Characteristics and Working Principle of IGBT

Catalog

Ⅰ Introduction

IGBT, Insulated Gate Bipolar Transistor, is a composite fully controlled voltage-driven power semiconductor device composed of BJT (bipolar junction transistor) and MOSFET (Metal-Oxide-Semiconductor Field-Effect Transistor). It combines the advantages of both the high input impedance of MOSFET and the low on-state voltage drop of BJT. While GTR (Giant Transistor Rectifier) has reduced saturation voltage and high current-carrying density, it requires larger driving current. MOSFET has very small drive power and fast switching speed, but suffers from high conduction voltage drop and small current-carrying density. IGBT integrates the advantages of both devices, offering small driving power and reduced saturation voltage. It is particularly suitable for converter systems with DC voltages of 600V and above, including applications in AC motor drives, inverters, switching power supplies, lighting circuits, traction drives, renewable energy systems, and electric vehicle powertrains.

An IGBT module is a modular semiconductor product that packages IGBT chips and FWD (freewheeling diodes) through a specific circuit bridge configuration. These packaged IGBT modules are directly applied to equipment such as inverters, UPS (uninterruptible power supplies), motor drives, and renewable energy converters. IGBT modules feature energy-saving capabilities, convenient installation and maintenance, and stable heat dissipation. The term "IGBT" commonly refers to both discrete devices and IGBT modules. With the advancement of energy conservation and environmental protection initiatives, these products have become increasingly prevalent in the market. IGBT is the core device for energy conversion and transmission, often referred to as the "CPU" of power electronic devices, and is widely used in rail transit, smart grids, aerospace, electric vehicles, renewable energy equipment, industrial automation, and consumer electronics.

Ⅱ Structure of IGBT

The figure below shows an N-channel enhanced insulated gate bipolar transistor structure. The N+ region is called the source region, with its electrode designated as the source (emitter E). The N- base is called the drift region. The control area of the device is the gate region, with its electrode called the gate (G). The channel forms at the boundary of the gate region. The P-type region between the collector (C) and emitter (E) terminals is called the sub-channel region. The P+ area on the collector side of the drift region is called the collector injector. This is a unique functional area of the IGBT that forms a PNP bipolar transistor together with the drift region and the sub-channel area. It acts as an emitter, injecting holes into the drift region, enabling conductivity modulation, and reducing the on-state voltage of the device. The electrode on the collector injection area is called the collector (C).

N-channel enhanced insulated gate bipolar transistor structure

The switching function of the IGBT is achieved by applying a positive gate voltage to form a channel, which provides base current to the PNP transistor to turn on the IGBT. Conversely, applying a negative or zero gate voltage eliminates the channel and cuts off the base current, turning off the IGBT. The driving method of IGBT is fundamentally similar to that of MOSFET, requiring only control of the input N-channel MOSFET gate, resulting in high input impedance characteristics. After the MOSFET channel is formed, holes (minority carriers) are injected from the P+ base into the N- layer, modulating the conductance of the N- layer and reducing its resistance. Consequently, the IGBT maintains a low on-state voltage even at high voltage levels.

IGBT is a three-terminal device with gate (G), collector (C), and emitter (E) terminals. The structure and simplified equivalent circuit of the IGBT are shown in the figure below.

IGBT internal structure and equivalent circuit 

The figure shows the cross-sectional schematic diagram of the internal structure of the N-IGBT, combining the N-channel VDMOSFET and BJT. IGBT has one additional P+ implantation layer compared to VDMOSFET, forming a large-area PN junction J1. The P+ injection region emits minority carriers into the N- base region when the IGBT is turned on, modulating the conductivity of the drift region and enabling the IGBT to carry substantial current. The N+ layer between the P+ injection region and the N- drift region is called the buffer zone. The presence or absence of this buffer determines different IGBT characteristics. IGBTs with an N+ buffer are called asymmetric or punch-through (PT) IGBTs. They feature advantages such as small forward voltage drop, short turn-off time, and small tail current during turn-off, but have relatively weak reverse blocking capability. IGBTs without an N+ buffer are called symmetric or non-punch-through (NPT) IGBTs. They possess strong forward and reverse blocking capability, but their other characteristics are not as optimal as asymmetric IGBTs.

