IGBT vs MOSFET: When to Use Each – A Practical Selection Guide for Engineers

Expert guide on IGBT vs MOSFET: When to Use Each – A Practical Selection Guide for Engineers. Technical specs, applications, sourcing tips for engineers and buyers.

IGBT vs MOSFET: When to Use Each – A Practical Selection Guide for Engineers

Why the IGBT-MOSFET Trade-Off Hits Harder in 2025’s Supply Chain

You’re not imagining it—the old voltage‑versus‑frequency decision tree has gotten a lot more complex. Today’s power stage designers face a dual squeeze: the physical limits of silicon and a supply chain that can turn a routine BOM refresh into a production‑stopping crisis. Shunlongwei’s 2025 guide captures the classic voltage‑frequency‑power triangle, but the on‑the‑ground problem is bigger. When an IGBT’s turn‑off tail current meets a 30‑week lead time, the only thing that matters is whether you can ship product.

At the device level, LCSC details the hybrid IGBT structure—a MOSFET gate driving a bipolar output—that delivers conductivity modulation at the expense of a stubborn turn‑off delay. That tail current isn’t just a loss mechanism; it forces you to delay the complementary switch turn‑on, which eats into your dead time budget and can push you into a harder switching regime than the datasheet numbers suggest. Meanwhile, lead times for IGBTs and MOSFETs have stretched to 30 weeks or more, and several major IDMs are fully allocated on popular voltage classes. The result: a design decision that was once purely about conduction losses and switching frequency is now a sourcing strategy problem. Picking the wrong device can lead to overheating, oversized heatsinks, EMI problems, or early failure—and when you can’t get the right part, you’re forced to qualify a second source under pressure, a scenario where every unchecked parameter becomes a latent field risk.

The engineering community now talks about the IGBT‑MOSFET selection as a three‑dimensional optimization: performance, cost, and availability. You can’t afford to treat any axis as fixed. The rest of this guide breaks down the physics, the real‑world alternatives, and a practical checklist that gets you from datasheet to BOM without leaving procurement in the dark.

How Conductivity Modulation and Tail Current Shape Power Losses

To choose the right switch, you need to understand why the two devices behave so differently under load. A power MOSFET is a unipolar device; current flows through a resistive channel and drift region, and on‑resistance (Rds(on)) rises with blocking voltage—roughly proportional to VBR(DSS)2.5. That makes a 600 V silicon MOSFET painfully lossy at high current unless you parallel multiple die or move to wide‑bandgap alternatives. The IGBT takes a different path. It uses conductivity modulation: the bipolar output stage floods the drift region with minority carriers, collapsing the voltage drop to a near‑constant VCE(sat) that stays low even at high voltage ratings. The penalty is that those carriers must be swept out during turn‑off, creating the notorious tail current that persists until recombination completes.

This tail current translates directly into turn‑off switching loss, and it limits the practical switching frequency of IGBTs to roughly 20–50 kHz for hard‑switched applications, compared to hundreds of kilohertz for a MOSFET. EE Times notes the industry trend of IGBTs replacing MOSFETs except in very low current applications, precisely because the IGBT’s conduction advantage grows with voltage and current, while the switching penalty becomes tolerable at lower frequencies. The table below distills the fundamental differences you’ll see reflected in every datasheet.

ParameterPower MOSFETIGBT
Conduction mechanismUnipolar, resistive channelMOSFET input + bipolar output (conductivity modulation)
On‑state voltage dropRds(on) × ID; strongly temperature‑dependentVCE(sat) ≈ 1.5–2.5 V, with positive temperature coefficient at high current
Switching speedFast; turn‑on/off times in tens of nsSlower turn‑off due to tail current; typical tf 100–500 ns
Typical voltage range20–900 V (silicon); up to 1.7 kV (SiC)600–6.5 kV
Practical frequency limit (hard switching)100 kHz–1 MHz+10–50 kHz (silicon); up to 60 kHz for fast‑switching types
Turn‑off behaviorFast current fall, negligible tailFast current fall followed by exponential tail current
Primary loss component at high voltageConduction loss (I²R) dominatesConduction loss low; turn‑off switching loss dominates

These physical differences explain why IGBTs own the high‑voltage, high‑current, low‑frequency corner of the application space, while MOSFETs dominate fast‑switching, lower‑voltage designs. But the picture blurs once you introduce silicon carbide MOSFETs, which combine low on‑resistance with near‑zero tail current and can operate at frequencies that would smoke a silicon IGBT. That’s where the next section picks up.

