.953482 0.843759
0.948214 0.948993 0.847246
0.95134
0.95529 0.848443
0.946221 0.956532 0.843956
0.933699 0.925652 0.830273
0.943162 0.931407
0.84207
0.953771
0.95681 0.847292
0.954261
0.9627 0.880288
0.95347 0.958053 0.869421
0.951091 0.949328 0.879504
0.953152 0.957619
0.949365
0.87429
0.95415 0.870919
0.953134 0.953933 0.877099
0.968774 0.963719
0.89289
0.952916 0.966476 0.876342
0.954018 0.952685 0.863714
0.951849 0.959844 0.872888
0.980003 0.982069
0.92967
0.981075 0.973505 0.927509
0.980086 0.980667 0.930726
0.978951 0.987282 0.929598
0.980074 0.979024
0.93264
0.980558 0.986103 0.927707
0.979791 0.984907 0.930487
0.978494
0.97633 0.928615
0.979366 0.978295 0.927173
0.979656 0.973514
0.92914
Table E.16: IXYS IGBT R2 values (continued)
Vce
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
2000
2000
2000
2000
2000
2000
2000
15 Ω 3000
3000
3000
3000
3000
3000
3000
3000
3000
3000
3500
3500
3500
3500
3500
3500
3500
3500
3500
3500
Ic Shot #
63.2
1
63.2
2
63.2
3
63.2
4
63.2
5
63.2
6
63.2
7
63.2
8
63.2
9
63.2
10
126
1
126
2
126
3
126
4
126
5
126
6
126
7
126
8
126
9
126
10
190
1
190
2
190
3
190
4
190
5
190
6
190
7
190
8
190
9
190
10
222
1
222
2
222
3
222
4
222
5
222
6
222
7
222
8
222
9
222
10
154
VCE R2
IC R2
VGE R2
-1.08766 -1.271015 -1.489065
0.96483 0.944875 0.840991
-0.91456 -1.057673 -1.369561
0.956165 0.956136 0.829943
0.959434 0.949376 0.832938
0.959367 0.959446 0.833776
0.960342 0.956821 0.834038
0.960104 0.951316 0.830393
0.952305
0.95201 0.829097
0.963246 0.961469 0.837227
0.954841 0.948289 0.838224
0.953581 0.952048 0.836748
0.982587
0.98042 0.868805
0.977129
0.97299 0.869897
0.962324 0.960003 0.845077
0.983314 0.981285 0.870811
0.957898 0.956001
0.83799
0.957344 0.954264 0.837478
0.95422 0.953281 0.832727
0.978666 0.971485 0.860143
0.956078 0.953505 0.403514
0.909913 0.891104
0.49056
0.909103 0.895492 0.491686
0.94183 0.934048 0.432441
0.959226 0.941579
0.41506
0.959053 0.936254
0.40329
0.959619 0.958105 0.412455
0.943542 0.946411 0.427949
0.946696 0.923803 0.438119
0.90932 0.910108 0.483532
-1.21795 -1.266581 -1.719735
0.954912
0.94597 0.780313
-1.25474 -1.412497
-1.76396
0.954068 0.945791 0.781706
0.952005 0.947621 0.784641
0.974156 0.964048 0.808092
-1.20493 -1.304989
0.958018
-1.71669
0.95372 0.782359
0.955877 0.946721 0.782281
0.950047 0.940379 0.772625
Bibliography
[1] W.C. Nunnally, “Critical Component Requirements for Compact Pulse
Power System Architectures,” Plasma Science, IEEE Transactions on,
vol. 33, Aug. 2005, pp. 1262 - 1267.
[2] J.A. Gaudet, R.J. Barker, C.J. Buchenauer, C. Christodoulou, J. Dickens,
M.A. Gundersen, R.P. Joshi, H.G. Krompholz, J.F. Kolb, A. Kuthi, M.
