Common Problems of IGBT
Continuously get the latest content about IGBT
What is an IGBT?
IGBT stands for insulated-gate bipolar transistor. Figure (a) shows the symbol of an IGBT. It is a power transistor that combines an input MOS and an output bipolar transistor. Figure (b) shows an example of the IGBT structure. A P region is formed on the drain side of the MOSFET. The resistivity of the high-resistance N- drift region decreases when holes are injected from this P region at turn-on. This phenomenon is called conductivity modulation. Consequently, an IGBT is a switching transistor with low ON voltage even at high breakdown voltage.
Although its internal equivalent circuit is complicated, it can be simplified as consisting of an N-channel MOSFET with variable on-resistance and a diode connected in series as shown in Figure (c).
The IGBT is a transistor ideal for high-voltage, high-current applications. Available with a voltage rating ranging from 400 V to 2000 V and a current rating ranging from 5 A to 1000 A(*1), the IGBT is widely used for industrial applications such as inverter systems and uninterruptible power supplies (UPS), consumer applications such as air conditioners and induction cookers, and automotive applications such as electric vehicle (EV) motor controllers.
(*1) IGBTs with up to 6 kV and up to 4500 A are also available for railway, high-voltage direct-current (HVDC) transmission, and other large applications.

(a) Symbol of an N-channel IGBT
(b) Typical IGBT structure
What is the difference between MOSFETs and IGBTs?
There are three major types of transistors: bipolar transistors, MOSFETs, and IGBTs. The following table compares the performance and characteristics of these transistors. Bipolar transistors are now hardly ever used for power electronics and switching applications because of the need for drive and protection circuits and slow switching speed. Instead, MOSFETs and IGBTs are selectively used according to the required characteristics. The figure given alongside shows the on-state voltage characteristics of a 30-A IGBT and a 31-A super-junction MOSFET (SJMOS).
In the low-current region, the MOSFET exhibits a lower on-state voltage than the IGBT. However, in the high-current region, the IGBT exhibits lower on-state voltage than the MOSFET, particularly at high temperature. IGBTs are commonly used at a switching frequency lower than 20 kHz because they exhibit higher switching loss than unipolar MOSFETs.
Forward characteristic comparison : IGBT vs. MOSFET
Comparison of the performance of different types of transistors
Type |
Bipolar transistors |
MOSFETs |
IGBTs |
Gate (base) drive |
Current drive (Low input impedance) |
Voltage drive (High input impedance) |
Voltage drive (High input impedance) |
Gate (base) drive circuit |
Complicated for switching applications |
Relatively simple |
Relatively simple |
On-state voltage characteristics |
Low VCE(sat) |
On-resistance x drain current Without built-in voltage(*1) |
Low VCE(sat) With built-in voltage(*1) |
Switching time |
Slow (Carrier accumulation effect) |
Ultra-high speed (Unipolar device) |
High speed (Faster than bipolar transistors and slower than MOSFETs) |
Parasitic diode |
Not present |
Present (body diode) |
Present only in RC-IGBTs |
(*1) The built-in voltage is a threshold voltage inherent to a device. Here, the built-in voltage refers to the forward threshold voltage.
For what applications should MOSFETs and IGBTs be used?
Because IGBTs are bipolar switching devices that use conductivity modulation, they exhibit slower switching speed, particularly longer turn-off time at high temperature, than unipolar MOSFETs. Therefore, IGBTs cause higher switching losses.In contrast, IGBTs have advantages in that they can easily achieve high withstand voltage and have relatively low on-state voltage even at high current and high temperature.
Therefore, IGBTs and MOSFETs fabricated using silicon material have the following application areas:
1. MOSFETs: Low-voltage applications (below 200 to 300 V)
2. IGBTs: High-voltage applications (above 1200 V)
3. IGBTs and MOSFETs are used for different purposes for 400- to 1200-V applications:
(1) IGBTs are used for inverter applications with a switching frequency of less than 20 kHz requiring high overload endurance.
(2) MOSFETs are used for inverter applications with a switching frequency exceeding 20 kHz.
(3) MOSFETs are used for some low-capacity inverter applications whereas IGBTs are used for soft-switching and high-current-density applications. IGBTs and MOSFETs should be used properly according to their characteristics.
