MOSFET or Transistor: Twelve Practical Scenarios That Pick the Winner

The MOSFET-versus-transistor question is usually answered with equations, but most builders face it as a list of loads: an LED here, a motor there, a sensor needing a signal. Twelve practical scenarios, each solved by one parameter, put the decision where it belongs — in the application. This article leads with the twelve-scenario table, walks the first six small-load and level-shift cases, walks the last six motor, PFC, and load-switch cases, names the one parameter that decides each, and closes with the real-world caveats that change the textbook answers.

Last updated: October 11, 2026

Twelve Scenarios, Twelve Winners: The Table up Front

The complete set up front is the fastest way to use this article, and the table is the tool. Each row names the load, the winner, and the single parameter that decided it.

Scenario	Load	Winner	The one parameter
1	Level shift (3.3 V to 5 V signal)	MOSFET	Ease of gating a small signal
2	LED on/off switch	Either	Current and drive simplicity
3	Relay or solenoid drive	MOSFET	Big current at low drive voltage
4	Small sensor read	BJT	Low-signal linear behavior
5	Piezo buzzer	Either	Frequency and drive level
6	Logic rail switch	MOSFET	Low quiescent leak at rest
7	DC motor (load switch)	MOSFET	High current and low heat
8	PFC / rectifier role	MOSFET	Fast switching at power level
9	Audio signal switch	BJT	Linear gain behavior
10	High-voltage load switch	MOSFET (or IGBT)	Voltage class and margin
11	Photo / light sensor preamp	BJT	Low-signal gain
12	Heater or resistive load	MOSFET	Current and pulse control

The table is the whole decision in one view, and the scenario walkthroughs give the reason each cell landed where it did. The MOSFET vs BJT comparison and the NPN transistor articles provide the theory this article deliberately leaves out; here the parameter and the load decide.

Each row has exactly one deciding parameter because that is how builders actually choose: the load asks one hard question, and the device that answers it wins. The rest of the article walks the twelve rows in two groups, then names the parameters explicitly and the caveats that complicate them.

Scenario Walkthroughs 1-6: Small Loads and Level Shifts

The first six scenarios cover the small-signal and level-shift territory where the two families are genuinely close, and the deciding parameter is usually the drive or the linearity.

A level shift from a 3.3 V logic rail to a 5 V signal favors the MOSFET: its gate can be driven directly from the logic rail, and its switching is clean at the small signal level. An LED on-off switch is nearly a draw — both families switch small currents easily, and the decision falls to what the board already has. A relay or solenoid drive favors the MOSFET because it needs a big current at a low drive voltage, which is the MOSFET’s strength. A small sensor read favors the BJT, whose linear gain behavior suits low-signal amplification. A piezo buzzer is another near-draw, decided by the frequency and the drive level. A logic-rail switch favors the MOSFET, whose low quiescent leak at rest keeps the sleeping rail quiet. The logic-level MOSFET article covers the direct-drive cases, and the digital switching transistor article covers the transistor side.

The pattern in the first six is the divide that repeats: when the load needs clean switching at low drive voltage, the MOSFET wins; when it needs linear small-signal gain, the BJT wins. The parameter is the drive-versus-linearity question, and the scenarios below carry the same divide into the heavier loads.

Scenario Walkthroughs 7-12: Motors, PFC, and Load Switches

The last six scenarios move to the power territory, where the MOSFET’s ratings and thermal behavior dominate and the BJT mostly drops out.

A DC motor load switch favors the MOSFET: the motor draws high current, and the MOSFET carries it with low on-resistance heat and easy gate drive. A PFC or rectifier role favors the MOSFET, whose fast switching at the power level carries the stage. An audio signal switch reverts to the BJT, whose linear gain suits the analog path. A high-voltage load switch favors the MOSFET when the voltage class fits — and an IGBT when it does not; the voltage and margin decide. A photo or light-sensor preamp favors the BJT for the low-signal gain. A heater or resistive load favors the MOSFET for the current and pulse control. The MOSFET for SMPS article and the power MOSFET selection guide cover the power-territory cases, and the IGBT guide takes the high-voltage relay when the voltage class passes the MOSFET’s corner.

The pattern in the last six is the same divide at higher power: switching and current favor the MOSFET, linear signal gain favors the BJT, and the voltage class introduces the IGBT as the MOSFET’s high-voltage extension. The twelve rows are variations on the same two questions — can the device be driven, and does the load need linear gain.

Where the Numbers Separate a MOSFET from a BJT

The scenario table above settles most cases on intent, but one calculation settles the rest: dissipation. Compare a small-signal BJT against a modern power MOSFET on the same load and the crossover stops being a rule of thumb.

onsemi specifies the 2N3904 at 200 mA continuous collector current and 625 mW total dissipation at 25 °C, falling to 250 mW at 60 °C, with a 200 °C/W junction-to-ambient resistance. The NTMFS5C430NL power MOSFET is rated 40 V with an on-resistance of 1.4 mΩ at a 10 V gate drive and 2.2 mΩ at 4.5 V, 38 A continuous at a 25 °C ambient on a 650 mm² copper pad, a 40 °C/W junction-to-ambient path, and a 175 °C junction limit.

Switching 5 A	Small-signal BJT (2N3904)	Power MOSFET (NTMFS5C430NL)
Conduction loss	VCE(sat) × 5 A ≈ 1.5 W at 0.3 V	I²R = 25 × 0.0014 ≈ 0.035 W
Against its own package	Six times the 250 mW available at 60 °C	Well inside a 38 A ambient rating
Junction rise on the datasheet path	300 °C per watt — the part cannot survive it	About 1.4 °C for the same current
What the designer must add	More base drive, which does not reduce VCE(sat) much	A gate driver that reaches the voltage the resistance is quoted at

Read the last row carefully, because it is where a MOSFET selection quietly goes wrong. The on-resistance has two published values for a reason: driving the same device from a 4.5 V rail instead of 10 V raises its resistance by about 57 %. At 20 A that is the difference between 0.56 W and 0.88 W of heat from a change in gate drive alone, with no change to the load. This is a model with stated assumptions: substitute your own gate voltage and current, then re-solve once, because resistance rises with junction temperature.

