DC Voltage Doubler Circuits: How Switched-Capacitor Charge Pumps Work

A DC voltage doubler is a switched-capacitor circuit that stores energy in a "flying" capacitor during one clock phase and transfers it in series with the input supply during the next phase to yield an output roughly equal to twice the input. In a real circuit, this changes a low-voltage single-rail DC source into a higher-voltage rail without the electromagnetic interference (EMI), physical bulk, or cost of an inductor or transformer. While they cannot deliver the massive current of an inductive boost converter, they are the undisputed champions of low-power, low-noise voltage translation on the bench and in commercial silicon.

The Core Mechanism: How Switched Capacitors Move Charge

To understand a switched-capacitor charge pump, you have to look at the internal oscillator and the MOSFET switches that route the current. The circuit operates in two distinct, alternating phases, typically switching between 10 kHz and 100 kHz depending on the IC.

1. Phase 1 (Charge): Internal switches connect the flying capacitor ($C_{FLY}$) directly across the input voltage ($V_{IN}$). The capacitor charges to $V_{IN}$.
2. Phase 2 (Transfer): The switches flip. The flying capacitor is disconnected from ground and placed in series with the input supply. The positive side of $C_{FLY}$ is now tied to $V_{IN}$, and its negative side is routed to the output reservoir capacitor ($C_{RES}$). The output sees $V_{IN} + V_{CFLY}$, which equals $2 \times V_{IN}$.

Think of it like two water buckets. You fill one bucket from the main tap (input voltage), then physically lift it and stack it on top of a second bucket that is already at the tap's height, pouring them together into a higher reservoir. You aren't creating new water; you are just changing its potential energy by stacking it.

A Worked Numeric Example

Let's look at a modern, diode-free doubler like the Microchip TC7660. We feed it 5.0V DC. In an ideal world, the output is 10.0V. But real silicon has internal switch resistance ($R_{SW}$) and the capacitors have Equivalent Series Resistance (ESR).

If we draw a 20 mA load from the output, the voltage drops according to the output impedance ($R_{OUT}$). For the TC7660, $R_{OUT}$ is typically around 12Ω under these conditions. Using Ohm's Law:

V_{DROP} = I_{LOAD} \times R_{OUT} = 0.020A \times 12\Omega = 0.24V

Because the voltage drop happens across the charge transfer path, the actual output is roughly $2 \times V_{IN} - V_{DROP}$, yielding 9.76V. If you were using an older ICL7660 with external silicon diodes instead of internal MOSFETs, you would lose an additional ~1.2V to diode forward voltage drops, landing you at a disappointing 8.5V.

Where You Meet DC Voltage Doublers in Practice

You rarely see discrete voltage doublers built from scratch on a modern PCB, but the topology is hiding inside dozens of common integrated circuits:

• RS-232 Transceivers: The legendary MAX232 chip uses an internal charge pump to generate the ±10V rails required for serial communication from a single 5V supply.
• High-Side Gate Drivers: To fully turn on an N-channel MOSFET on the high side of a load, the gate must be driven 10V-15V above the source pin. A bootstrap charge pump provides this elevated gate voltage.
• LED Bias and LCD Contrast: Driving strings of white LEDs or generating the negative bias voltage for LCD contrast often relies on multi-stage charge pumps (Dickson charge pumps) to step 3.3V up to 15V or 20V at just a few milliamps.
• EEPROM Programming: Writing data to onboard flash or EEPROM requires a brief ~12V pulse, which is generated internally by a charge pump so the microcontroller doesn't need an external 12V rail.

Bench Scenario: When a Doubler Fails Under Load

Charge Pump vs. Inductive Boost: Choosing the Right Topology

People frequently confuse capacitive DC voltage doublers with inductive boost converters because both step up DC voltage. However, their underlying physics dictate entirely different use cases. Refer to the Analog Devices MT-073 tutorial for a deeper look at charge transfer limitations.

Common Confusions and Design Traps

Confusion 1: The AC Villard Multiplier vs. DC Charge Pumps
Many beginners search for a "voltage doubler" and find the Villard or Greinacher circuits, which use diodes and capacitors. Those circuits require an AC input to function, relying on the alternating polarity to push charge up the ladder. A true DC voltage doubler requires active switching (MOSFETs or a mechanical clock) to simulate that AC polarity reversal internally.

Trap 1: Ignoring DC Bias in Ceramic Capacitors
When selecting the flying capacitor, you might grab a cheap 10µF X7R ceramic capacitor to keep ESR low. However, Class II ceramics suffer from severe DC bias capacitance drop. A "10µF" capacitor with 5V across it might actually only store 4µF. This starves the charge pump, increasing output ripple. For the flying cap in a 5V doubler, a low-ESR tantalum or a high-voltage-rated (e.g., 16V or 25V) ceramic is mandatory to maintain true capacitance under bias.

Trap 2: Assuming Exact 2x Multiplication
Never design a downstream circuit that requires exactly 10.0V from a 5.0V doubler. The output will always be $2 \times V_{IN}$ minus the diode drops (if external) or the $I \times R$ losses of the internal switches. Always design your load to tolerate a 10% to 15% droop from the ideal doubled voltage.