Laroussi, A. Neuber, W. Nunnally, E. Schamiloglu, K.H. Schoenbach, J.S.
Tyo, and R.J. Vidmar, “Research issues in developing compact pulsed
power for high peak power applications on mobile platforms,”
Proceedings of the IEEE, vol. 92, Jul. 2004, pp. 1144 - 1165.
[3] M. Buttram, “Some future directions for repetitive pulsed power,” Plasma
Science, IEEE Transactions on, vol. 30, Feb. 2002, pp. 262 -266.
[4] J. Biela, C. Marxgut, D. Bortis, and J.W. Kolar, “Solid state modulator for
plasma channel drilling,” Dielectrics and Electrical Insulation, IEEE
Transactions on, vol. 16, Aug. 2009, pp. 1093 -1099.
[5] P.E. Gaudreau, J.A. Casey, M.A. Kempkes, T.J. Hawkey, and J.M.
Mulvaney, “Solid-state modulators for plasma immersion ion
implantation applications,” Journal of Vacuum Science Technology B:
Microelectronics and Nanometer Structures, vol. 17, Mar. 1999, pp. 888
-894.
[6] T. Namihira, S. Sakai, T. Yamaguchi, K. Yamamoto, C. Yamada, T.
Kiyan, T. Sakugawa, S. Katsuki, and H. Akiyama, “Electron Temperature
and Electron Density of Underwater Pulsed Discharge Plasma Produced
by Solid-State Pulsed-Power Generator,” Plasma Science, IEEE
Transactions on, vol. 35, Jun. 2007, pp. 614 -618.
[7] W. Hartmann, M. Romheld, and K.-. Rohde, “All-Solid-State Power
Modulator for Pulsed Corona Plasma Reactors,” Dielectrics and Electrical
Insulation, IEEE Transactions on, vol. 14, Aug. 2007, pp. 858 -862.
[8] A. Ludmirsky, E. Moshe, Y. Horovitz, M. Fraenkel, Z. Henis, and I.B.
Goldberg, “High power solid state single-frequency pulsed laser for
Doppler velocimeters,” Review of Scientific Instruments, vol. 71, Aug.
2000, pp. 2992 -2994.
[9] R. Petr, D. Reilly, J. Freshman, N. Orozco, D. Pham, L. Ngo, and J.
Mangano, “Solid-state pulsed power for driving a high-power dense
plasma focus x-ray source,” Review of Scientific Instruments, vol. 71,
Mar. 2000, pp. 1360 -1362.
155
[10] W. Jiang, K. Yatsui, K. Takayama, M. Akemoto, E. Nakamura, N.
Shimizu, A. Tokuchi, S. Rukin, V. Tarasenko, and A. Panchenko,
“Compact solid-State switched pulsed power and its applications,”
Proceedings of the IEEE, vol. 92, Jul. 2004, pp. 1180 - 1196.
[11] M.V. Fazio and H.C. Kirbie, “Ultracompact pulsed power,” Proceedings of
the IEEE, vol. 92, Jul. 2004, pp. 1197 - 1204.
[12] A. Benwell, J. VanGordon, D. Sullivan, B. Hutsel, S. Kovaleski, and J.
Gahl, “A 2.8 MV, 600 kA pulsed power driver constructed at the
University of Missouri-Columbia,” Plasma Science, 2007. ICOPS 2007.
IEEE 34th International Conference on, 2007, pp. 209 -209.
[13] M.A. Kemp, R.D. Curry, and S.D. Kovaleski, “Experimental study of the
multichanneling self-break section of the rimfire switch,” Plasma
Science, IEEE Transactions on, vol. 34, Feb. 2006, pp. 95 - 103.
[14] B.T. Hutsel, A. Benwell, S.D. Kovaleski, M.A. Kemp, D.L. Sullivan, and
J.M. Gahl, “Runtime and Jitter on a Laser-Triggered Spark-Gap Switch,”
Plasma Science, IEEE Transactions on, vol. 36, Oct. 2008, pp. 2541 2545.