Comparison of different types of transistors
What is the principle of operation of the IGBT?
The equivalent circuit of an IGBT is shown below. When both the gate-emitter (G-E) and collector-emitter (C-E) paths are positively biased, the N-channel MOSFET conducts, causing drain current to flow. This drain current also flows to the base of QPNP and causes the IGBT to turn on. Since the DC current gain (α) of QPNP is very small, almost the entire emitter current (IE(pnp)) flows as base current (IB(pnp)). However, part of IE(pnp) flows as collector current (IC(pnp)). The IC(pnp) does not turn on QNPN because it bypasses the RBE inserted between the base and emitter of the QNPN.
Therefore, almost the entire collector current of the IGBT flows as the drain current of the N-channel MOSFET via the emitter-base paths of QPNP. At this time, holes are injected into the high-resistance drift region of the N-channel MOSFET from the emitter of QPNP. This causes the resistivity of the drift region (Rd(MOS)) to decrease considerably, reducing the on-resistance during a conduction period. This phenomenon is called conductivity modulation.
Turning off the gate (G) signal causes the N-channel MOSFET to turn off and therefore causes the IGBT to turn off.
Equivalent circuit of an IGBT
In what structures are IGBTs available?
The common structures of IGBTs include: (a) punch-through (PT), and (b) non-punch-through (NPT), and (c) thin-wafer punch-through (thin-wafer PT), which is also called field-stop (FS).
(d) Reverse-conducting IGBTs (RC-IGBTs) are a recent addition to IGBT variations in which part of the collector P region of the FS IGBT is replaced by an N region and a freewheeling diode is integrated like a MOSFET. The following table shows generations of IGBTs and their structures.
● PT IGBTs
● The PT structure has been used since the inception of IGBTs. The P layer on the collector side is thick, and the forward voltage in the low-current region is high.NPT IGBTs
● NPT IGBTs appeared, following PT IGBTs. NPN IGBTs have high ruggedness and are used for hard switching and other inverter applications.Thin-PT IGBTs
● Thin-PT is one of the latest IGBT structures that uses thin-wafer technology to improve trade-offs between forward voltage drop and switching speed. Because of low loss, thin-PT IGBTs are widely used.RC-IGBTs
RC-IGBTs use the latest thin-wafer technology and incorporate a fast-recovery diode (FRD). RC-IGBTs are available for voltage resonance and other applications.
(a) PT IGBT
(b) NPT IGBT (NPT : Non-Punch-Through)
(c) Thin-PT IGBT (FS-IGBT)
(d) RC-IGBT
What is a reverse-conducting IGBT (RC-IGBT)?
A reverse-conducting IGBT (RC-IGBT) integrates an IGBT and a freewheeling diode (FWD) on a single chip. In many IGBT applications, there is a mode in which freewheeling current flows from the emitter to the collector. For this freewheeling operation, the freewheeling diode is connected anti-parallel to the IGBT. Figure (b) shows an example of the internal structure of an RC-IGBT. Part of the P region in the collector electrode is replaced by an N region to form a PIN diode(*1) (P-N--P) with the P region in the emitter electrode. This PIN diode is connected anti-parallel to the IGBT like the FWD shown in Figure (a) and acts as a freewheeling diode. Nowadays, the applications of RC-IGBTs are expanding to the hard-switching fields in addition to voltage resonance.
(*1) A PIN diode is a diode with a high-resistance intrinsic (I) semiconductor region between P and N regions. Because of low dopant concentration, the intrinsic region has high resistance close to that of an intrinsic semiconductor. The FWD in an IGBT requiring high withstand voltage has this structure.
(a) IGBT+FWD
(b) Example of the internal structure of an RC-IGBT
What is conductivity modulation?
While IGBTs and other bipolar devices(*1) are on, carrier injection into the high-resistance drift region causes its resistivity to decrease. This is called conductivity modulation.
The N- drift region in IGBTs and other high-voltage switching devices is thick and has low dopant concentration. Therefore, the N- drift region has extremely high resistivity. The IGBT turns on when the gate-emitter and collector-emitter paths are positively biased as shown below. At this time, holes are injected into the N- region from the collector P region via the N region. Consequently, the carrier concentration in the high-resistance N- region increases, causing its resistivity to decrease. As a result, its forward voltage drop decreases, causing the IGBT to act as a switching device with low on-state voltage.
