Voltage doubler

A voltage doubler is a diode–capacitor circuit whose DC output is about twice the peak of the AC input. Cascade a Clamper circuit with a Peak rectifier: the clamper offsets the waveform so its excursion reaches $2V_p$, and the peak rectifier captures that doubled peak.

How the two stages combine

Stage 1 — clamper ($C_1$ + $D_1$). The clamper shifts the AC input so that its negative peak sits at 0 V. An input that swung from $-V_p$ to $+V_p$ (peak-to-peak $2V_p$) slides up by $V_p$, so it now swings from 0 to $+2V_p$. Same waveform shape, just a shifted DC level, so its peak is now $2V_p$ instead of $V_p$.

Stage 2 — peak rectifier ($D_2$ + $C_2$). Fed with the clamped waveform whose peak is $2V_p$, the peak rectifier charges $C_2$ up to that peak and holds it there. The output is therefore a DC voltage of magnitude $\approx 2V_p$ (minus the two diode drops, so really $2V_p-2V_D$ with the Constant-voltage-drop model).

A clamper stage followed by a peak rectifier; output magnitude ≈ $2V_p$.

On polarity: with the diode orientation drawn here the output settles at about $-2V_p$ (minus the diode drops); reversing both diodes gives the more common $+2V_p$ version. Magnitude is twice the input peak either way, only the sign depends on diode direction.

Why you would want one

A voltage doubler gives a DC voltage higher than the rectified AC peak available to you, without a bigger transformer. A common use: generating the higher bias rail a small LCD backlight needs from a 5 V USB rail. Stacking more clamper/peak-rectifier stages gives a voltage multiplier (tripler, quadrupler, …), each stage adding another $V_p$ to the output; the Cockcroft–Walton ladder used for high-voltage generation is the same trick repeated. The price of multiplying is poor current capability: each stage’s capacitors must supply the load through a long chain, so doublers and multipliers are for low-current, high-voltage needs, not power delivery.