MOSFET vs IGBT: Structure, Switching Behavior, and How to Choose

Choosing between a power MOSFET and an IGBT is one of the recurring decisions in power-electronics design, and the right answer almost always depends on the voltage, current, and switching frequency the converter has to handle. Both are MOS-gated switches that an engineer drives with a voltage rather than a continuous base current, so they feel similar from the gate driver's point of view. Underneath that shared front end, however, they conduct and switch through different physics, and those differences decide which one keeps a design efficient, cool, and reliable. This guide walks through the structure of each device, explains why their conduction and switching behavior diverge, and turns that physics into a practical selection method.

• MOSFET vs IGBT at a Glance

• Inside the Device

• Conduction Characteristics

• Switching Behavior

• Where Each Device Wins

• How to Choose

• The Wide-Bandgap Shift

• Frequently Asked Questions

• Sources and References

MOSFET vs IGBT at a Glance

A power MOSFET is a fast, majority-carrier switch. Current flows through a conductive channel and drift region using only electrons (in an n-channel device), with no minority-carrier storage to slow it down, so it turns on and off quickly and tolerates high switching frequencies. Its on-state behaves like a resistor, which keeps losses low at lower voltage ratings but causes the conduction drop to climb steeply as the voltage rating rises.

An IGBT is a MOS-gated bipolar device. It keeps the easy, high-impedance gate of a MOSFET but adds a bipolar output stage, so at high voltage and high current it holds a more contained on-state voltage than a comparable silicon MOSFET. The tradeoff is speed: stored charge in the bipolar section produces a turn-off tail that adds switching loss and limits how fast the device can run. The table below summarizes the core contrast between the two devices.

Inside the Device

The behavioral differences come straight from how each device is built, so it helps to look at the silicon before reasoning about losses.

MOSFETs and IGBTs share a MOS-gated input, but their internal current paths and carrier behavior are different.

Vertical power MOSFET structure and the intrinsic body diode

Most power MOSFETs use a vertical structure: the source and gate sit on the top surface, the drain is the bottom of the die, and current flows vertically through a lightly doped drift region that supports the blocking voltage. Applying a positive gate-to-source voltage inverts a thin channel under the insulated gate oxide, connecting source to drift region so electrons carry the current. Because conduction relies only on these majority carriers, there is no stored minority charge to remove at turn-off.

That drift region is also where the voltage rating and the on-resistance fight each other. To block more voltage, the drift region must be longer and more lightly doped, and both changes raise resistance. This is why a high-voltage silicon MOSFET tends to have a much larger RDS(on) than a low-voltage part of similar die size. Power MOSFETs also include an intrinsic body diode formed by the body-to-drain junction, which conducts in the reverse direction and is frequently used as the freewheeling path in half-bridge and inverter legs.

How the IGBT's p+ injection layer enables conductivity modulation

An IGBT starts from a similar MOS front end but adds a p+ layer on the collector side of the die, beneath the drift region. When the device conducts, this layer injects holes into the drift region. The presence of both electrons and injected holes floods the drift region with extra carriers, a mechanism called conductivity modulation that effectively lowers the resistance of that region while the device is on.

That injected charge is the source of the IGBT's strength and its main limitation. Conductivity modulation lets a high-voltage IGBT hold a lower on-state drop than a high-voltage MOSFET would, but the same stored charge has to be cleared when the device turns off, which produces the tail current discussed later. The output therefore behaves like a bipolar stage controlled by a MOS gate, combining easy drive with a low high-voltage conduction drop.

Conduction Characteristics

In conduction the two devices follow different curves. The MOSFET drop is roughly ohmic: it rises in proportion to current, set by RDS(on), so at light load the voltage across the device is very small. The IGBT instead shows a junction-like offset, the saturation voltage VCE(sat), so there is a knee voltage that must be overcome before the device conducts heavily, after which the drop rises only gradually with current.