This increase in conductivity (i.e., a reduction in resistivity) during a conduction period is called conductivity modulation.
(*1) A bipolar device is a type of semiconductor device that uses both electrons and holes as charge carriers for current conduction. In contrast, unipolar devices such as MOSFETs use only one type of charge carrier.
(a) Symbol of an N-channel IGBT
What is a safe operating area?
The safe operating area (SOA) is defined as the current and voltage conditions over which an IGBT can be expected to operate without self-damage or degradation. In practice, it is necessary not only to use an IGBT within the safe operating area but also to derate its area for temperature. There are forward-bias and reverse-bias safe operating areas (FBSOA and RBSOA). The forward-bias safe operating area defines the usable current and voltage conditions for the period of time while the IGBT is on. The reverse-bias safe operating area defines the usable current and voltage conditions during the turn-off period of the IGBT.
Figure (a) shows an example of a forward-bias safe operating area, which consists of four regions: (1) a region limited by the maximum collector current rating, (2) a region limited by collector power dissipation (thermal breakdown), (3) a region limited by secondary breakdown, and (4) a region limited by the maximum collector-emitter voltage rating. Care should be exercised as to the region limited by secondary breakdown because it differs, depending on the device design.
Figure (b) shows an example of a reverse-bias safe operating area, which consists of three regions: (1) a region limited by the maximum collector current rating (ICP), (2) a region limited by the inherent characteristics of a device, and (3) a region limited by the maximum collector-emitter voltage rating (VCES). In the second region, the maximum current is in inverse proportion to VCE. Hard-switching applications, in particular, should be designed so as to satisfy the limit of this SOA region.
(a) Forward-bias SOA
(b) Reverse-bias SOA
What is the definition of IGBT power dissipation?
The power dissipation of an IGBT is specified as collector power dissipation (PC) in its datasheet.
Collector power dissipation (PC) is defined as the maximum permissible power dissipation that the IGBT can consume continuously and expressed as:
Collector power dissipation (PC) = permissible_rise_in_temperature (Tj(max) – 25°C) / thermal_resistance (Rth)
Collector power dissipation and thermal resistance are shown in the datasheet. The conditions under which they are guaranteed are specified for individual IGBTs. Heat dissipation conditions are specified for collector power dissipation whereas two positions with a temperature difference are specified for thermal resistance.
In Table (a), a PC of 2 W is specified under free-air cooling conditions without any thermal fin whereas a PC of 40 W is specified under cooling conditions with infinite thermal conductivity. In Table (b), an Rth(j-a) of 62.5°C/W is junction-to-ambient thermal resistance under free-air cooling conditions whereas Rth(j-c) is junction-to-case thermal resistance. To calculate thermal resistance under actual usage conditions, it is necessary to take insulation materials, thermal fins, and the contact thermal resistance of each component into consideration. The thermal resistance thus calculated should be used to calculate collector power dissipation using the equation shown above.
Absolute Maximum Ratings (Note) (Ta=25℃, unless otherwise specified)
Table (a) Examples of absolute maximum ratings and collector power dissipation (PC) of an IGBT
Thermal Characteristics
Table (b) Example of thermal resistances of an IGBT
What is the tail current of an IGBT?
The IGBT is a type of power transistor that operates in bipolar mode because of the P layer formed on the drain side of a MOSFET. The IGBT uses a phenomenon called conductivity modulation that exhibits a reduction in the resistivity of the high-resistance N- drift region at turn-on when holes are injected from this P region.
On-state voltage can be reduced because of conductivity modulation, but the IGBT needs to remove minority carriers from the N- drift region when it turns off.
When the IGBT begins to turn off, minority carriers are swept out to external circuitry. When the collector-emitter voltage (VCE) of the IGBT has risen to a certain level (i.e., after the depletion region has expanded), minority carriers contribute to internal recombination current. This current is called tail current. Because tail current is the collector current with a high VCE voltage being applied, it is one of the significant contributors to switching loss.
To reduce tail current and thereby switching loss, IGBTs are principally designed so as to reduce 1) the lifetime of minority carriers and 2) the amount of holes injected from the collector. However, both these techniques cause a rise in on-state voltage. Therefore, IGBTs are designed with optimum trade-offs among these characteristics according to their intended applications.