The simplified equivalent circuit demonstrates that the IGBT functions as a Darlington configuration composed of a BJT and MOSFET. This structure is driven by the MOSFET, while the other part consists of a thick-base PNP transistor.

The ideal equivalent circuit and actual equivalent circuit of IGBT

From the equivalent circuit perspective, the IGBT can be understood as a monolithic Bi-MOS transistor formed by a Darlington connection of a PNP bipolar transistor and a power MOSFET.

When a positive voltage is applied between the gate and emitter to turn on the power MOSFET, the base-collector junction of the PNP transistor is connected through low resistance, placing the PNP transistor in a conductive state. The P+ layer injects holes into the N- base during the on-state, triggering conductivity modulation. This enables extremely low on-resistance compared to power MOSFETs.

When the gate-emitter voltage is reduced to 0V or below the threshold voltage, the power MOSFET enters the off-state, and the base current of the PNP transistor is cut off, placing it in the off-state.

As described above, the IGBT, like the power MOSFET, can control turn-on and turn-off through voltage signals, making it a voltage-controlled device with minimal gate drive requirements.

Ⅲ Working Characteristics

1. Static Characteristics

The static characteristics of IGBT primarily include volt-ampere characteristics and transfer characteristics.

The volt-ampere characteristic of IGBT refers to the relationship curve between collector current (I_C) and collector-emitter voltage (V_CE) when the gate-emitter voltage (V_GE) is used as a parameter. The output collector current is controlled by the gate-emitter voltage V_GE. Higher V_GE results in larger I_C. This characteristic is similar to the output characteristics of BJT and can be divided into three regions: the saturation region (linear region), active region (amplification region), and breakdown region. In the off-state of IGBT, forward voltage is supported by the J2 junction, while reverse voltage is supported by the J1 junction. Without an N+ buffer layer, forward and reverse blocking voltages can reach similar levels. With the addition of an N+ buffer layer, reverse blocking voltage is limited to tens of volts, which restricts certain IGBT applications requiring bidirectional blocking capability.

The transfer characteristic of IGBT refers to the relationship curve between the output collector current I_C and the gate-emitter voltage V_GE. It exhibits similar transfer characteristics to MOSFETs. When the gate-emitter voltage is less than the threshold voltage V_GE(th) (typically 3-6V for modern IGBTs), the IGBT remains in the off-state. Throughout most of the collector current range after the IGBT turns on, I_C maintains a linear relationship with V_GE. The maximum gate-emitter voltage is limited by the maximum collector current rating and device reliability considerations, with optimal values typically around 15V for standard IGBTs and up to 20V for some high-performance devices.

2. Dynamic Characteristics

Dynamic characteristics, also called switching characteristics, are critical for IGBT performance in practical applications. The switching characteristics of IGBTs comprise two main aspects: switching speed, indicated by the duration of each phase of the switching process; and switching losses, which occur during transitions between on and off states.

The switching characteristic of IGBT describes the relationship between collector current and collector-emitter voltage during transitions. When the IGBT is in the ON state, its current gain (β) is relatively low because the PNP transistor has a wide base region. Although the equivalent circuit forms a Darlington structure, the current flowing through the MOSFET constitutes the main portion of the total IGBT current. At this time, the on-state voltage V_CE(on) can be expressed by the following formula:

V_CE(on) = V_J1 + V_dr + I_CR_ch

Where V_J1 is the forward voltage of the J1 junction, typically 0.7~1V; V_dr is the voltage drop across the drift region resistance R_dr; and R_ch is the channel resistance.

The on-state current I_C can be expressed by the following formula:

I_C = (1 + β_PNP)I_MOS

Where I_MOS is the current flowing through the MOSFET portion.

Due to the conductivity modulation effect in the N- drift region, the on-state voltage drop of the IGBT is relatively small. Modern IGBTs with blocking voltages of 1200V typically exhibit on-state voltage drops of 1.5 to 2.5V at rated current, while 1700V devices show 2 to 3V. When the IGBT is in the off-state, only a small leakage current exists, typically in the microampere to milliampere range depending on voltage rating and temperature.