This shapes where each device is efficient. At lower voltage and moderate current, the MOSFET's small resistive drop usually wins, because there is no fixed offset voltage to pay. As the required blocking voltage climbs, the MOSFET's drift-region resistance grows steeply and its conduction loss can become large, while the IGBT's conductivity modulation keeps its high-voltage, high-current drop more contained. There is a crossover region where the IGBT becomes the more efficient conductor, but exactly where that happens depends on the specific parts, the current level, and the temperature. Rather than trusting a single rule-of-thumb number, take the actual RDS(on) and VCE(sat) curves from the manufacturer documents, such as the Bourns conduction-loss white paper and the device datasheets, and compare them at your operating point.

Switching Behavior

Switching is where the MOSFET's majority-carrier nature pays off. With no stored minority charge to remove, it turns off cleanly and quickly, so its switching energy per transition is low and it can run at high frequency without overheating. That makes it the natural choice when a converter relies on a high switching frequency to shrink magnetics and filters.

The IGBT is slower to turn off because the holes injected during conduction do not disappear instantly. As the channel closes, that remaining charge continues to flow for a short time, producing a turn-off tail current that overlaps with rising voltage and dissipates extra energy each cycle. The practical consequences are concrete: the designer must allow enough dead time to let the tail decay, the switching loss grows with frequency, and there is an effective frequency ceiling beyond which the IGBT runs too hot to be useful. This is the heart of the MOSFET-versus-IGBT decision. The MOSFET trades a higher high-voltage conduction loss for low switching loss; the IGBT trades a low high-voltage conduction loss for higher switching loss. For the actual loss equations, including how the tail-current term is modeled, the Renesas IGBT loss-calculation note and general power-switch loss literature from Texas Instruments lay out the method.

The MOSFET switches quickly because it has no stored minority charge, while the IGBT’s tail current increases turn-off loss.

Where Each Device Wins

Because conduction loss and switching loss pull in opposite directions, the winning device tracks the operating envelope. The table below maps common application classes to the device that is typically preferred and the reason behind it.

Typical MOSFET applications

MOSFETs dominate lower-voltage, higher-frequency conversion. Point-of-load regulators, DC-DC converters, switch-mode supplies, battery management, and consumer power stages all benefit from fast switching and a small resistive on-state at these voltage ratings. In these designs the frequency is high enough that switching loss would punish a slower device, and the voltages are low enough that the MOSFET's resistance stays modest.

Typical IGBT applications

IGBTs are the workhorse of higher-voltage, higher-current, lower-frequency power conversion. Motor drives, industrial and grid inverters, electric-vehicle traction, induction heating, and welding equipment all push voltage and current into a range where the IGBT's contained conduction drop matters more than raw switching speed. These systems usually run at frequencies the IGBT can tolerate, so the tail-current penalty is an acceptable price for efficient high-power conduction.

How to Choose

A workable decision follows the requirements in order. Start with the system voltage, because it eliminates the most options first. If the bus voltage is low, a MOSFET is usually the default. If it is high, the comparison tilts toward the IGBT unless a high-voltage wide-bandgap MOSFET is on the table.

Next, weigh the switching frequency against the current. A high target frequency rewards the MOSFET's low switching loss, while high current at a moderate frequency rewards the IGBT's low conduction drop. When voltage, current, and frequency point in conflicting directions, that is exactly the crossover region where you should compare candidate parts at your operating point rather than rely on a single threshold; the device choice is set by the interaction of these three variables, not by any one of them alone.

Then check the thermal budget. Estimate conduction and switching loss at the worst-case operating point using the manufacturer loss documents, confirm the package and heatsink can remove that heat, and revisit the frequency if the numbers are tight. Finally, plan the freewheeling path. A power MOSFET usually offers its intrinsic body diode for inductive freewheeling, while a standard IGBT has no inherent body diode and may need a co-packaged or external anti-parallel diode for inductive loads, so confirm on the datasheet whether the part includes one before you commit to a topology.

The best device choice depends on the combined voltage, current, switching frequency, and thermal requirements of the design.

The Wide-Bandgap Shift

Silicon is no longer the only option for these roles. Wide-bandgap devices, namely silicon-carbide (SiC) MOSFETs and gallium-nitride (GaN) transistors, are increasingly used where designs demand higher efficiency, higher switching frequency, or higher operating voltage than silicon comfortably delivers. SiC MOSFETs in particular extend fast, majority-carrier switching into voltage ranges that were once IGBT territory, which can reduce switching loss compared with a silicon IGBT in suitable applications.

