A valve for electricity
Up to now every component you have met is a two-way street. A resistor does not care which way the current flows through it; a divider divides the same whether the supply sits on top or bottom. The diode is the first part that breaks that symmetry. It is a one-way valve: it lets current pass freely in one direction and slams shut against it in the other, exactly like the check valve in a bicycle pump that lets air out but not back in.
A diode has two leads, the anode and the cathode. When the anode is pushed more positive than the cathode, the valve opens and current flows anode-to-cathode — we call this forward bias. Flip the pressure the other way, cathode more positive, and the valve stays shut: that is reverse bias, and almost no current gets through. One device, two completely different personalities depending on which way you push.
Inside the valve: the PN junction
The one-way trick lives at a boundary inside a sliver of silicon, a semiconductor — a material that is neither a good conductor like copper nor a stubborn insulator like glass, but somewhere in between, and whose conductivity we can deliberately tune. We tune it by doping: lacing one side with impurities that leave it with spare mobile electrons (the n-side, n for negative carriers) and the other side with impurities that leave it short of electrons, i.e. full of mobile 'holes' (the p-side). Where these two regions meet is the PN junction, and that meeting point is the whole story.
Right at the junction, spare electrons from the n-side drift across and fill holes on the p-side. That leaves a thin no-man's-land swept clean of mobile carriers — the depletion region — and the exposed fixed charges there build a small built-in electric field, like a spring-loaded gate that has settled closed. To get current flowing, you have to overcome that built-in barrier.
Now the asymmetry makes sense. Push with forward bias — anode (p-side) positive — and you fight the built-in field, narrow the depletion region, and once you push hard enough carriers come flooding across: the gate swings open. Push the other way with reverse bias and you only widen the depletion region and pull the gate harder shut. Same junction, but the geometry only collapses for one polarity of push. That is the physical root of the one-way valve.
The 0.7 V toll and the bent curve
Overcoming that built-in barrier costs a roughly fixed amount of voltage. For ordinary silicon diodes it takes about 0.6 to 0.7 V of forward push before meaningful current flows, and once it is conducting that drop barely moves even as the current climbs. This is the famous forward voltage drop — think of it as a toll booth at the entrance to the valve: pay roughly 0.7 V and you may pass; the harder you press beyond that, the more current pours through while the toll itself hardly changes.
If you actually plot current against forward voltage — the I-V curve — you do not get the straight line of a resistor. You get a flat floor that hugs near-zero up to about half a volt, then a sharp 'knee' where the curve sweeps almost vertically upward. The rise is exponential: each extra ~60 mV multiplies the current by about ten. That is why the drop feels 'fixed' — to go from 1 mA to 10 mA you only add ~60 mV, so over any normal operating range the diode sits stubbornly near its knee voltage no matter how much current you ask of it.
Three models you can compute with
That exponential curve is honest but awkward to do arithmetic with, so engineers reach for a simplified diode model sized to the job. There are three rungs of detail. The ideal model treats the diode as a perfect switch: a dead short when forward, an open circuit when reverse — good for a quick sanity check or for digital logic. The constant-voltage-drop model adds the toll: when conducting it behaves like a small 0.7 V battery, otherwise open. This is the workhorse for everyday design. The most refined model adds a small dynamic resistance r_d in series with that 0.7 V, capturing the slight tilt of the curve above the knee.
Model Reverse bias V < ~0.6 V V > ~0.7 V (on) ---------------------------------------------------------------------- Ideal (switch) open circuit open circuit short, 0 V drop Constant-drop open circuit open circuit fixed ~0.7 V drop 0.7 V + r_d open + leakage nearly open 0.7 V + I times r_d
- Take a 9 V supply feeding a 1 kohm resistor in series with a forward silicon diode. First decide the diode is forward biased — the supply pushes anode-to-cathode, so yes, it conducts.
- Ideal model: the diode is a dead short, so all 9 V lands on the resistor and I = 9 / 1000 = 9 mA. Fast, but it ignores the toll.
- Constant-drop model: the diode eats 0.7 V, leaving 9 - 0.7 = 8.3 V across the resistor. By Kirchhoff's voltage law, I = (9 - 0.7) / 1000 = 8.3 mA. This is the number you would actually design to.
- Notice the two answers bracket reality and are close — because the 1 kohm resistor, not the diode, dominates. That is the design lesson: let a resistor set the current and the exact diode drop stops mattering much, just as it did when you sized an LED's current-limiting resistor earlier.
Reverse bias, breakdown, and cousins
Reverse bias is where the valve earns its keep, but the seal is not perfect. Under reverse bias a tiny leakage current still seeps through — typically nanoamps to a few microamps, and it grows with temperature. For almost every circuit that is negligible, so we say the reverse diode is 'open'. But honest modeling remembers it is a very high resistance, not a true infinity.
Push the reverse voltage too far, though, and the dam bursts. Beyond a part's reverse breakdown voltage the junction suddenly conducts hard backwards, and an ordinary diode not built for it can be destroyed. Every diode datasheet states a maximum reverse voltage you must respect. That same breakdown, tamed and made precise, is exactly what a Zener diode exploits to pin a steady voltage — the subject of a later guide in this rung.
Not all diodes are the same silicon valve. A Schottky diode uses a metal-to-semiconductor junction instead of a PN one, giving it a much smaller forward drop (~0.2 to 0.4 V) and far faster switching — invaluable where the 0.7 V toll wastes precious headroom or where the diode must keep up with high-frequency signals. With this one-way behavior in hand you are ready for its first great application: ganging diodes to turn a back-and-forth AC wave into one-directional DC, which is exactly what the rectifier in the next guide does.