During the turn-on process, the IGBT initially operates as a MOSFET. In the late stage of the collector-emitter voltage V_CE falling process, the PNP transistor transitions from the active region to saturation, introducing additional delay time. The turn-on delay time t_d(on) represents the initial delay, while t_ri represents the current rise time. In practical applications, the collector current turn-on time t_on is the sum of t_d(on) and t_ri, while the collector-emitter voltage fall time comprises t_fv1 and t_fv2.

Triggering and turning off IGBT requires applying positive and negative (or zero) voltages between its gate and emitter. Gate voltage can be generated by various driving circuits. When selecting drive circuits, consideration must be given to: device turn-off bias requirements, gate charge requirements (Q_g), isolation requirements, and power supply conditions. Because the IGBT gate-emitter impedance is high, MOSFET driving techniques can be adapted for IGBT triggering. However, since the input capacitance of IGBTs is larger than that of equivalent MOSFETs, IGBT gate drive circuits typically require higher current capability and may benefit from negative turn-off bias (typically -5V to -15V) to ensure fast and reliable turn-off.

During IGBT turn-off, the collector current waveform exhibits two distinct segments. After the MOSFET portion turns off, stored charge in the PNP transistor cannot be quickly eliminated, resulting in a characteristic tail current with extended duration. The turn-off delay time t_d(off) represents the initial delay, while t_rv is the rise time of V_CE. The fall time t_fi of the collector current, commonly specified in datasheets, comprises the initial fall time t_fi1 and the tail current fall time t_fi2. The total turn-off time t_off = t_d(off) + t_rv + t_fi. The sum of t_d(off) and t_rv is also called storage time.

The switching speed of IGBT is lower than that of MOSFET but significantly higher than traditional BJTs. Modern IGBTs can achieve switching frequencies from several kHz to over 100 kHz, with silicon carbide (SiC) and gallium nitride (GaN) alternatives pushing frequencies even higher. IGBTs do not require negative gate voltage to reduce turn-off time, though negative bias can improve turn-off speed and reduce switching losses. Turn-off time increases with the gate-emitter parallel resistance. The threshold voltage of modern IGBTs is typically 3 to 6V. The saturation voltage drop when the IGBT is turned on is lower than that of MOSFETs and approaches that of BJTs, decreasing with increasing gate voltage up to the recommended maximum gate voltage.

As of 2025, commercially available IGBT devices have significantly expanded their voltage and current capabilities. High-voltage IGBTs now routinely reach 6.5kV with current ratings exceeding 1000A, while ultra-high-voltage devices of 10kV and above are available for specialized applications. Advanced technologies including field-stop (FS) trench-gate structures, improved lifetime control techniques, and enhanced chip designs have enabled these improvements. Major manufacturers including Infineon, Mitsubishi Electric, Fuji Electric, and ON Semiconductor continue developing IGBTs with higher voltage ratings, lower losses, and improved thermal performance. Silicon carbide (SiC) MOSFETs are increasingly competing with IGBTs in certain applications, particularly where higher switching frequencies and efficiency are critical, though IGBTs maintain advantages in cost-effectiveness for many medium-frequency, high-power applications.

Ⅳ Working Principle of IGBT

1. Turn On

The structure of the IGBT silicon chip is very similar to that of the power MOSFET. The main difference is that the IGBT incorporates a P+ substrate and an N+ buffer layer. The MOSFET portion drives the bipolar portion of the device. The substrate creates a J1 junction between the P+ and N- regions. When positive gate bias causes inversion of the P-base region under the gate, an N-channel forms. Simultaneously, an electron current appears, generated in exactly the same manner as in a power MOSFET. If the voltage drop produced by this electron flow across the N- drift region reaches approximately 0.7V, the J1 junction becomes forward biased. Holes are then injected from the P+ substrate into the N- drift region, modulating its resistivity and reducing resistance between collector and emitter. This results in reduced total power conduction losses and initiates a second charge carrier flow. The final result is two different current components within the semiconductor structure: an electron flow (MOSFET current) and a hole current (bipolar current). This conductivity modulation is the key mechanism enabling IGBTs to achieve low on-state voltage drops even at high voltage ratings.