The useful way to reason about this is as a gradual widening of where each technology fits, not as one part swapping in for another. Wide-bandgap devices add capability at the high-voltage, high-frequency end and reshape the boundaries of the silicon MOSFET and IGBT domains, while silicon parts remain widely used where their cost and maturity are a good match. Whether a wide-bandgap device improves a specific design depends on the operating point, the system cost target, and the loss comparison for the actual parts, which is laid out in the Toshiba SiC-versus-IGBT loss material.

Frequently Asked Questions

What is the main difference between a MOSFET and an IGBT?
A MOSFET conducts using only majority carriers, so it switches fast and behaves resistively, while an IGBT adds a bipolar output stage that gives a lower on-state drop at high voltage at the cost of slower turn-off.

When should I use an IGBT instead of a MOSFET?
Reach for an IGBT when the design runs at high voltage and high current with a switching frequency the device can tolerate, such as a motor drive or an inverter, where its contained conduction drop outweighs its switching loss.

Why does a MOSFET handle higher switching frequencies better?
Because it stores no minority charge, a MOSFET turns off quickly with low switching energy per cycle, so it dissipates less per transition and stays cooler at high frequency.

What is IGBT tail current and why does it matter?
Tail current is the residual current that keeps flowing at turn-off as injected charge in the drift region clears. It overlaps with rising voltage, adds switching loss, forces longer dead time, and sets a practical frequency ceiling.

Is a MOSFET or an IGBT better for an inverter or a welding machine?
Both applications run at high power with manageable frequency, so an IGBT is usually the better fit because its low high-voltage conduction loss matters more there than fast switching.

Does an IGBT have a body diode like a MOSFET?
A standard IGBT does not have an inherent body diode the way a power MOSFET does, so many IGBTs are co-packaged with an anti-parallel freewheeling diode; confirm this on the datasheet before using it in an inductive circuit.

How do SiC MOSFETs affect the MOSFET vs IGBT decision?
SiC MOSFETs bring fast switching into higher-voltage ranges, so in high-efficiency, high-frequency designs they can reduce switching loss relative to a silicon IGBT, widening where a MOSFET-type switch makes sense.

Sources and References

Fuji Electric's explainer on the difference between an IGBT and a MOSFET (americas.fujielectric.com) lays out the structural distinction between a majority-carrier MOSFET and a MOS-gated bipolar IGBT and the typical application split; it is an introductory overview, so the exact device behavior still depends on the specific part you choose.

The Bourns white paper on measuring IGBT conduction loss (bourns.com) explains how saturation voltage governs IGBT conduction loss, which supports the resistive-versus-junction contrast here; the precise VCE(sat) curves it discusses are tied to the particular devices measured.

The Bourns article on choosing between an IGBT and a MOSFET (bourns.com) frames the selection logic of frequency and voltage that this guide follows, though the crossover point it describes shifts with each design's operating conditions.

The Renesas IGBT loss-calculation application note (renesas.com) shows how conduction and switching losses, including the turn-off tail term, are computed; the figures it produces are only as accurate as the device parameters you enter for your part.

Texas Instruments power-switch loss literature (ti.com) describes the general conduction-versus-switching loss mechanisms used to frame the tradeoff, and the specific numbers depend on the topology and components in your converter.

The Toshiba IGBT FAQ (toshiba.semicon-storage.com) confirms that IGBTs suit relatively low switching frequency at high output power while MOSFETs handle higher frequency; it speaks in general terms, so a given part may sit differently on that scale.

The Toshiba comparison of loss between a SiC MOSFET and a Si IGBT (toshiba.semicon-storage.com) supports the wide-bandgap trend framing, with the caveat that any efficiency advantage is specific to the devices and operating points compared.

The University of Tennessee EECS power-device lecture notes (web.eecs.utk.edu) provide textbook grounding for the device physics described here; as educational material, it is a conceptual reference rather than a substitute for a datasheet.