Side-by-Side: When Silicon MOSFETs, SiC, and IGBTs Each Win

The real decision isn’t just MOSFET versus IGBT. It’s silicon MOSFET, silicon IGBT, or wide‑bandgap FET—and the answer depends on your exact operating point. IC Online’s sourcing guide frames this as a cost‑driven selection, noting that a SiC MOSFET may carry a 2–3× price premium over an IGBT, but the system‑level savings in magnetics and thermal management can flip the equation. Key considerations for power electronics design emphasize that you can’t just compare Rds(on) and VCE(sat) in isolation; you must weigh the total loss budget, switching frequency, and the cost of cooling. Cost‑optimization strategies further highlight that IGBTs often win in 3‑phase inverters and motor drives above 1 kW, while SiC is carving out territory in solar inverters and EV on‑board chargers where efficiency and power density are non‑negotiable.

The table below captures the practical trade‑offs across four metrics that directly impact your BOM and PCB layout. Values are representative for 650‑1200 V class parts at 25 °C, unless otherwise noted. Always check the datasheet at your worst‑case junction temperature.

Comparison MetricSilicon MOSFET (e.g., 650 V C3M™)SiC MOSFET (e.g., 1200 V)IGBT (e.g., 650 V Trench‑Field Stop)Selection Criteria & Failure Boundary
Rds(on) / VCE(sat)90–360 mΩ (depends on Rds(on) class)15–80 mΩ (low Rds(on) per die area)VCE(sat) 1.5–1.9 V at IC ratingAt 20 A, IGBT conduction loss may be lower than 300 mΩ Si MOSFET; above 100 A, IGBT wins hands down
Eon + Eoff (typical, 400 V, 20 A)~100–200 µJ~50–120 µJ (fast, low Qg)~500–1500 µJ (tail‑dominated)Switching loss penalty crosses over below 20 kHz; SiC excels at 50 kHz+
Safe Operating Area (SOA)Full forward‑biased SOA to Tj max; linear‑mode capabilitySquare SOA; limited linear‑mode capability; care with short‑circuitSquare SOA; short‑circuit withstand 5–10 µsIGBTs survive short‑circuit better; MOSFETs offer wider linear operation for e‑fuse/load switch
Typical package footprintTO‑247, D2PAK, PowerPAKTO‑247‑4, D2PAK‑7, SMDTO‑247, TO‑264, modulesSiC often requires Kelvin‑source drive pin; check gate‑driver layout

Note: The VCE(sat) of an IGBT drops only marginally with increasing current, while MOSFET conduction loss rises quadratically. That’s why IGBT modules dominate in 100 A+ motor drives. SiC MOSFETs, however, can compete head‑to‑head with IGBTs in 1200 V applications if you can push the switching frequency high enough to shrink the magnetics. The crossover point is shifting: a 20 kHz, 10 kW inverter might still favor IGBTs on cost, but at 50 kHz the SiC total system cost can be lower because the heatsink and inductors become smaller and cheaper.

From Datasheet to BOM: A 5‑Step Selection Checklist for Power Stage Design

Rather than jumping straight to a part number, work through the following checklist. It forces you to validate the device in the context of your topology, thermal environment, and supply chain—not just the headline specs. At each step, cross‑reference the parameters that matter most.

  1. Define steady‑state voltage and current, including transients. Derate VDS or VCES by at least 20% above the maximum DC‑link voltage, and account for ringing. For IGBTs, check the RBSOA (reverse‑bias safe operating area) at turn‑off; for MOSFETs, verify the avalanche energy rating if you expect unclamped inductive spikes.
  2. Check switching frequency against the 10%‑of‑period rule. Florisera’s MOSFET switching‑time guide recommends that the combined turn‑on and turn‑off time should be less than 10% of the switching period to keep switching losses manageable. For an IGBT, the turn‑off time includes the tail; if the tail extends beyond 10% of the period, you’ll need to either lower the frequency or accept a larger heatsink.
  3. Scrutinize Rds(on) or VCE(sat) at real junction temperature. Silicon MOSFET on‑resistance can double between 25 °C and 125 °C. IGBT VCE(sat) has a flatter temperature coefficient, but still rises. Use Bourns’ IGBT datasheet parameter reference to understand how VCE(sat) and switching energies are specified at Tc = 25 °C and at elevated temperatures. For MOSFETs, IC Online’s practical MOSFET datasheet guide walks you through the normalized Rds(on) curves and gate‑charge figures that determine your gate‑drive power.
  4. Calculate total power loss and thermal impedance. Sum conduction and switching losses, then use the junction‑to‑ambient (or junction‑to‑case) thermal resistance to estimate junction temperature. If Tj exceeds 125 °C for a silicon IGBT or 150 °C for a SiC MOSFET, you’ll need a bigger heatsink or a different part. Pay special attention to the thermal interface material and mounting torque.
  5. Map the part to second sources and lead‑time risk. Use distributor stock data to identify direct drop‑in alternatives or near‑equivalents. When allocation hits, having a cross‑reference already vetted avoids a last‑minute scramble. The sourcing guide from IC Online recommends that you compare gate charge, threshold voltage, and package pinout to ensure the alternative doesn’t require a gate‑driver respin.

The table below summarizes the key datasheet parameters you must verify at each step, along with typical values and why they matter. This isn’t exhaustive, but skipping any of these is a common source of field failures.

