2N2222 transistor: pinout for every package, ratings and substitutes

Read the print on the flat face. P2N2222A (printed as P2N2 over 222A) is onsemi's TO-92 part and runs collector, base, emitter from the left. PN2222A is the Fairchild-designed part that onsemi now sells, and it runs emitter, base, collector. A metal can marked 2N2222A is the original TO-18 part, with the emitter next to the tab. A tiny three-lead part marked 1P is an MMBT2222A in SOT-23.

A TO-92 marked only "2N2222" or "2N2222A" could have come from any maker. Do not trust a drawing for it. Check it with a meter, as described further down, before you solder it.

The versions without the A are older and weaker. In onsemi's SOT-23 datasheet the plain MMBT2222 is rated 30 V collector to emitter against 40 V for the MMBT2222A, and 5 V emitter to base against 6 V. Buy the A.

The makers agree on the voltages and disagree on current and power. The P2N2222A is rated 600 mA, the PN2222A 1 A and the TO-18 can 800 mA. The SOT-23 part is rated 225 mW, about a third of what the TO-92 parts can dissipate, so it runs much hotter at the same current. Microchip's 2N2222A, qualified to MIL-PRF-19500, is rated 50 V collector to emitter where Central's is rated 40 V.

These are limits, not operating points. Design for well under them: a 2N2222A carrying 500 mA works, but by the datasheet its saturation voltage can reach 1 V at that current, which is half a watt in a part rated for 625 mW.

Two things in this table matter more than the rest. First, gain is a range. At 150 mA and 10 V the hFE can be anywhere from 100 to 300, and at 1 V across the transistor (a switch that is nearly on) only 50 is guaranteed. Never design for the 300. Second, the 0.3 V saturation voltage is specified with the base current at one tenth of the collector current (150 mA and 15 mA). That ratio is what the switching example below uses.

The 300 MHz transition frequency is why the 2N2222 turns up in RF oscillators and fast pulse circuits as well as relay drivers. The onsemi datasheet gives storage time as 225 ns at most at 150 mA, so switching at tens of kilohertz, as in PWM motor or LED drive, is well within reach.

• Switching a relay, a buzzer, a small motor or a short LED strip from a microcontroller pin, up to a few hundred milliamps. The BJT as a switch lesson covers the theory.
• Small-signal amplifiers, such as a common-emitter stage or an emitter follower.
• Oscillators and pulse circuits into the tens of megahertz.

It is the wrong part above about half an amp, and for anything that needs to run cool at currents over a few hundred milliamps. At that point a logic-level MOSFET wastes far less power; MOSFET or BJT compares the two.

A 12 V load that draws 100 mA, say a small relay coil, switched from a 5 V microcontroller pin. The emitter goes to ground, the load sits between +12 V and the collector, and a resistor feeds the base from the pin.

From a 3.3 V pin the same working gives R ≤ 210 Ω, so 180 Ω, and a base current between 11.7 mA and 15 mA. Look up your chip's per-pin current limit before you settle on that. If the pin cannot supply it, a logic-level MOSFET needs almost no gate current (see the MOSFET as a switch).

Two parts finish the circuit. A relay coil or motor needs a diode across it (cathode to +12 V), or the voltage spike at switch-off can exceed the 40 V rating; the flyback diode and relay driving lessons explain why. A 10 kΩ resistor from base to emitter holds the transistor off while the microcontroller is resetting and its pin is floating. It takes only about 0.1 mA from the base drive.

The two parts people reach for when a 2N2222 is not in the drawer are the 2N3904 and the BC547. Both are NPN and both come in TO-92, but they use different pin orders, and neither handles as much current.

For the 100 mA example above, neither is a good swap. The BC547 would be at its absolute maximum current. The 2N3904 is rated for 200 mA, but its datasheet specifies saturation voltage only up to 50 mA, and guarantees an hFE of just 30 at 100 mA, so you would be relying on behaviour nobody promises. For loads under 50 mA both are fine, provided the pins go in the right holes.

If a board was designed for a PN2222A or 2N3904 (E-B-C) and you only have a P2N2222A or BC547 (C-B-E), turn the part around so the flat face points the other way. The outer two leads swap and the base stays in the middle. Then check it with a meter.

An NPN transistor looks like two diodes to a meter, with their anodes joined at the base. Set the meter to diode test.

1. Find the lead that conducts to both others with the red probe on it. That lead is the base, and each junction reads roughly 0.6 to 0.7 V. If the black probe is the one that finds a common lead, the part is PNP, not a 2N2222.
2. Reverse the probes: both junctions should read open (OL). Between the two outer leads you should read open in both directions.
3. The base-emitter reading is usually a few millivolts higher than the base-collector one. That is a hint, not proof. If your meter has an hFE socket, try the part both ways round: the orientation that shows the higher gain has collector and emitter in the right holes.

The transistor testing lesson works through the same method step by step, and continuity and diode test explains what the meter is doing in that mode. The Datasheet Guru will pull the same figures from any other transistor's datasheet.

A loose TO-92 sold as "2N2222", in a market or online, with no maker named could have either pin order, because the maker decides it. Buy from a seller that names the maker and the full part number (onsemi orders its TO-92 part as P2N2222AG, for example), or meter every part from an unmarked bag before it goes into a board. Our guide to buying electronic components in India compares ten online sellers and covers how to spot fakes.