Typical collector current and collector-emitter voltage waveforms during turn-off
Please give some application examples for IGBTs.
Because of their high-voltage and high-current characteristics, IGBTs are used as switching devices for motor drive systems, uninterruptible power supplies (UPS), induction cookers, and other applications.
Figure (a) shows an example of a motor drive circuit. IGBTs are widely used as switching devices in the inverter circuit (for DC-to-AC conversion) for driving small to large motors. IGBTs for inverter applications are used in home appliances such as air conditioners and refrigerators, industrial motors, and automotive main motor controllers to improve their efficiency.
Figure (b) shows an example of a UPS circuit. IGBTs are in middle- and large-capacity (several kVA or higher-capacity) models, contributing to high efficiency and space saving.
Figure (c) shows an example of an induction-heating (IH) circuit. Induction heating uses LC resonance for zero-voltage switching (ZVS) or zero-current switching (ZCS) to reduce switching loss. Because of high resonance voltage or resonance current, IGBTs are commonly used. Specifically, IGBT applications include induction cookers, induction rice cookers and microwave ovens.
(a) Motor drive circuit
(b) UPS circuit
(c) Induction cooktop
Please explain hard switching and soft switching using IGBTs.
The terms “hard switching” and “soft switching” refer to the methods of switching based on the relationship of current and voltage during the turn-on and turn-off of the IGBT. Hard switching is a switching method that simply uses a device’s own ability.
Figure (a) shows a typical hard-switching current, voltage waveforms and its operating locus. During on-off transitions, both voltage and current are applied to the device. With hard switching, collector current and collector-emitter voltage change sharply, causing switching noise and loss. Hard switching is used for simple switch, motor drive inverter, and switched-mode power supply applications.
In contrast, soft switching uses an LC resonant circuit to turn on and off a device at zero current or voltage. Or the voltage and current switching timing is controlled in order to minimize the intersection of their waveforms. Figure (b) shows typical current and voltage waveforms of a soft-switched device and its operating locus. Soft switching helps reduce the switching noise and loss because switching devices turn on and off at zero or nearly zero voltage or current. Soft switching is commonly used for induction rice cookers, induction cooktops, and microwave ovens.
Soft switching has an added advantage over hard switching in terms of a safe operating area (SOA) as shown below.
(a) Hard switching
(b) Soft switching
Please explain the operation of voltage-resonant soft switching of an IGBT.
Figure (a) shows the schematic of a voltage-resonant induction cooktop as an application example of soft switching. Figure (b) shows its operation and waveforms.In the circuit of Figure (a), when the IGBT turns on, current flows through the heating coil (L1). When the IGBT turns off, L1 and C1 go into resonance, causing sinusoidal voltage to be applied to the IGBT. The direction of resonance between L1 and C1 reverses, causing the C1 voltage to offset the C2 voltage. When the C1 voltage exceeds the C2 voltage, current begins to flow through the C1-C2-FWD-C1 loop. During this period, the collector-emitter voltage of the IGBT is equal to the forward voltage (VF) of the freewheeling diode (FWD), which is almost zero. At this time, the IGBT turns back on. As a result, current flows through the heating coil (L1) from the input side again. This sequence is repeated.
A voltage-resonant circuit is inexpensive because it does not require many components. However, when a system needs a high power capacity, an IGBT with very high withstand voltage is required so as to handle high resonance voltage. Therefore, a voltage-resonant circuit is used in many induction home appliances with a capacity of up to 1.5 kW at 100 VAC and up to 3 kW at 200 VAC. The smoothing capacitor (C2) on the input side has low capacitance because it receives electric power during a single pulse period. The voltage across C2 has a full-sine waveform, leading to a high power factor on the input side. Therefore, a voltage-resonant circuit eliminates the need for a power factor correction (PFC) circuit.
(a) Example of a voltage-resonant circuit
(b) Example of operating waveforms
1. The IGBT turns on.
2. Collector current flows via L1. The IGBT turns off.
3. L1 and C1 go into resonance, causing voltage to increase.The FWD conducts.
The energy stored in C1 flows back to the power supply via the FWD. An “on” signal is applied to the IGBT when the resonance voltage drops below the reference voltage.Steps 1 to 3 are repeated.
