Diodes & Transistors

What to observe

1. Start with the PN diode and the supply at zero. Nothing flows, yet nothing is still: every carrier jitters. Look at the middle. Holes stop short on the left, electrons stop short on the right, and between them sits a strip of naked dopant ions, the depletion region. Its field is exactly what keeps the two crowds apart.
2. Push the supply up slowly. Below about 0.6 V almost nothing happens, then the current takes off. Watch why: forward bias shrinks the depletion region until the crowds can reach each other, and every crossing ends in a recombination flash. The knee is not a threshold anybody built in, it is an exponential caught in the act.
3. Now drive the supply negative. The depletion region swells, the carriers are pushed back to their contacts, and the current falls to a leak worth picoamps. One device, two completely different behaviours, from nothing but which way the barrier leans.
4. Raise the temperature with the diode forward biased. The forward drop falls by roughly 2 mV per kelvin and the reverse leak climbs fast, because heat generates electron and hole pairs on its own. This is why datasheets never quote 0.7 V without a temperature next to it.
5. Switch to the Zener and sweep the supply to −10 V. Past V_Z the reverse barrier gives way and the current comes back hard, but the voltage across the device barely moves. That flat wall in the I-V plot is the whole reason Zeners exist: a voltage reference made of a controlled failure.
6. Pick the LED and try each colour. Blue needs about 2.9 V, red about 1.9 V, and the photons only appear where electrons and holes actually meet. The forward drop is not a design choice, it is the band gap, and the band gap is the colour.
7. Watch a recombination closely, then switch between the silicon diode and the LED. It is the same event either way, an electron dropping into a hole, and the energy has to leave. Silicon hands it to the crystal asheat, the LED hands it over as a photon. That single difference, direct gap against indirect gap, is why one of them lights up and the other just gets warm.
8. Go to the NPN and set the base current to a few microamps. Count the streams: a trickle of holes goes into the base, a flood of electrons crosses it into the collector. The measured gain settles near β, and yet the base is doing nothing but letting the emitter's electrons over a hill it has lowered.
9. Keep raising I_B. The collector current climbs, then stops: V_CE has hit the bottom and the resistor now owns the current. That is saturation, the transistor used as a closed switch. Between cutoff and saturation lies the active region where it is an amplifier.
10. Switch to the NMOS with the gate at 0 V. No channel exists at all, just a depletion region under the oxide. Raise V_GS past V_thand an inversion layer of electrons condenses at the surface out of nowhere. No current ever flows into the gate: it is a capacitor plate, and that is why CMOS logic burns almost nothing while it holds a value.
11. With the channel on, raise the drain voltage. The channel tapers, thins to a point, and pinches off, and from there the current stops caring about V_DS. Note the carriers speed up where the channel is thinnest: the same current through a narrower pipe.