The boundary condition for the BJT side is thermal rather than electrical. Its 625 mW rating is quoted at 25 °C and derates by 5.0 mW/°C, so the usable dissipation in a 60 °C enclosure is barely a quarter of a watt — which is why the same 5 A load that looks marginal in a datasheet table is not marginal at all in practice, it is impossible.

The One Parameter That Decides Each

The twelve scenarios reduce to a short list of deciding parameters, and naming the list turns the article into a decision method. Four parameters cover all twelve rows.

The first is drive voltage: can the logic rail or the controller drive the device directly? The MOSFET’s easy gate drive wins where the drive is low. The second is current and heat: does the load draw enough current to make on-resistance and thermal path the deciding number? The MOSFET wins the high-current rows. The third is linearity: does the load need small-signal linear gain? The BJT wins the sensor and audio rows. The fourth is voltage class: does the load’s voltage exceed the MOSFET’s comfortable blocking corner? The IGBT (or a higher-voltage MOSFET) takes over. The MOSFET selection guide and the MOSFET polarity article develop the first two parameters, and the IGBT guide develops the fourth.

The parameter map is the article’s compact answer: drive, current, linearity, and voltage class — ask the load which one it cares about, and the device follows. The twelve scenarios are the demonstration, and the four parameters are the method.

The four parameters also explain why the twelve rows cluster the way they do, and reading the cluster is the shortcut that a builder internalizes with practice. The small-signal rows – the sensor, the photo preamp, the audio switch – all share the linearity parameter and cluster around the BJT. The power rows – the motor, the relay, the heater, the PFC – all share the current-and-drive parameter and cluster around the MOSFET. The level-shift and logic-rail rows share the drive-voltage parameter and sit with the MOSFET’s easy gating. The high-voltage row is the one that clusters by voltage class and reaches for the IGBT. Recognizing the cluster means a builder can answer most future scenarios by asking which cluster the load belongs to, without re-reading the twelve rows each time. The power MOSFET selection guide and the NPN transistor articles give the family context that the clustering draws on.

The same clustering is why the article deliberately avoids equations. Each scenario is solved by one parameter because the builder’s real question – what load am I switching and what is my drive – is a one-parameter question at the decision point. The equations that a full comparison would add are the subject of the MOSFET vs BJT comparison; here the twelve loads and four parameters are the practical method, and the caveats keep it honest against the real board.

Real-World Caveats That Change Twelve Textbook Answers

The final section is the honesty check: the twelve answers are textbook answers, and the real world changes some of them. Four caveats matter most.

First, duty and thermal reality beat the page: a scenario that “should” use a BJT at high duty will fail thermally while a MOSFET carries it, because the page did not include the heat. Second, drive penalties are real: a MOSFET that needs a driver, or a BJT that needs a bias network, can cost more than the part it replaces. Third, voltage derating changes the high-voltage row: a load at the edge of the MOSFET’s class needs the derating rule before the part is trusted. Fourth, the existing board wins ties: with the electrical parameters close, the part already in the BOM and in stock is the right answer, and the sourcing strategy article covers the supply reason the tie-break exists. The derating guide and the field reliability checklist cover the thermal and margin shape of the caveats.

The twelve caveats also make the method reusable beyond the twelve rows: any new load is solved by asking the four parameters, finding the cluster, and then checking the duty, the drive cost, the derating, and the BOM. That is the complete decision habit this article is built to leave behind – the table starts the reader, the parameters generalize it, and the caveats keep it honest with the real board.

The twelve scenarios answer the MOSFET-or-transistor question without equations, and the four caveats keep the answers honest. Drive, current, linearity, and voltage class decide the load; duty, drive cost, derating, and the existing BOM change the verdict. The builder who reads the table, walks the scenario, and then applies the caveats ends with the part that works in the real board, not the one that wins the textbook page. The power MOSFET category and the transistor category supply the parts both answers land on.

FAQs

At what current should I switch from a BJT to a MOSFET?

There is no single current. The crossover is where the conduction loss exceeds what the package can dissipate at your ambient, and that depends on the forward drop of the BJT against the on-resistance of the MOSFET at your gate drive voltage.

Why does a MOSFET need a gate driver if it is voltage controlled?

Because the quoted on-resistance is tied to a specific gate voltage. Driving a part specified at 10 V from a 4.5 V rail raises its resistance, and at high current that difference turns directly into heat.

Is a BJT ever the better switch?

Yes, at low currents where its simplicity, low cost, and predictable saturation voltage matter more than efficiency, and where the load is small enough that the package dissipation stays well inside its derated rating.

Why does the dissipation rating fall so much with temperature?

The rating is set by how much heat the package can pass at a given ambient for a given junction limit. As the ambient rises, less headroom remains, so the allowable dissipation falls along the thermal resistance line.

Continue through this engineering topic

For the complete selection, design, verification, and troubleshooting path, explore the MOSFET IGBT and Transistor Hub.

• Practical guide: Testing a MOSFET with a Multimeter: The Capacitance Sign, Diode Check, and Short Detection
• Practical guide: Zener Gate Protection for MOSFETs: Rescuing Gates from Overvoltage and Spikes
• Practical guide: IGBT Basics for Power Electronics: Ratings, Structure, and the Selection Questions That Matter
• Ask Good-Ark technical support about this application