ICMASS.COM - Global Electronic Component Solutions Provider

Q: Is the TO-252 tab ground or drain?

Drain - electrically the same net as pin 2. A heatsink pad tied to the ground plane shorts the 100V rail. If the tab sits on an isolated copper island, it must not connect to GND unless the schematic says so.

Q: What are pins 1, 2, and 3 on the NCE0140KA?

1 is Gate, 2 is Drain, 3 is Source. The tab is also Drain. In a low-side switch, the load connects to pin 2 and the tab, pin 3 goes to ground, and the gate drive goes to pin 1.

Q: Is it pin-compatible with SOP-8 MOSFETs like the NCE0110AS?

No - completely different pinouts. SOP-8 uses 1-3 Source / 4 Gate / 5-8 Drain; TO-252 uses 1 Gate / 2 Drain / 3 Source + tab. Moving packages is a re-layout, not a swap.

Q: Can I drive it from a 3.3V MCU pin?

It turns on, but not fully - there is no 3V spec. RDS(on) is guaranteed from 4.5V; at 3.3V the die runs partially enhanced at 2-3× the rated resistance. For a 3.3V rail, the NCE0125AK has the 3V spec.

Q: What's the best way to dissipate heat?

The tab is the heat path - feed it copper. 2 oz pours extending 1-3mm past the pad, 6-10 thermal vias of 0.3-0.5mm into inner planes, solder mask opened over the vias. The 1.07°C/W RθJC only helps if the board has somewhere to put the heat.

Q: What is the 160A pulse rating for?

Microsecond-scale inrush and stall events: hot-plug capacitor banks, motor-start spikes, changeover glitches. The die must return to normal temperature between events - it's not a license for 160A average.

Q: How do I verify a genuine NCE0140KA?

Measure RDS(on) at 4.5V and 10V. A genuine part reads near 14mΩ typ at 10V and 13mΩ at 4.5V. Counterfeit parts remarked from smaller dies fail the 4.5V check by 2-3× while passing visual inspection.

All Categories

Language

Home > Blog > Blog

NCE0140KA Pinout, Specifications & Common Mistakes

2026/9/1 21:04:50

The NCE0140KA is a single N-Channel power MOSFET in TO-252 (DPAK): 100V/40A with R_DS(on) of 17mΩ max @ 10V and a 94nC gate charge. Pin 1 is Gate, pin 2 is Drain, pin 3 is Source — and the tab is Drain, electrically the same net as pin 2.

This page covers the pin diagram, the specs that matter, the circuit that uses it most (low-side switching), and the mistakes that show up in support queues.

NCE0140KA Pin Diagram

Pin mapping at a glance: Gate on 1, Drain on 2, Source on 3. The large tab on top is the Drain — the same net as pin 2, and the package's main thermal path out of the die.

Pin	Name	Type	Description
1	G	Gate	Gate — drive to V_GS ≥ 4.5V to fully enhance (N-Channel)
2	D	Drain	Drain — high-current path (load side in low-side use)
3	S	Source	Source — ties to GND in low-side use
Tab	D	Drain	Tab = Drain (same net as pin 2) — heatsink path, never tie to GND

Key Specifications

Parameter	Value
Drain-Source Voltage (V_DS)	100V (typ 110V breakdown)
Gate-Source Voltage (V_GS)	±20V
Continuous Drain Current	40A @ 25°C / 28A @ 100°C
Pulsed Drain Current	160A
R_DS(on) @ 10V	17mΩ max (14mΩ typ)
R_DS(on) @ 4.5V	18mΩ max (13mΩ typ)
V_GS(th)	0.9V to 1.5V (typ 1.1V)
Power Dissipation	140W @ 25°C (derate 0.94W/°C)
Thermal Resistance R_θJC	1.07°C/W
Gate Charge	94nC @ 10V (Q_gs 16nC / Q_gd 24nC)
Input Capacitance	3400pF typ
Body Diode Forward Voltage	0.85V typ / 1.2V max @ 28A
Body Diode Reverse Recovery	33ns
Single-Pulse Avalanche Energy	520mJ (100% UIS tested)
Junction Temperature	-55°C to +175°C

Specs that matter: per the NCE0140KA datasheet (Rev 1.2), R_DS(on) is guaranteed at 4.5V and 10V — a 5V rail turns it fully on, but there is no 3V spec (that's the NCE0125AK's slot in the family).

The 140W P_D is a T_C = 25°C number; with a realistic case temperature the honest envelope is 20–28A. The tab copper decides which number you actually get.

Typical Application Circuits

Circuit 1 — low-side load switch: load between VIN and the drain, source to ground, gate driven from a 5V GPIO through 100Ω with a 100kΩ pull-down to keep the channel off at power-up. The 4.5V spec floor matches the 5V rail directly.

The tab needs the copper it was designed for: 2 oz pours extending 1–3mm past the pad, 6–10 thermal vias to inner planes. That footprint is the difference between the 140W paper rating and 20–28A sustained.

Off-time on a 12V rail with the 94nC gate charge (Q_g × R ÷ V):

1kΩ pull-down path~8µs

10kΩ pull-down path~78µs

But why does drive stiffness keep coming up? Because 94nC is the highest gate charge in the 100V family — a 10kΩ gate resistor that was fine on a small SOT-23 turns this part into a slow-motion switch.

