Flip the lever, press the button, and the lamp locks in the lever's value. A ready-made memory circuit you can open in the TorchAnvil simulator.
Set the lever. Press Latch. The lamp snaps to match the lever and stays there, even if you flip the lever afterward. Press Latch again and it re-samples whatever the lever is doing right now.
That's a D flip-flop — arguably the memory cell of the modern world. Most RAM, most registers, most "hold this signal until I say otherwise" circuits you've heard of are built from these.
The D flip-flop has two inputs:
On every rising edge of CLK (the moment it goes from low to high), Q samples whatever D is at that instant and holds it until the next rising edge. In between edges, D can wiggle all it wants — Q doesn't care.
| D (at CLK rise) | Q (next) |
|---|---|
| 0 | 0 |
| 1 | 1 |
(No more complicated than that. The magic is when it samples.)
Compare this to the SR flip-flop. SR needs you to drive Set or
Reset explicitly — two wires, two buttons, careful not to press both.
D gives you one data line and one clock. Line up a bunch of D flip-flops
sharing a clock and you get a register: a synchronized snapshot of
multiple bits, all captured on the same edge. Every CPU is full of them.