Power Discretes

MOSFETs

Trench, super-junction and P-channel devices for switching, load control and reverse-polarity protection, from 20 V logic-level parts to 1,200 V super-junction devices.

5 variants 5 catalogue items 1.2 mΩ and up 500-900 V simple high side

Need a specific part?

Send the part number, quantity and target price. We come back with availability, cross-references and the datasheet.

Request a quote Ask an engineer

Lead timeQuoted per line

MOQBy line item

Cross-referenceOn request

Most MOSFET selection starts and ends with on-resistance, which is the least interesting parameter in a switching circuit. Conduction loss is RDS(on) × I² × duty, but switching loss is a function of gate charge, output capacitance and the speed of the driver, and at a few hundred kilohertz the switching term usually dominates. A part with half the RDS(on) and twice the gate charge can easily be the hotter of the two.

The second trap is the gate drive voltage. A device specified at 10 V will not deliver its datasheet RDS(on) from a 4.5 V logic driver, and the difference is often a factor of two or three. Logic-level parts are specified at 4.5 V or 2.5 V for exactly this reason, and mixing the two in a BOM is a reliable way to build a board that runs hot only in production.

We supply the voltage classes that cover industrial and automotive switching: 20-100 V trench devices for synchronous rectification and load switching, 500-900 V super-junction parts for PFC and LLC primaries, and P-channel devices for high-side and reverse-polarity duty. Automotive grades to AEC-Q101 are available with the qualification documentation.

Range Overview

MOSFETs variants

Each variant exists because a different circuit position asks for something different: voltage, current, package and stability trade against each other.

Trench LV (20-100 V)

1.2 mΩ and up

Low RDS(on) per unit area, fast body diode, the standard choice for synchronous buck, OR-ing and load switches.

Super-junction HV (500-900 V)

500-900 V

Charge-compensated structure that keeps RDS(on) low at high voltage, used in PFC, LLC and flyback primaries.

P-channel

simple high side

Conducts with a gate below the source, so a high-side switch needs no charge pump. Higher RDS(on) than an N-channel of the same size.

Dual and complementary

half bridge

Two devices in one package for half-bridge, synchronous rectifier and level-shifter stages, saving board area and matching thermal behaviour.

GaN HEMT (adjacent)

high frequency

Not a MOSFET, but the natural upgrade where switching frequency or power density has reached the silicon limit. Zero reverse recovery and much lower Qg.

Catalogue

MOSFETs — quoted on request

Browse all products

Technical Range

Typical specifications

Figures below are the range across the family, not a single model. Specify the exact range, accuracy and connection on the quote request and we will confirm which variant meets it.

Request a specification check
Parameter Typical
VDS 20 V to 1,200 V
ID (continuous) 1 A to 300 A
RDS(on) 1.2 mΩ to 10 Ω
Total gate charge Qg 2 nC to 300 nC
Output charge Qoss 2 nC to 500 nC
VGS(th) 1 V to 5 V
Packages DFN 3×3, SO-8, TO-220, D2PAK, TO-247
Qualification AEC-Q101, automotive grade

Applications

Where this family is used

Synchronous buck and boost

High-side control FET and low-side synchronous FET, where the low-side part is chosen mainly on RDS(on) and the high-side on Qg and Coss.

Motor drive H-bridges

Three-phase and brushed DC bridges, where SOA during stall and the body diode recovery shape the choice more than steady-state loss.

Load switching and hot-plug

Inrush limiting and rail sequencing, where the safe operating area and the SOA derating curve decide whether the part survives.

PFC and LLC primary

Super-junction devices in continuous-conduction PFC and resonant half-bridge primaries, selected for low Qrr and controlled dv/dt.

Selection Guide

How to specify correctly

  1. 1

    Match RDS(on) to the actual gate drive

    Read the RDS(on) row for the VGS your driver produces. A 10 V-specified part driven from 4.5 V may show three times the on-resistance, which turns a 30 W loss calculation into 90 W.

  2. 2

    Budget gate charge against driver current

    Switching time is Qg divided by drive current. If the driver sources 1 A and the part needs 60 nC, the transition takes 60 ns, and the switching loss scales with the time spent in the linear region.

  3. 3

    Check the SOA for hot-plug and stall

    The DC curve in the safe operating area graph is far below the pulse curve. Load switches and motor drives spend milliseconds in a high-current low-voltage state, which is exactly where an undersized part fails.

  4. 4

    Thermal resistance decides the real current

    The same die in a DFN 3×3 and a TO-247 has the same RDS(on) but a junction-to-ambient resistance that differs by a factor of three. The package, the copper area and the airflow set the continuous current, not the ID rating.

  5. 5

    Paralleling needs matched parts and separate gates

    Two devices share current by RDS(on) only above the threshold. Below it the faster device takes everything, so each needs its own gate resistor and the pair should come from the same production lot.

FAQ

MOSFETs questions

Why is RDS(on) higher than the datasheet when driven at 4.5 V?

The datasheet has separate rows for each gate voltage, and the 4.5 V row is usually two to three times the 10 V figure. A logic-level part is specified to be fully enhanced at 4.5 V, but its on-resistance at that voltage is still higher than at 10 V. Always read the row that matches the driver.

Can I parallel two MOSFETs to double the current?

Yes, with care. Above threshold they share by RDS(on), which has a positive temperature coefficient, so the hotter device takes less current and the pair self-balances. The risk is during switching, when threshold voltage mismatch makes one device carry the transient. Use separate gate resistors and parts from the same lot.

What does Qrr do to my efficiency?

In a hard-switched bridge, the body diode of the low-side device conducts during dead time and must recover when the high side turns on. The recovery charge Qrr flows as a current spike through both devices, producing a loss proportional to Qrr × VDS × frequency. At high voltage and high frequency it can dominate the total loss.

When is a P-channel better than an N-channel plus a charge pump?

Below about 30 V and a few amps, a P-channel high-side switch is simpler and cheaper than an N-channel with a charge pump, and it needs no bootstrap. Above that the RDS(on) penalty and the die area make the N-channel plus driver the better engineering choice.

How do I read the safe operating area graph?

Each curve is a permitted combination of drain current and drain-source voltage for a given pulse width, bounded by the RDS(on) line, the current limit and the thermal limit. For a hot-plug or stall event, find the pulse width that matches the event, apply the derating for case temperature, and confirm the operating point sits inside the curve with margin.

Related categories

Tell us the application and we will confirm the right variant.