Design StepParameter (Symbol)Why It MattersTypical Value to Watch
1. Voltage & current ratingVBR(DSS) / VCES, ID / ICPrevents avalanche breakdown; determines SOA margin≥ 20% headroom above DC‑link max
2. Switching frequencytd(on), tr, td(off), tf (MOSFET); td(on), tr, td(off), tf + tail (IGBT)Switching loss must stay below 10% of periodTotal ton+toff < 1 µs for 100 kHz
3. Conduction lossRds(on) at 125 °C; VCE(sat) at 125 °CDominates loss at low frequencyRds(on) < 200 mΩ for 20 A; VCE(sat) < 2.0 V
4. Gate driveQg, VGS(th), gate plateau voltageDetermines driver current and power dissipationQg < 100 nC for high‑frequency SiC
5. Thermal impedanceRthJC, RthJASets maximum power dissipation for given TambRthJC < 0.5 K/W for TO‑247
5. Supply chainPackage footprint, pinout, gate‑threshold voltageEnsures second‑source compatibilityMatch industry standard TO‑247‑3 or ‑4 pinout

Once you’ve locked in a preferred part, upload the BOM to IC Online to check cross‑references and availability across multiple distributors. A mixed BOM strategy—combining allocated IGBTs with a qualified SiC fallback—can keep your production line running while you negotiate longer‑term contracts.

IGBT vs MOSFET: Questions Engineers and Buyers Ask Before Signing Off

Q: At what voltage and current do IGBTs unequivocally outperform MOSFETs?
Generally above 600 V and 20 A, IGBTs exhibit lower conduction losses due to conductivity modulation. The exact crossover depends on the specific MOSFET Rds(on) and IGBT VCE(sat), but for a 600 V class, a 25 A IGBT will show lower total loss than a silicon MOSFET below 20 kHz. SiC MOSFETs now compete up to 1.2 kV, especially in high‑frequency designs, but at 15 kHz and 100 A, an IGBT module still wins on cost and thermal performance.

Q: How do I mitigate turn‑off tail current in hard‑switched half‑bridges?
Use a negative gate drive voltage (-5 to -15 V) to accelerate carrier extraction and prevent parasitic turn‑on. Optimize the gate resistor to balance turn‑off speed against dv/dt‑induced shoot‑through. Tighten the dead time as much as the tail current allows, and consider soft‑switching topologies (ZVS or ZCS) if the frequency is high enough. Always verify the IGBT’s dv/dt ruggedness and reverse‑bias SOA during the tail period; exceeding these limits can trigger latch‑up or secondary breakdown.

Q: What are realistic lead times for 650 V IGBTs versus SiC MOSFETs right now?
As of early 2025, many 650 V IGBTs are quoted at 30+ weeks, while SiC MOSFETs often run 20–26 weeks. Allocation is tighter in the 600–1200 V range, where EV and industrial demand overlaps. Check distributor stock daily and secure multiple sources; the IC Online platform provides real‑time availability cross‑references that reduce the risk of a single‑source line‑stop.

Q: Can I substitute a MOSFET into an IGBT design if I increase gate drive voltage?
Not directly. A MOSFET may have higher Rds(on) at high voltage, increasing conduction losses and potentially overheating. Even if you adjust gate drive, the SOA, avalanche rating, and dv/dt immunity must be re‑evaluated. The IGBT’s tail current also shapes the dead‑time and commutation loop; a MOSFET replacement may require a different gate‑drive circuit and snubber network. Always perform a full thermal and SOA analysis before substituting.

Q: Which datasheet parameters are critical for outdoor inverter reliability?
Thermal resistance (RthJC) and power dissipation curves at elevated Tj are paramount. Short‑circuit withstand time (tSC) and avalanche energy (EAS) dictate ruggedness under fault conditions. For outdoor use, check package creepage and clearance distances against pollution degree and humidity requirements; a TO‑247 package with 2.5 mm creepage may be insufficient for 600 V in a condensing environment. Look for high‑temperature reverse bias (HTRB) and high‑humidity high‑temperature reverse bias (H3TRB) qualification data, if available.

Q: How do I convince procurement to qualify a second source for IGBTs?
Quantify the cost of a line‑stop: lost revenue per hour, rework costs, and potential damage to customer relationships. Show that a single‑source failure during allocation can shut down a production line for weeks. Present field failure risks from forced substitutions—a similar part number may have different gate threshold voltage or thermal impedance, leading to elevated field returns. Use cross‑reference tools on IC Online to propose a drop‑in alternative and demonstrate that the second source has been pre‑qualified at the datasheet level. The conversation shifts from “why spend extra” to “what’s the cost of not having a backup.”

References & Further Reading

For mixed BOM sourcing and flexible MOQ options, visit IC Online to search real‑time inventory, compare cross‑references, and access second‑source alternatives that keep your power stage designs on schedule.

Related Articles