2. Shut Down

When negative bias is applied to the gate or the gate voltage falls below the threshold, the channel is eliminated and hole injection into the N- region ceases. However, if the MOSFET current drops rapidly during the switching phase, the collector current decreases gradually. This is because, after commutation begins, minority carriers (holes) remain in the N- drift layer. The reduction of this residual current (tail current) depends entirely on the recombination of minority carriers, which is influenced by several factors including doping concentration and profile, layer thickness, lifetime control techniques, and junction temperature. The decay of minority carriers causes the collector current to exhibit a characteristic tail current waveform. This tail current causes the following issues: 1. Increased switching losses and reduced efficiency; 2. Cross-conduction problems in bridge circuits, particularly in devices using freewheeling diodes; 3. Limitations on maximum switching frequency.

Since the tail current is related to minority carrier recombination, its magnitude is closely related to chip temperature and hole mobility, which correlate with I_C and V_CE. Therefore, depending on the operating temperature, it is possible to reduce this undesirable tail current effect through proper thermal management and gate drive optimization in the terminal equipment design. Modern IGBT designs employ various lifetime control techniques, including electron irradiation, proton irradiation, and optimized doping profiles, to minimize tail current while maintaining acceptable on-state voltage.

3. Blocking and Latching

When reverse voltage is applied to the collector (negative with respect to emitter), the J1 junction becomes reverse biased, and the depletion layer extends into the N- drift region. If this layer is too thin, effective blocking capability cannot be achieved. Therefore, proper drift region design is critical. Conversely, if this region is excessively thick, on-state voltage drop increases continuously. This trade-off explains why NPT (non-punch-through) devices typically exhibit higher on-state voltage drops than equivalent PT (punch-through) devices with the same current rating and switching speed.

When the gate and emitter are shorted and positive voltage is applied to the collector terminal, the P/N J3 junction (or J2 junction in some nomenclatures) is reverse biased. The depletion layer in the N- drift region supports the externally applied voltage, enabling forward blocking capability.

Under certain conditions, the parasitic thyristor structure inherent in the IGBT can be triggered, leading to latch-up. The main differences between static and dynamic latch-up are as follows:

Static latch-up occurs when the parasitic thyristor is fully turned on and cannot be turned off by the gate signal. Dynamic latch-up occurs during the turn-off transient when high dI_C/dt or dV_CE/dt triggers the parasitic thyristor. This phenomenon severely limits the safe operating area (SOA). To prevent the harmful latch-up phenomenon caused by parasitic NPN and PNP transistors, several measures are necessary: prevent the NPN portion from turning on through proper layout design; optimize doping profiles and concentrations; reduce the total current gain of the NPN and PNP transistors through lifetime control; implement low-inductance layouts to minimize voltage spikes; and ensure adequate gate drive with negative turn-off bias. Additionally, the latching current has a direct relationship with the current gains of the PNP and NPN parasitic devices. Therefore, it is closely related to junction temperature; as junction temperature and gain increase, the resistivity of the P-base region increases, degrading overall device characteristics and reducing latch-up immunity. Device manufacturers maintain a specific ratio between maximum collector current and latching current, typically 1:5 or higher, to ensure adequate safety margin. Modern IGBT designs with optimized cell structures and improved processing techniques have significantly increased latch-up immunity, with some devices achieving ratios exceeding 1:10.

Ⅴ History of IGBT

In 1979, MOS-gated power switching devices were introduced as the conceptual predecessor to the IGBT. These devices featured a thyristor-like structure (P-N-P-N four-layer composition), characterized by V-groove gates formed through strong alkaline wet etching processes.

