The Voltage Sets a Slope, Not a Position
The Voltage Sets a Slope, Not a Position
A voltage-controlled oscillator's control input does not choose a frequency the way a dial chooses a number. What the control voltage actually sets is a rate — an instantaneous frequency, in hertz per volt. But the thing the oscillator produces, the thing a listener hears or a receiver locks onto, is phase — and phase is defined as the integral of that rate over all elapsed time.
Read that in order: voltage sets frequency, and frequency is the derivative. Phase is the integral. This has a sharp consequence that holds even on a wire with no capacitor, no inductor, and no physical mass to damp anything: the VCO's phase cannot jump, no matter how instantaneously the control voltage jumps. Not because any filter or component is fighting to keep up — because of what integration itself is. An integral is a running sum of everything that came before. You can change the slope added to that sum at any instant you like, but the sum itself carries forward unbroken, because a step in the rate only bends the accumulated curve; it cannot erase what was already summed into it.
This is a deliberate mirror image of the direct digital synthesis piece already on this shelf. There, a stored register value changes at a clock edge with zero inertia, because a number has no physical history to carry — any settling behavior lives entirely downstream, in a real reconstruction filter. Here, nothing is manufactured downstream at all. Continuity falls straight out of the definition of an integral. The VCO's phase has no choice but to be continuous, even with a control voltage as sharp-edged as a digital signal — not an engineering achievement bolted on afterward, but a property of accumulation itself.
Two entirely separate places continuity can live in a signal chain, now both drawn on this shelf, and they are not the same mechanism wearing two masks.