[15] D.L. Sullivan, J.M. Gahl, S.D. Kovaleski, and B.T. Hutsel, “Study of laser
target triggering for spark gap switches,” Dielectrics and Electrical
Insulation, IEEE Transactions on, vol. 16, Aug. 2009, pp. 956 -960.
[16] N. Mohan, T. Undeland, and W. Robbins, Power Electronics Converters,
Applications, and Design, John Wiley & Sons, 2003.
[17] B.R. Geil, S.B. Bayne, D. Ibitayo, and M.G. Koebke, “Thermal and
Electrical Evaluation of SiC GTOs for Pulsed Power Applications,” Plasma
Science, IEEE Transactions on, vol. 33, Aug. 2005, pp. 1226 - 1234.
[18] R. Hironaka, M. Watanabe, E. Hotta, A. Okino, M. Maeyama, K.-. Ko, and
N. Shimizu, “Performance of pulsed power generator using high-voltage
static induction thyristor,” Plasma Science, IEEE Transactions on, vol.
28, Oct. 2000, pp. 1524 -1527.
[19] H. Singh and C.R. Hummer, “Advanced semiconductor switches for EM
launchers,” Magnetics, IEEE Transactions on, vol. 37, Jan. 2001, pp.
394 -397.
[20] T. Heeren, T. Ueno, D. Wang, T. Namihira, S. Katsuki, and H. Akiyama,
“Novel Dual Marx Generator for Microplasma Applications,” Plasma
Science, IEEE Transactions on, vol. 33, Aug. 2005, pp. 1205 - 1209.
156
[21] R.J. Baker, “High voltage pulse generation using current mode second
breakdown in a bipolar junction transistor,” Review of Scientific
Instruments, vol. 62, Apr. 1991, pp. 1031 -1036.
[22] H. O'Brien, W. Shaheen, R.L. Thomas, T. Crowley, S.B. Bayne, and C.J.
Scozzie, “Evaluation of Advanced Si and SiC Switching Components for
Army Pulsed Power Applications,” Magnetics, IEEE Transactions on, vol.
43, Jan. 2007, pp. 259 -264.
[23] M. Bhatnagar and B.J. Baliga, “Comparison of 6H-SiC, 3C-SiC, and Si for
power devices,” Electron Devices, IEEE Transactions on, vol. 40, Mar.
1993, pp. 645 -655.
[24] V. Khana, Insulated Gate Bipolar Transistors (IGBT):
Design, IEEE Press, 2003.
Theory and
[25] J. Yedinak, J. Merges, J. Wojslawowicz, A. Bhalla, D. Burke, and G.
Dolny, “Operation of an IGBT in a self-clamped inductive switching
circuit (SCIS) for automotive ignition,” Power Semiconductor Devices
and ICs, 1998. ISPSD 98. Proceedings of the 10th International
Symposium on, 1998, pp. 399 -402.
[26] P.M. Shenoy, S. Shekhawat, and B. Brockway, “Application specific
1200V planar and trench IGBTs,” Applied Power Electronics Conference
and Exposition, 2006. APEC '06. Twenty-First Annual IEEE, 2006, p. 5
pp.
[27] L. Schnur, G. Debled, S. Dewar, and J. Marous, “Low inductance,
explosion robust IGBT modules in high power inverter applications,”
Industry Applications Conference, 1998. Thirty-Third IAS Annual
Meeting. The 1998 IEEE, 1998, pp. 1056 -1060 vol.2.
[28] H.J. Ryoo, J.S. Kim, G.H. Rim, and G. Goussev, “High Repetitive Pulsed
Power Modulator Based on IGBT switches for PSII Application,” Plasma
Science, 2007. ICOPS 2007. IEEE 34th International Conference on,
2007, pp. 942 -942.