Please explain the operation of current-resonant soft switching of the IGBT.
Figure (a) shows the schematic of a current-resonant induction cooktop as an application example of soft switching. Figure (b) shows its operation and waveforms. The heating coil (L1) and the capacitor (C1) go into resonance, causing load current to be sinusoidal. (The voltage applied to the IGBT has a square waveform.) IGBT1 turns on, causing the current generated by the resonance between L1 and C1 to flow. When IGBT1 turns off thereafter, resonance current flows through the closed L1ーC1ーFWD2 loop. When resonance current reverses its direction, IGBT2 turns on, causing resonance current to flow through the closed C1ーL1ーIGBT2 loop. When IGBT2 turns off, resonance current freewheels through the L1ーFWD2ーC2ーC1 loop. When it drops to zero, IGBT1 turns on again. This sequence is repeated.
A current-resonant circuit requires more components than a voltage-resonant circuit. The current-resonant circuit is used for high-power stationary cookers and aluminum pan (all-metal) cookers requiring high-frequency switching. A DC power supply from a PFC circuit is desirable for stable operation. In practice, however, current resonance provides a high power factor without PFC. C2 can be a small capacitor having enough capacitance to provide a stable switching operation for one cycle. The AC power supply is designed to provide a sinusoidal load current.
(a) Example of a current-resonant circuit
Symbol and measuring point
Symbol |
Measuring point |
VGE1 |
IGBT1 Gate - Emitter Voltage |
VCE1 |
IGBT1 Gate - Emitter Voltage |
VGE2 |
IGBT2 Gate - Emitter Voltage |
VCE2 |
IGBT2 Gate - Emitter Voltage |
(b) Example of operating waveforms
1. IGBT1 is on.
2. Resonance current flows through C2ーIGBT1ーL1ーC1ーC2.IGBT1 turns off.
3. Resonance current flows through the closed L1ーC1ーFWD2ーL1 loop.IGBT2 turns on.
4. When the current flowing through L1 reverses its direction, IGBT2 turns on, causing current to flow through the closed C1ーL1ーIGBT2ーC1 loop.IGBT2 turns off.
Current flows through the closed L1ーFWD1ーC2ーC1ーL1 loop.Steps 1 to 4 are repeated.
How can I provide protection against the surge voltage generated by the turn-off of an IGBT?
Figure (a) shows an example of an IGBT application circuit. Surge voltage is generated when the IGBT turns off while it is carrying the load current. Surge voltage is caused by a sharp change in IGBT current (-diC/dt) as well as the package and wire stray inductance (LS). At this time, VCEP = LS・diC/dt+VCC is applied to the IGBT instantaneously. The IGBT is permanently damaged if a voltage exceeding its breakdown voltage is applied. A primary solution for this is to reduce the stray inductance (LS) of the main current path. It is therefore necessary to increase the width and reduce the length of the wires. If it is difficult to reduce stray inductance, the switching speed of the IGBT should be reduced by increasing the value of an external gate resistor connected in series with the gate. Care should be exercised, however, because this causes its switching loss to increase. Surge voltage can also be reduced by inserting a snubber circuit between the collector and emitter terminals of the IGBT. However, the snubber circuit causes charge/discharge loss, increasing overall circuit loss.
(a) IGBT switching circuit
(b) Turn-off waveform
Is it possible to connect multiple IGBTs in parallel? If so, is there anything to note about parallel connection?
We recommend using an IGBT with higher ratings instead of connecting two IGBTs in parallel. In case of unavoidable parallel operation, please pay attention to the following points when designing.
1. In the high-current region, the collector-emitter saturation voltage (VCE(sat)) of the IGBT has positive temperature dependence. However, many IGBTs exhibit negative temperature dependence in the low-current region as well as in the region in which they are actually used. If these IGBTs are connected in parallel, a greater proportion of current flows to the one with lower VCE(sat). This proportion increases as temperature rises. To balance current among parallel IGBTs, they should have an equal VCE(sat).
2. If parallel-connected IGBTs are driven by a single external gate resistor (RG) connected in series with the gate, the IGBT gate-emitter voltage might oscillate, resulting in the oscillation of the collector voltage and current. To prevent such oscillation, it is necessary to add separate gate resistors (RG1 and RG2) to each IGBT and suppress the parasitic inductance of the closed-loop gate drive circuit.