Circuit 2 — synchronous rectification (secondary side): in a 48V-to-12V brick, the NCE0140KA replaces the Schottky on the secondary side.

The 33ns body-diode recovery keeps reverse-recovery loss small, and the 4.5V drive matches a 5V bias rail.

Drive it from a real half-bridge driver — at SR frequencies, 94nC of gate charge is beyond a GPIO's budget.

Common Mistakes When Using NCE0140KA

Tying the tab to ground. The DPAK tab is the drain — a heatsink pad connected to the GND plane shorts the 100V rail. This is the most common TO-252 mistake in support queues, and it fails at power-up, not at load.
Treating the 40A rating as continuous. It's a T_C = 25°C number. The datasheet itself derates to 28A at 100°C; with realistic board cooling the honest envelope is 20–28A. At 17mΩ, 40A is 27W — the tab copper decides whether that's a heat spot or a fire.
Driving from a 3.3V pin and expecting full enhancement. The spec stops at 4.5V. V_GS(th) is typ 1.1V, so 3.3V turns it on partially — at 2–3× rated R_DS(on). If the rail is 3.3V, the family member with a 3V spec is the NCE0125AK.
Using the 160A pulse rating as a continuous number. The pulse rating assumes microsecond-scale events with the die returning to temperature between them — hot-plug inrush and stall currents, not an hour of overload.
Assuming 520mJ avalanche means no freewheeling diode. Per the SE discussions on avalanche-rated MOSFETs, the rating is a design margin, not a substitute for clamping. The diode handles the normal kickback; the 520mJ covers the fault the diode misses.
Carrying SOP-8 pin habits over. An SOP-8 MOSFET uses 1–3 Source / 4 Gate / 5–8 Drain; the TO-252 uses 1 Gate / 2 Drain / 3 Source + tab. Swapping packages without re-checking the pinout is how the gate ends up on the drain net.
Buying by marking alone. The 40A marking is a counterfeit magnet, and generic "40N10" labels are a remarking favorite. Measure R_DS(on) at 4.5V and 10V — a genuine part reads near 14mΩ typ at 10V, and a remarked smaller die fails by 2–3×.

How do you know if your drive is strong enough? Measure the off-time — if the switch takes tens of microseconds to drop, the gate resistor is too weak for this 94nC gate charge.

Where to Buy NCE0140KA

NCE0140KA is typically in stock at ICMASS with same-day dispatch from Shenzhen. We lot-test R_DS(on) at 4.5V and 10V before shipping. Contact us for current pricing and quantity pricing on the 2500-piece reel.

Frequently Asked Questions

Q1: Is the TO-252 tab ground or drain?

A: Drain — electrically the same net as pin 2. This is the classic DPAK mistake: a heatsink pad tied to the ground plane shorts the 100V rail. If the tab sits on an isolated copper island, the island must not connect to GND unless the schematic says so.

Q2: What are pins 1, 2, and 3 on the NCE0140KA?

A: 1 is Gate, 2 is Drain, 3 is Source. The tab is also Drain. In a low-side switch, the load connects to pin 2 (and the tab), pin 3 goes to ground, and the gate drive goes to pin 1.

Q3: Is it pin-compatible with SOP-8 MOSFETs like the NCE0110AS?

A: No — completely different pinouts. The SOP-8 uses 1–3 Source / 4 Gate / 5–8 Drain; the TO-252 uses 1 Gate / 2 Drain / 3 Source + tab. Moving between packages is a re-layout, not a swap — including the tab copper the TO-252 needs.

Q4: Is it pin-compatible with the IRF540 TO-220?

A: Electrically, yes — the pin order matches: 1 Gate, 2 Drain, 3 Source, tab = Drain. The difference is mechanical: TO-220 is through-hole with a screw-mount tab; TO-252 reflows onto the board. Same electrical order, completely different board design.

Q5: Can I drive it from a 3.3V MCU pin?

A: It turns on, but not fully — there is no 3V spec. R_DS(on) is guaranteed from 4.5V; at 3.3V the die runs partially enhanced at 2–3× the rated resistance. For a 3.3V rail, the family member with a 3V spec is the NCE0125AK.

Q6: What's the best way to dissipate heat?

A: The tab is the heat path — feed it copper. 2 oz pours extending 1–3mm past the pad on every side, 6–10 thermal vias of 0.3–0.5mm into inner planes, solder mask opened over the vias so they fill. The 1.07°C/W R_θJC only helps if the board side has somewhere to put the heat.

Q7: What is the 160A pulse rating for?

A: Microsecond-scale inrush and stall events. Hot-plug capacitor banks, motor-start spikes, and changeover glitches. The die must return to normal temperature between events — it's not a license for 160A average, and the 520mJ energy limit is the real constraint for inductive loads.

Q8: How do I verify a genuine NCE0140KA?

A: Measure R_DS(on) at 4.5V and 10V with a bench supply and milliohm meter. A genuine part reads near 14mΩ typ at 10V and 13mΩ at 4.5V. Counterfeit parts remarked from smaller dies fail the 4.5V check by 2–3× while passing visual inspection — the reason we batch-test before shipping.