[29] W. Jiang, N. Oshima, T. Yokoo, K. Nakahiro, H. Honma, K. Takayama, M.
Wake, N. Shimizu, and A. Tokuchi, “Repetitive Pulsed Power Based on
Semiconductor Switching Devices,” Plasma Science, 2007. ICOPS 2007.
IEEE 34th International Conference on, 2007, pp. 421 -421.
[30] D. Bortis, J. Biela, and J.W. Kolar, “Active Gate Control for Current
Balancing of Parallel-Connected IGBT Modules in Solid-State
Modulators,” Plasma Science, IEEE Transactions on, vol. 36, Oct. 2008,
157
pp. 2632 -2637.
[31] V. Zorngiebel, E. Spahn, G. Buderer, A. Welleman, and W. Fleischmann,
“Compact High Voltage IGBT Switch for Pulsed Power Applications,”
Magnetics, IEEE Transactions on, vol. 45, Jan. 2009, pp. 531 -535.
[32] J. Kim, B. Min, S.V. Shenderey, and G. Rim, “High Voltage Pulsed Power
Supply Using IGBT Stacks,” Dielectrics and Electrical Insulation, IEEE
Transactions on, vol. 14, Aug. 2007, pp. 921 -926.
[33] M. Giesselmann, B. Palmer, A. Neuber, and J. Donlon, “High Voltage
Impulse Generator Using HV-IGBTs,” Pulsed Power Conference, 2005
IEEE, 2005, pp. 763 -766.
[34] J. Qiu, K. Liu, and Y. Wu, “A Pulsed Power Supply Based on Power
Semiconductor Switches and Transmission Line Transformer,” Dielectrics
and Electrical Insulation, IEEE Transactions on, vol. 14, Aug. 2007, pp.
927 -930.
[35] J. Kim, B. Min, S. Shenderey, and G. Rim, “High Voltage Marx Generator
Implementation using IGBT Stacks,” Dielectrics and Electrical Insulation,
IEEE Transactions on, vol. 14, Aug. 2007, pp. 931 -936.
[36] Spectrum Software, Micro-Cap 9 Professional Version.
[37] G. Oziemkiewicz, “Implementation and Development of the NIST IGBT
Model in a SPICE-based Commercial Circuit Simulator,” Degree of
Engineer, University of Florida, 1995.
[38] Powerex, Inc., “Powerex QIS4506001 Datasheet,” 2006.
[39] IXYS, “IXYS IXEL40N400 Datasheet,” 2009.
[40] The MathWorks, Inc., MATLAB.
[41] K. Sheng, S.J. Finney, and B.W. Williams, “Fast and accurate IGBT
model for PSpice,” Electronics Letters, vol. 32, Dec. 1996, pp. 2294 2295.
[42] A.F. Petrie and C. Hymowitz, “A Spice Model for IGBTs,” Applied Power
Electronics Conference and Exposition, vol. 1, 1995, pp. 147-152.
[43] K. Sheng, B.W. Williams, and S.J. Finney, “A review of IGBT models,”
Power Electronics, IEEE Transactions on, vol. 15, Nov. 2000, pp. 1250 1266.
[44] A.R.
Hefner,
“Device
Models,
Circuit
158
Simulation,
And
Computer-
controlled Measurements For The IGBT,” Computers
Electronics, 1990 IEEE Workshop on, 1990, pp. 233 -243.
in
Power
[45] A.R. Hefner and D.M. Diebolt, “An experimentally verified IGBT model
implemented in the Saber circuit simulator,” Power Electronics
Specialists Conference, 1991. PESC '91 Record., 22nd Annual IEEE,
1991, pp. 10 -19.
[46] A.R. Hefner, “Analytical modeling of device-circuit interactions for the
power insulated gate bipolar transistor (IGBT),” Industry Applications,
IEEE Transactions on, vol. 26, Dec. 1990, pp. 995 -1005.