In the early 1980s, the DMOS (Double Diffusion Metal-Oxide-Semiconductor) process used in power MOSFET manufacturing was adapted for IGBT production. The first commercial IGBTs were introduced by several companies including General Electric and RCA. At that time, silicon chip structures employed thicker NPT (non-punch-through) designs. Subsequently, through adoption of PT (punch-through) structures, significant improvements in parameter trade-offs were achieved. This advancement resulted from technological progress in epitaxial growth on silicon wafers and implementation of N+ buffer layers designed for specific blocking voltages. Over subsequent years, DMOS planar gate structure design rules for epitaxial PT wafers advanced from 5 micrometers to 3 micrometers, then to sub-micron dimensions.

In the mid-1990s, trench-gate structures emerged as a new IGBT concept, enabled by silicon dry etching technology borrowed from large-scale integration (LSI) processes. These devices initially maintained punch-through (PT) chip structures. The trench structure achieved more significant improvements in the trade-off between on-state voltage and turn-off time compared to planar gate designs, enabling higher current densities and lower conduction losses.

Punch-through (PT) technology features relatively high carrier injection efficiency but requires minority carrier lifetime control, which can deteriorate transport efficiency. Non-punch-through (NPT) technology, not requiring lifetime killing, maintains good transport efficiency but has relatively lower carrier injection efficiency. Subsequently, NPT technology evolved into field-stop (FS) technology, also called "soft punch-through" (SPT) or light punch-through (LPT), which further improved overall cost-performance characteristics by combining advantages of both approaches.

In 1996, CSTBT (Carrier Stored Trench-gate Bipolar Transistor) technology enabled fifth-generation IGBT modules, adopting field-stop chip structures with advanced wide-cell-pitch designs. This generation achieved significant reductions in both conduction and switching losses.

From 2000 onwards, IGBT technology continued advancing through multiple generations. The sixth generation (circa 2000-2005) introduced improved trench structures and optimized field-stop designs. The seventh generation (circa 2005-2012) featured further refinements in trench technology and micro-pattern trench (MPT) structures. Recent generations (2012-present) have incorporated advanced features including reverse-conducting (RC-IGBT) designs that integrate the antiparallel diode functionality, and ultra-thin wafer technology enabling higher power density. As of 2025, cutting-edge IGBTs feature switching frequencies exceeding 100 kHz, junction temperatures up to 175°C (with some devices rated for 200°C), and significantly improved short-circuit withstand capability.

Today, high-current and high-voltage IGBTs are extensively modularized, with integrated dedicated IGBT gate driver circuits manufactured alongside power modules. Modern IGBT modules offer superior performance, higher system reliability, and more compact form factors. Advanced packaging technologies including direct bonded copper (DBC) substrates, advanced solder materials, and improved thermal interface materials have enhanced thermal performance and reliability. Intelligent power modules (IPMs) integrate IGBTs with gate drivers, protection circuits, and temperature sensors in single packages, simplifying application design and improving system reliability.

Ⅵ Modern Applications and Future Trends

As of 2025, IGBT technology continues to evolve and expand into new application areas:

Key Applications:

• Electric Vehicles (EVs): IGBTs remain dominant in EV traction inverters for voltages up to 800V, though SiC MOSFETs are gaining market share in premium vehicles

• Renewable Energy: Solar inverters, wind turbine converters, and energy storage systems extensively use IGBT technology

• Industrial Drives: Motor control applications from fractional horsepower to multi-megawatt systems

• Rail Transportation: Traction systems for trains, metros, and high-speed rail

• Smart Grid: HVDC transmission, FACTS devices, and grid-scale energy storage

• Home Appliances: Air conditioners, refrigerators, induction cookers, and washing machines

Future Trends:

• Competition with wide-bandgap semiconductors (SiC, GaN) in certain applications

• Continued improvements in silicon IGBT technology to maintain cost-competitiveness

• Integration of advanced sensing and protection features

• Development of higher temperature operation capabilities (200°C+)

• Enhanced packaging technologies for improved thermal and electrical performance

• Hybrid modules combining IGBTs with SiC diodes for optimized performance

Ⅶ Frequently Asked Questions (FAQs)

Q1: What is the main difference between IGBT and MOSFET?

A: The main difference is that IGBT combines MOSFET and bipolar transistor structures, providing lower on-state voltage drop than MOSFET at high voltages, but with slower switching speed. IGBTs are preferred for high-voltage (>600V), high-current applications where conduction losses dominate, while MOSFETs excel in lower-voltage, high-frequency applications where switching losses are critical.

