NCE30H10K - 30V/100A N-Channel Power MOSFET (TO-252)

What Are the Technical Specifications of NCE30H10K?

Key numbers that matter: do the loss math before you trust the 100A rating. At 50A, I²R = 13.75W; at 100A it's 55W. The 110W P_D rating assumes the case stays cool - on a bare TO-252 pad you're nowhere near that.

The gate side is a driver's part, not an MCU's: V_GS(th) typ 3V and 70nC gate charge mean a GPIO through a resistor will hold the die in the linear region and roast it. Budget a gate driver and low-impedance drive at 10V.

Per the NCE30H10K datasheet, the 4.5V drive figure exists, but this is a 10V-driven device in practice - the 5.5mΩ max rating is at V_GS = 10V.

When Should You Use (and NOT Use) the NCE30H10K?

✅ Use NCE30H10K when:

• DC motor drive and H-bridges below 30V. Brush motors, e-bike hubs, and 12V/24V actuators pull 10–100A peaks - the 100A rating with 5.5mΩ is exactly this class.
• BMS discharge and protection paths. High-current pack protection where a logic-level dual like the NCE9926 runs out of headroom.
• Low-voltage inverter and UPS stages. The datasheet's listed applications - hard-switched circuits, UPS, power inverters - all assume a real gate driver.
• DC-DC synchronous rectification at 30V-class rails. Low R_DS(on) and 70nC Q_g balance conduction vs switching loss at tens of kHz.
• Hot-swap and load stages that need avalanche margin. The trench cell is avalanche-characterized, absorbing inductive kicks without dying.

❌ Don't use NCE30H10K when:

• You're driving the gate from an MCU pin. 70nC and 3V threshold mean the pin sits in the linear region during every edge. Use NCE9926 (2.5V-specified, 4nC) or add a proper driver.
• High-side switching without a level shifter. N-Channel high-side at 24V needs the gate 10V above the floating source. For simple high-side, use a P-Channel like NCE9435.
• The board can't dissipate 15W+. 30A through 5.5mΩ is already 5W. If the layout is copper-starved, the 100A rating is fiction.
• Rails above 30V. The rating is hard at 30V - flyback from motors will exceed it. Step to a 40V/60V-class trench part.

What Are the Alternatives to NCE30H10K?

The power-stage decision in one line: 30H10K when the current is real (10A+) and a driver exists; NCE9926 when logic drives directly; NCE9435 when the switch is high-side and simple.

What Are the Typical Applications of NCE30H10K?

DC motor H-bridge (12V/24V): Four NCE30H10Ks form a classic N-Channel H-bridge with a driver IC. At 30A per FET the conduction loss is 5W each - the TO-252 tab pours and driver dead-time control are what keep the bridge alive through locked-rotor peaks.

BMS high-current discharge path: In a 1–7S pack discharge stage, the 30H10K replaces multiple paralleled small FETs. One 100A-rated die with proper copper replaces three or four 30A parts and their gate drivers.

Low-voltage inverter stage: The datasheet lists UPS and inverters. At tens of kHz, the 70nC gate charge needs a driver with a few amps of peak current, and the low R_DS(on) keeps the switching-stage conduction loss small.

Battery tool and e-bike power stages: High-current interrupted flows - throttle, brake, and boost paths. The avalanche rating absorbs the motor's flyback without an external clamp diode in many designs.

Why Buy NCE30H10K from ICMASS?

Lot-tested for the ratings that burn. We verify R_DS(on) at 10V and V_GS(th) on every lot before shipping. At 100A class, a remarked die with 2× the resistance fails exactly where it hurts - heat.

Cross-reference support for the whole NCE power line. Not sure whether your stage wants the 30H10K, the 40H-class, or a paralleled SOP-8 pair? Send us the rail voltage, peak current, and switching frequency - we'll match the die.

BOM consolidation for power products. The same motor-drive BOM carries the 30H10K power stage, the NCE9926 logic switch, and the NCE9435 high-side rail. One shipment covers the entire FET population.

Same-day dispatch, 5–10 days worldwide. Orders before 15:00 CST ship same day via DHL or FedEx.

Frequently Asked Questions About NCE30H10K

Q1: Is 100A really a continuous rating?

A: Only under datasheet conditions - cool case, low ambient. The 100A figure assumes the case is held at 25°C. In the real world, I²R at 100A is 55W, so the package and PCB thermal path decide what you actually get. Budget 30–60A continuous on a well-thermalized board, with 100A as a peak capability.

Q2: Can I drive NCE30H10K directly from an MCU pin?

A: No - and this is the most common way to kill it. With V_GS(th) typ 3V and 70nC gate charge, a GPIO holds the die in the linear region during switching, dumping power into the FET instead of the load. Use a gate driver (peak current of 2–4A is typical for this Q_g) at 10V drive.

Q3: What gate resistor value should I use?

A: 10Ω is the typical starting point; 5Ω for fast switching. Too large a resistor forms an RC with the 3.4nF input capacitance and keeps the FET in the ohmic region too long - excess heat. Too small invites ringing and gate-oxide stress. For paralleled FETs, each die gets its own gate resistor.

Q4: Can I parallel NCE30H10Ks for more current?

A: Yes, but each FET needs its own gate resistor. Thresholds differ between dies; without individual resistors, the lower-threshold FET grabs the current and dies first. With separate gate resistors, parallel devices share reasonably well. The TO-252 tab pours still need to be symmetric for thermal sharing.

Q5: How thick does the PCB trace need to be for 100A?

A: Standard copper foil won't carry it - plan a copper bar or heavy pours. Roughly 15mm² of copper cross-section is the rule of thumb for 100A. On a typical 2oz (70µm) layer, that's a 215mm-wide trace - in practice, solder copper bars or bus bars over the FET tabs.

Q6: NCE30H10K vs NCE9926 vs NCE9435 - which do I need?

A: 30H10K for the power path, 9926 for logic-driven switching, 9435 for high-side. The 30H10K carries 10–100A with a real driver; the NCE9926's 2.5V-specified channels handle 2–4A from GPIOs; the NCE9435 does high-side P-Channel without a level shifter. Most boards need two of the three.

Q7: My motor drive keeps blowing the FETs - what should I check?

A: Check three things: dead time, gate drive, and layout inductance. At 100A, shoot-through (both FETs on) vaporizes dies in microseconds - dead time is mandatory. Undervoltage on the driver leaves FETs half-on, another classic killer. And source-lead inductance induces voltages that can spike the gate oxide - keep the driver low-impedance and the layout tight.

Q8: How do I calculate the heatsink requirement?

A: I²R first, then the thermal budget. At 40A, loss is 8.8W (40² × 5.5mΩ). The TO-252 junction-to-case is low, but junction-to-ambient on bare copper is terrible - size the pour or heatsink so junction stays under ~125°C (with 175°C max, keep 50°C margin).

Q9: My switch burns on power-up, not during operation - why?

A: Inrush into the output capacitance or motor is exceeding the safe area. A discharged capacitor bank looks like a short - 100A-capable die will charge it violently. Add soft-start (slower gate edge or an inrush limiter) and check the SOA during the transient, not just steady state.

Q10: What's the difference between NCE30H10K and NCE30H10K-VB?

A: The -VB is a VBsemi-sourced cross-reference with a lower-loss die: about 2mΩ @ 10V and V_th ~1.9V. Functionally compatible in the TO-252 socket, but the lower threshold changes drive requirements - verify against the specific datasheet before you swap it into a design tuned for the NCE original.