AO3400A PCB Design Guide: Footprint, Pinout, and Alternatives

The AO3400A is the SOT-23 workhorse of low-voltage load switching: a 30 V, 5.7 A N-channel trench MOSFET from Alpha & Omega Semiconductor that turns on from 3.3 V logic and fits on a grain of rice. It is the part you reach for when a TO-220 is absurd overkill — small LEDs, relay coils, low-power DC motors, and battery-protection cutoffs all sit comfortably within its ratings, and it costs pennies in volume.

Three gate-drive tiers are specified: RDS(on) is guaranteed at VGS = 10 V (26.5 mΩ max), 4.5 V (32 mΩ max), and 2.5 V (48 mΩ max). That 2.5 V spec is what sets the AO3400A apart from most SOT-23 MOSFETs — it means a 3.3 V MCU pin can enhance it to a usable on-resistance without an external level shifter. But the threshold can reach 1.45 V, so at 2.5 V you have barely a volt of overdrive in the worst case. The gate is fragile by MOSFET standards: VGS max is ±12 V, not the typical ±20 V, so a 12 V rail with an unclamped gate driver is a destruction recipe.

Thermally, the AO3400A lives and dies by its PCB. The headline 5.7 A uses the ≤ 10-second junction-to-ambient thermal resistance; steady-state RθJA is 125 °C/W, giving only 1.4 W of dissipation headroom at room temperature before the 150 °C junction limit is reached. At worst-case 48 mΩ that means about 5.4 A — but at 70 °C ambient you have less than half that. Every square millimeter of copper on the drain pin buys you more current. This is a switch, not a linear regulator: keep it fully on or fully off.

What breaks boards

1. 01

   VGS max is ±12 V — not the usual ±20 V

   Most power MOSFETs tolerate ±20 V on the gate, but the AO3400A is rated for only ±12 V. Driving it from a 12 V rail with a gate driver that swings to the rail gives zero margin; any overshoot, ringing, or ESD event that pushes the gate above 12 V can puncture the gate oxide. If your gate-drive rail is 12 V, add a clamp or pick a part rated for ±20 V.

2. 02

   5.7 A is a short-pulse number — steady-state is PD-limited

   The 5.7 A continuous rating uses the t ≤ 10 s thermal resistance (RθJA = 90 °C/W max). Under steady-state conditions RθJA is 125 °C/W max, which means 1.4 W of dissipation is your absolute ceiling at 25 °C ambient. At worst-case RDS(on) of 48 mΩ (2.5 V gate drive), you hit 1.4 W at only 5.4 A — and at 70 °C ambient with the 150 °C Tj(max) limit you have just 80 °C of headroom, or about 0.64 W. On a compact PCB with thin copper the thermal resistance is worse; derate accordingly.

3. 03

   2.5 V gate drive works — but barely at worst-case threshold

   The RDS(on) spec at VGS = 2.5 V (48 mΩ max) is what makes the AO3400A usable from 3.3 V logic. But VGS(th) can be as high as 1.45 V, leaving only about 1 V of overdrive in the worst case. RDS(on) rises steeply below 3 V (see Figure 5 in the datasheet). For a 3.3 V MCU pin sourcing a few mA into the gate, the FET will turn on, but don't count on the 48 mΩ number at high current with a corner-threshold part. From 1.8 V logic this part will not turn on at all — pick a FET specified at 1.5 V or 1.8 V gate drive.

4. 04

   The SOT-23 package is the heatsink — the PCB is your thermal design

   There is no exposed pad, no heatsink tab, and no bolt hole. The three tiny SOT-23 leads are the only thermal path, and the drain (pin 3) carries most of the heat. Maximize the copper area connected to the drain pad — use 2 oz copper, flood the drain node with a large polygon on both layers, and stitch with vias if you have a ground or power plane. Every 10 °C/W you can shave off the junction-to-ambient thermal resistance buys you measurable extra current.

5. 05

   AO3400A is not AO3400 — check the marking

   The non-A variant (AO3400) is a different MOSFET with lower current and slightly different RDS(on) specs. The AO3400A marking is 'X0' (laser-marked on the SOT-23 package). The green-product AO3400AL shares the same X0 marking. Counterfeit AO3400As from marketplace sellers and non-franchised distributors are endemic in the hobby world — they're often remarked lower-spec FETs or entirely different dice. Buy from franchised distribution and verify die marking when performance matters.

Key specifications

Verified against the manufacturer datasheet on 2026-07-10. Confirm the current revision before production use.

Alternatives

• IRLZ44N: through-hole TO-220 for higher currents. The IRLZ44N handles 47 A and has a real heatsink, but needs 5 V gate drive and takes up PCB real estate. Step up to this when the SOT-23 AO3400A runs out of thermal headroom.
• AO3401A: P-channel companion from the same AOS family in SOT-23. −30 V, −4.2 A. Use when you need a high-side switch without a charge pump.
• SI2302: Vishay/Siliconix SOT-23 N-channel, 20 V, 2.4 A. Lower ratings but widely multi-sourced. The common 'AO3400A' breakout boards from marketplace sellers often carry an SI2302 die instead — verify before trusting the silk screen.
• DMG2302U: Diodes Inc. SOT-23 N-channel, 20 V, 3.5 A. Also specifies RDS(on) at 2.5 V gate drive, so it's a functional drop-in from 3.3 V logic, but the 20 V VDS rating is a step down from the AO3400A's 30 V.

Common questions

Can I drive an AO3400A from a 3.3 V GPIO?

Yes — the AO3400A has an RDS(on) specification at VGS = 2.5 V (48 mΩ max), which means a 3.3 V MCU pin can enhance it. But VGS(th) can reach 1.45 V, so at the worst-case corner you have less than 2 V of overdrive. The FET will turn on, but RDS(on) is higher than the 2.5 V spec and rises sharply below 3 V. For a few hundred mA it's fine; above an amp, check the transfer curves and derate.

What is the difference between AO3400A and AO3400?

They are different parts. The AO3400 (no 'A') is an older variant with lower current ratings and different RDS(on) characteristics. The AO3400A is the current production part, marked 'X0' on the package. The AO3400AL is the green/RoHS designation and shares the same X0 marking and specifications.

Why is my AO3400A running hot at only 3 A?

Two likely culprits. First, gate drive: if you're driving the gate below 4.5 V, RDS(on) is higher than the headline 26.5 mΩ. At 2.5 V worst-case 48 mΩ, 3 A dissipates 0.43 W — enough to push the junction 54 °C above ambient at steady-state RθJA of 125 °C/W. Second, PCB copper: the SOT-23 relies on the drain pad copper for heatsinking; a tiny pad on a thin 1 oz board will have much worse thermal resistance than the datasheet's 1 in² 2 oz copper test condition.

Can I use an AO3400A as a high-side switch?

You can, but you need a gate voltage higher than the source voltage by at least the threshold plus overdrive. With an N-channel FET in a high-side configuration, the source sits at the load voltage, so the gate must be driven above the supply rail. Use a P-channel FET (like the AO3401A) for simple high-side switching, or add a bootstrap gate driver for N-channel high-side operation.

Is the AO3400A suitable for PWM at high frequency?

At 7 nC total gate charge and 630 pF input capacitance, the AO3400A switches quickly even from a modest GPIO — rise and fall times are in the single-digit nanoseconds. At low PWM frequencies (sub-kHz) a GPIO with a small series gate resistor works fine. Above 10 kHz at several amps, switching losses start to dominate, and ringing from the fast edges can couple into the gate. A gate resistor in the 10–100 Ω range is recommended to damp ringing, and consider a dedicated gate driver above 50 kHz or at higher currents.