3. The parasitic inductance of the main circuit might cause oscillation, surge voltage, current imbalance, and other abnormal conditions during switching. A primary cause of these problems is parasitic inductances (LS*) as shown below. The board trace layout must be designed in such a manner as to reduce these parasitic inductances to as close to zero as possible. For parallel operation, the board traces from the main circuit must be designed in such a manner as to reduce LS1 and LS2, in particular, to zero.
Gate resistor connections and parasitic inductances in the case of parallel-connected IGBTs
At what voltage should the IGBT gate be driven?
Consider driving the IGBT gate at a voltage equal to the VGE value shown in the datasheet as a test condition for the VCE(sat) and switching characteristics or at a voltage close to VGE (see the following tables). Increasing the gate-emitter voltage (VGE) reduces a margin with respect to the absolute maximum rating whereas reducing the gate-emitter voltage causes VCE(sat) to increase as shown below, increasing conduction loss. If the gate-emitter voltage is too low, a system might not operate satisfactorily because the IGBT is not driven sufficiently. Generally, we recommend a VGE level equal or close to 15 V except for special-purpose IGBTs such as those for strobe light applications.
Also, when there is a difference between the gate bias power supply voltage and the gate-emitter voltage of the IGBT, a design should be created such that the gate-emitter voltage of the IGBT becomes the above-mentioned value.
(Note: Generally, the gate-emitter voltage of an IGBT is set to zero when it is off. In some cases, however, the gate-emitter voltage is reverse-biased to stabilize the IGBT switching operation.)
Static Characteristics(Ta=25℃, unless otherwise specified)
Dynamic Characteristics(Ta=25℃, unless otherwise specified)
Table Electrical characteristics of an IGBT
Figure Example of IC-VCE curves
Technical documents for MOSFETs and bipolar transistors contain a safe operating area (SOA) graph. What is it?
The area where transistors and MOSFETs are not destroyed or prone to deterioration is called the safe operating area (SOA). It is important that the operating point is within this curve. In the case of a bipolar transistor, the curve in the SOA is divided into a thermal dissipation limit and the limit given by secondary breakdown. In the case of a MOSFET, the SOA is divided into a current limit, a thermal dissipation limit, a limit given by secondary breakdown, and a voltage limit. Recently, there are products for which an on-resistance limit area is specified in the data sheet.
Maximum Ratings: Power MOSFET Application Notes (PDF:1,075KB)
Are there any special considerations for thermal calculation?
Each device has different thermal conductivity paths, but basically, you should calculate the thermal resistance from a semiconductor chip in a package to ambient air. Use this information when considering thermal dissipation for a system design. Power devices, in particular, have considerable loss, which significantly affects the system reliability and life time. Extreme care should be exercised when considering thermal design. Allow sufficient margins relative to the maximum allowable junction temperature for each device (120°C, 150°C, etc).
Thermal Design and Attachment of a Thermal Fin: Power MOSFET Application Notes (PDF:1,061KB)
Neither Rth(ch-a) nor Rth(j-a) is specified for MOSFETs, IGBTs and bipolar transistors. Why is that?

self-standing power devices
Since it is common to install heat sinks in self-standing power devices, we are offering Rth(ch-c) and Rth(j-c). (There are exceptions in some products.)
Channel-to-case (or junction-to-case) thermal resistance values should be used for more accurate thermal calculations instead of channel-to-ambient (or junction-to-ambient) thermal resistance values. Be sure to use Rth(ch-c) or Rth(j-c) if it is specified.
Thermal Design and Attachment of a Thermal Fin: Power MOSFET Application Notes (PDF:1,061KB)
Are there any reasons why junction-to-case (or channel-to-case) thermal resistance is not specified for small-package devices?
Neither junction-to-case thermal resistance, Rth(j-c), nor channel-to-case thermal resistance, Rth(ch-c), is specified because small-package devices do not dissipate much heat and because temperatures on the surface of small packages cannot be measured accurately. For thermal design, use either junction-to-ambient thermal resistance, Rth(j-a), or channel-to-ambient thermal resistance, Rth(ch-a), instead.