Pnp Transistor

A p-n-p transistor is a bipolar junction transistor made of a thin n-type base between a p-type emitter and a p-type collector, so that its main current is carried by holes. It is the mirror image of the n-p-n: every voltage and current direction is reversed, and the transistor turns on when the base is pulled a little below the emitter instead of pushed above it. Its circuit symbol has the arrow on the emitter pointing inward, toward the base.

Holes do the carrying, and the bias polarity flips accordingly. In a p-n-p device the emitter is heavily doped p-type and injects holes, the absence of electrons that behave like positive carriers, into the narrow n-type base. This happens when the base is about 0.6 to 0.7 volts lower than the emitter in silicon, which makes the emitter the most positive terminal. The collector sits at a lower voltage than the base, so the base-collector junction is reverse-biased and sweeps the holes through. Conventional current therefore enters at the emitter and leaves through the collector and the base, the reverse of the n-p-n. Because the emitter must be above the base, the supply connects to the emitter and the load goes on the collector side toward ground.

The p-n-p is the natural partner for switching a load from the positive supply, the high-side arrangement. With the emitter tied to the supply rail and the load between the collector and ground, pulling the base low turns the load on, and letting the base rise to the supply voltage turns it off. This keeps the load grounded on one side, which is often wanted for safety and for sensing. The catch is that a microcontroller pin at 3.3 volts cannot switch off a p-n-p whose emitter sits at 12 volts, because the base would still be 8.7 volts below the emitter and the transistor would stay on, so a level-shifting n-p-n or resistor network is added. Mixing up which side the transistor sits on is the most common error with the p-n-p.

Slower carriers make the p-n-p the weaker partner, which is why it is rarely used alone for speed. Holes in silicon move at roughly one third the speed of electrons, so for the same geometry a p-n-p has lower gain, lower bandwidth and, in integrated circuits, a larger area. Process designers compensated by making lateral p-n-p transistors with different geometries and by using the n-p-n wherever possible. This is a reason that digital designers, and later CMOS logic, leaned on the electron-based device for the heavy pulling down of signals. The gap is a handicap rather than a barrier, and many bipolar processes include respectable p-n-p devices for analogue work.

Complementary pairs combine the two polarities into push-pull stages, which is the p-n-p's classic role. In a push-pull audio output stage an n-p-n handles the positive half of the waveform by pushing current into the speaker, while a p-n-p handles the negative half by pulling current out. Matched parts are sold as complementary pairs, such as the small-signal 2N3904 (n-p-n) with the 2N3906 (p-n-p), the BC547 with the BC557, and the power types 2N3055 and MJ2955. Because both devices are driven from the same signal and share the load, the stage needs no output transformer and wastes little idle current, a design that took hold in transistor audio and operational-amplifier output stages from the 1960s onward. The ideal match is imperfect, since the two polarities differ slightly in gain and capacitance, and designers use feedback to hide the difference.

The p-n-p has a historical role as well, though the story is more mixed than a single date. Germanium made a convenient p-n-p because alloying indium into an n-type germanium wafer was an easy way to form both junctions, and many commercial germanium transistors of the 1950s, including the small audio and radio types sold in huge numbers, were p-n-p. By contrast, the first grown-junction transistor of 1950, from Gordon Teal and Morgan Sparks at Bell Labs, was an n-p-n, and silicon soon favoured the n-p-n because of electron mobility. In today's chips the p-channel MOS transistor takes the p-n-p's place, with holes in the channel pulling the output up toward the supply in every CMOS gate, while discrete p-n-p transistors survive in power supplies, battery protection and audio amplifiers.

Read the p-n-p transistor as the hole-carried mirror image of the n-p-n: pull the base down to open the path from emitter to collector, and put it where a load must hang from the positive rail or where an amplifier needs a partner to pull current out.