Q2: What is the typical gate voltage required to turn on an IGBT?

A: Modern IGBTs typically have threshold voltages (V_GE(th)) between 3-6V. For reliable operation, gate voltages of 15V are commonly used, with some devices rated up to 20V. Negative bias of -5V to -15V is often applied during turn-off to ensure fast switching and prevent false triggering from noise.

Q3: What causes tail current in IGBTs and how can it be minimized?

A: Tail current results from stored minority carriers (holes) in the drift region that must recombine before the device fully turns off. It can be minimized through: lifetime control techniques during manufacturing, negative gate bias during turn-off, optimized gate drive circuits, proper thermal management, and selection of IGBTs with field-stop or other advanced structures designed for reduced tail current.

Q4: What is latch-up in IGBTs and how can it be prevented?

A: Latch-up occurs when the parasitic thyristor structure within the IGBT is triggered, causing loss of gate control. It can be prevented by: staying within the device's safe operating area (SOA), using adequate gate drive with negative turn-off bias, minimizing circuit inductance to reduce voltage spikes, implementing proper snubber circuits, ensuring good thermal management, and selecting IGBTs with high latch-up immunity designed with optimized cell structures.

Q5: How do I choose between PT (Punch-Through) and NPT (Non-Punch-Through) IGBTs?

A: PT IGBTs offer lower on-state voltage and faster switching but limited reverse blocking capability, making them suitable for applications requiring unidirectional blocking with antiparallel diodes. NPT IGBTs provide bidirectional blocking capability and better short-circuit withstand time but higher on-state voltage. Field-stop (FS) IGBTs combine advantages of both and are now the most common choice for general applications. Selection depends on specific application requirements including voltage rating, switching frequency, efficiency targets, and cost constraints.

Q6: What are the advantages of IGBT modules over discrete IGBTs?

A: IGBT modules offer several advantages: integrated freewheeling diodes optimized for the IGBT characteristics, reduced parasitic inductance through optimized internal layout, improved thermal management with integrated heat sinks or cooling interfaces, simplified PCB design and assembly, better reliability through tested and qualified configurations, and often integrated temperature sensors and protection features. Intelligent Power Modules (IPMs) additionally include gate drivers and protection circuits.

Q7: How does temperature affect IGBT performance?

A: Temperature significantly impacts IGBT characteristics: on-state voltage typically has a positive temperature coefficient (increases with temperature), which aids in current sharing in parallel configurations; switching losses generally increase with temperature; threshold voltage decreases with temperature; tail current increases with temperature; and latch-up immunity decreases at higher temperatures. Proper thermal management is critical, with most IGBTs rated for maximum junction temperatures of 150-175°C, and some advanced devices rated to 200°C.

Q8: When should I consider using SiC MOSFETs instead of IGBTs?

A: Consider SiC MOSFETs when: switching frequencies above 50-100 kHz are required, maximum efficiency is critical, high-temperature operation (>175°C) is needed, or system size and weight reduction justifies higher component cost. IGBTs remain advantageous when: operating at lower switching frequencies (typically<20-30 cost="" is="" a="" primary="" very="" high="" current="" ratings="" are="" required="">1000A), or the application is well-established with proven IGBT solutions. Many applications in the 600-1700V range are transitioning to SiC, while IGBTs maintain strong positions in ultra-high-voltage (>3.3kV) and cost-sensitive applications.

Related Articles:

How many Transistors in a CPU?

What are the Testing Method and Types of Transistors?

An Overview of Bipolar Transistors

How Do Transistors Work?

Article Update Information:
       • Last Updated: October 31, 2025
   • Original Publication Date: 2020
   • Updates Include: Corrected technical terminology and formulas, updated voltage and current ratings to reflect 2025 technology standards, added information on modern IGBT generations and packaging technologies, included comparison with SiC and GaN alternatives, expanded applications section with current market trends, added comprehensive FAQ section addressing common technical questions, corrected spelling and grammatical errors, and updated all technical specifications to reflect current industry standards.