Rev. 1.0 1/15

Copyright © 2015 by Silicon Laboratories

AN857

AN857

1. Introduction

In the switch-mode power supply world, capacitor-based charge pumps (or Q-pumps) generally aren’t useful for

heavy lifting, but work well in niche micropower applications where space is at a premium. They work best in

applications where the output voltage is an integer multiple of the input voltage, which are operating points that

result in peak efficiency. However, they can also shine when powered from a variable input like a battery,

particularly when quiescent battery drain is more important than heavy-load efficiency. This might be the case

when powering a microcontroller that spends most of its life sleeping.

Low-voltage microcontrollers, such as PIC24s or MSP430s, are generally powered from a regulated supply voltage

such as 2.5 V and, if clocked slowly, might draw as little as 25 µA or 50 µA. In standby mode, with only the RTC

clock running, the current can be vanishingly small, often less than a microamp. This is a good application for the

regulated Q-pump described here, which boosts a single alkaline or NiMH cell to 2.5 V using a two-stage Q-Pump.

The “wings” of a Q-pump are the flying capacitors that connect first to the input, then to the output. If the capacitor

is stacked on the input voltage, it forms a voltage doubler. In the case of a regulated charge pump with a fixed

output voltage, the voltage across the flying capacitor may differ significantly from the voltage across the output

filter capacitor. When you connect two capacitors that are initially charged to differing voltages, you get a spark, or

power dissipation, in the switches as the capacitors equalize in voltage. This is why a simple voltage doubler

typically exhibits better efficiency than a regulated Q-pump.

This regulated Q-pump has an on-demand oscillator, a feedback regulation loop made from an op-amp and

reference, and a two-stage pump circuit. It has two flying capacitors, C2 and C4. The first pump stage is driven

directly by the TS12011 comparator that forms the oscillator, while the second stage is driven by an inverter

powered from the output voltage of the first stage. The full-load efficiency varies from 70% to 40% over a 1 V to

2.5 V input range, which is comparable to what you might achieve with a linear regulator.