For a standard 2:1 step-down switched capacitor converter (charge pump) converting a 5.0V DC input to a nominal 2.5V output at a 100mA load, the actual converted output voltage is 2.35V. This assumes a typical equivalent output resistance (Rout) of 1.5Ω, which is standard for basic CMOS charge pumps like the Microchip TC7660 or TI TPS60500 when using 1μF flying capacitors. The formula used is Vout(actual) = (Vin × Ratio) - (Iload × Rout). Substituting our exact values: Vout = (5.0V × 0.5) - (0.1A × 1.5Ω) = 2.5V - 0.15V = 2.35V.

The Core Conversion Math and Load Droop

Unlike inductor-based buck converters, a switched capacitor converter does not use a magnetic field to store energy. Instead, it transfers discrete packets of charge between a 'flying' capacitor and an output reservoir capacitor using internal MOSFET switches. Because of this, the 'conversion' is highly dependent on the physical limitations of the capacitors and the silicon switches.

The three assumptions that fix this answer are:

  • Flying Capacitor Value & ESR: A 1μF X7R ceramic capacitor with low Equivalent Series Resistance (ESR). If you use a high-ESR Y5V capacitor, Rout increases and your voltage drops.
  • Switching Frequency: Typically 10kHz to 1MHz. Higher frequencies reduce Rout by transferring charge more often, but increase switching losses.
  • Load Current: The 100mA draw dictates the voltage droop across the internal impedance.

Here is how the actual converted voltage shifts across a ±20% range of our baseline 100mA load, assuming the 1.5Ω Rout remains constant:

Load Current Ideal Vout (5V × 0.5) Voltage Droop (I × 1.5Ω) Actual Vout
80 mA (-20%) 2.50V -0.12V 2.38V
90 mA (-10%) 2.50V -0.135V 2.365V
100 mA (Base) 2.50V -0.15V 2.35V
110 mA (+10%) 2.50V -0.165V 2.335V
120 mA (+20%) 2.50V -0.18V 2.32V

Real-World Switched Capacitor IC Specifications

Theoretical ratios only get you so far on the bench. When selecting a switched capacitor converter, you must look at the silicon's maximum current capability and its typical output impedance. Below is a data-dense comparison of common charge pump topologies used in modern embedded designs.

Manufacturer / Part Topology Vin Range Vout Ratio Max Iout Typical Rout
TI TPS60500 Step-Down 1.8V - 3.3V 1.5x, 2x 250 mA ~1.2 Ω
Analog Devices LTC3260 Inverting / Step-Up 4.5V - 36V -1x, 2x 100 mA ~15 Ω
Microchip TC7660 Inverting 1.5V - 10V -1x 20 mA ~10 Ω
TI TPS60400 Inverting 1.6V - 5.5V -1x 60 mA ~4.5 Ω
TI LM2776 Inverting 2.7V - 5.5V -1x 200 mA ~2.0 Ω

Source data aggregated from Texas Instruments Charge Pump Portfolio and manufacturer datasheets.

High-Voltage Mains (120V/230V/3-Phase) and Meaningless Conversions

CRITICAL TOPOLOGY BOUNDARY: Switched capacitor converters are strictly low-voltage DC-DC topologies. Applying 120VAC, 230VAC, or 3-phase AC mains to a charge pump IC will result in immediate, catastrophic dielectric breakdown of the internal MOSFETs and flying capacitors.

Because these are DC-DC devices, the concept of how the answer 'shifts' for 120V vs 230V vs 3-phase power is fundamentally inapplicable. If you need to step down 120VAC or 230VAC to a low DC voltage, you must use an AC-DC flyback converter, a buck-boost with a rectifier stage, or a traditional 50/60Hz iron-core transformer followed by a linear regulator. The maximum input voltage for specialized high-voltage charge pumps (like the LTC3260) tops out at 36V DC.

When is the conversion meaningless?

  1. When Power Factor (PF) is queried: PF is an AC sinusoidal concept. A switched capacitor converter draws pulsed DC current from the source. Therefore, asking for the PF of a charge pump is meaningless; the relevant metric is efficiency and input ripple current.
  2. When Load Exceeds Charge Transfer Capacity: If you attempt to pull 150mA from a Microchip TC7660 (rated for 20mA), the output voltage will collapse to near zero. The nominal 2:1 or -1x ratio becomes mathematically meaningless because the pump cannot physically move enough electrons per switching cycle to sustain the voltage against the load.
  3. When Flying Cap ESR is Too High: If you substitute a 1μF electrolytic capacitor (high ESR) for the required 1μF MLCC ceramic (low ESR), the internal voltage drop during the charge transfer phase will destroy your conversion ratio, often yielding 30-40% less voltage than calculated.

FAQ: Design Edge Cases and Component Selection

Q: Can I parallel two switched capacitor converters to double the output current?
A: Generally, no. Unlike inductor-based buck converters with power-good or clock-sync pins, basic charge pumps lack phase-interleaving circuitry. Paralleling them often results in one IC dominating the load due to slight mismatches in internal Rout, leading to thermal runaway in the dominant chip. If you need more current, select a single IC with a lower Rout or switch to a synchronous buck converter.

Q: Why does my 5V to -5V inverting charge pump output only -3.8V on the bench?
A: You are likely seeing the combined voltage drop of the internal MOSFET RDS(on) and the ESR of your flying capacitor. Furthermore, if you are measuring with a standard multimeter, the high-frequency switching ripple (often 50mV to 200mV peak-to-peak) can confuse the ADC in cheaper meters. Measure the output using an oscilloscope with a tip-and-barrel probe to see the true DC baseline plus the AC ripple component.

Q: Do I need an input capacitor for a charge pump?
A: Yes. While the flying capacitor handles the energy transfer, the charge pump draws sharp, high-di/dt current pulses from the input supply during the charging phase. A 1μF to 10μF low-ESR ceramic input capacitor placed as close to the Vin pin as possible is mandatory to prevent voltage sags that would otherwise trigger the IC's undervoltage lockout (UVLO) or inject noise back into your main power rail.

For deeper theoretical analysis on charge transfer mechanics, refer to the All About Circuits guide on charge pump voltage converters.