decoupling & bypass capacitors

A decoupling capacitor is the tiny local energy reservoir parked right beside every chip's power pin, and it is the single most common component on any circuit board. Here's the problem it solves: a digital chip's current demand is not steady — every clock edge, millions of transistors switch at once and gulp a sudden burst of current. The power supply is too far away (through inches of inductive copper trace) to react in nanoseconds, so the chip's voltage would momentarily sag and create noise. The capacitor sits inches away holding a small charge ready to dump instantly, like a bucket of water kept right next to a thirsty worker so they don't have to run to the well every time.

Engineers usually place several in parallel — a bigger one (e.g. 10 µF) for slower, bulk current swings and a small one (0.1 µF, the classic) close to the pin for the fast, high-frequency transients, because each value is most effective over a particular frequency band. The placement is half the magic: even a couple of millimetres of extra trace adds enough inductance to ruin the cap's high-frequency punch, which is why layout guides scream 'put the 0.1 µF as close to the pin as physically possible.' Done well, decoupling keeps the power rail quiet AND stops those current spikes from radiating out as EMI; done badly, it's a top cause of mysterious glitches and failed emissions tests.