[47] A.R. Hefner, “Modeling buffer layer IGBTs for circuit simulation,” Power
Electronics Specialists Conference, 1993. PESC '93 Record., 24th Annual
IEEE, 1993, pp. 60 -69.
[48] A. Hefner and D. Blackburn, “An Analytical Model for the Steady-State
and Transient Characteristics of the Power Insulated-Gate Bipolar
Transistor,” Solid-State Electronics, vol. 31, 1988, pp. 1513-1532.
[49] D. Berning, J. Reichl, A. Hefner, M. Hernandez, C. Ellenwood, and J.-.
Lai, “High speed IGBT module transient thermal response measurements
for model validation,” Industry Applications Conference, 2003. 38th IAS
Annual Meeting. Conference Record of the, 2003, pp. 1826 - 1832 vol.3.
[50] A.R. Hefner, “An improved understanding for the transient operation of
the power insulated gate bipolar transistor (IGBT),” Power Electronics,
IEEE Transactions on, vol. 5, Oct. 1990, pp. 459 -468.
[51] A.R. Hefner, “An investigation of the drive circuit requirements for the
power insulated gate bipolar transistor (IGBT),” Power Electronics, IEEE
Transactions on, vol. 6, Apr. 1991, pp. 208 -219.
[52] A.R. Hefner, “A dynamic electro-thermal model for the IGBT,” Industry
Applications, IEEE Transactions on, vol. 30, Apr. 1994, pp. 394 -405.
[53] A. Hefner and S. Bouche, “Automated parameter extraction software for
advanced IGBT modeling,” Computers in Power Electronics, 2000.
COMPEL 2000. The 7th Workshop on, 2000, pp. 10 -18.
[54] H.A. Mantooth and A.R. Hefner, “Electrothermal simulation of an IGBT
PWM inverter,” Power Electronics, IEEE Transactions on, vol. 12, May.
1997, pp. 474 -484.
[55] C. Shen, A.R. Hefner, D.W. Berning, and J.B. Bernstein, “Failure
dynamics of the IGBT during turn-off for unclamped inductive loading
159
conditions,” Industry Applications, IEEE Transactions on,
2000, pp. 614 -624.
vol. 36, Apr.
[56] D.W. Berning and A.R. Hefner, “IGBT half-bridge shoot-through
characterization for model validation,” Industry Applications Conference,
1996. Thirty-First IAS Annual Meeting, IAS '96., Conference Record of
the 1996 IEEE, 1996, pp. 1491 -1499 vol.3.
[57] D.W. Berning and A.R. Hefner, “IGBT model validation,” Industry
Applications Magazine, IEEE, vol. 4, Dec. 1998, pp. 23 -34.
[58] D.W. Berning and A.R. Hefner, “IGBT model validation for soft-switching
applications,” Industry Applications, IEEE Transactions on, vol. 37, Apr.
2001, pp. 650 -660.
[59] A.G. Strollo, “A new IGBT circuit model for SPICE simulation,” Power
Electronics Specialists Conference, 1997. PESC '97 Record., 28th Annual
IEEE, 1997, pp. 133 -138 vol.1.
[60] D.M. Caughey and R.E. Thomas, “Carrier mobilities in silicon empirically
related to doping and field,” Proceedings of the IEEE, vol. 55, Dec.
1967, pp. 2192 - 2193.
[61] D. Neamen, Microelectronics Circuit Analysis and Design,
McGraw-Hill, 2007.
New York:
[62] Micrel, Inc., “Micrel MIC4451/4452 Datasheet,” 2005.
[63] M.A. Kemp, S.D. Kovaleski, B.T. Hutsel, A. Benwell, and J.M. Gahl,
“Particle Swarm Optimization of Pulsed Power Circuit Models,” Plasma
Science, IEEE Transactions on, vol. 36, Oct. 2008, pp. 2722 -2729.
[64] B. Baliga, Fundamentals of Power Semiconductor Devices, Springer,
2008.
160