October 3, 2010:
Revised: v2.0

ZL2PD AA-cell Switchmode Boost Regulator

Results of some tests on a low cost AA-cell boost converter sold as a celphone charger in many supermarkets.


A couple of months ago, I spent some time building and testing a number of different boost switchmode power supplies (SMPS) while developing another microprocessor-based design. A 'boost' regulator is one of two basic categories of SMPS regulators, the other being the ‘buck’ regulator.

As the names suggest, the output voltage of a  buck regulator is less than the input voltage while a boost regulator’s output voltage is greater. For example, with a 9V battery, a buck regulator might efficiently reduce that voltage to 5V or 3.3V, to power a microprocessor, for example. By contrast, a boost regulator might increase that 9V voltage to, say, 90V to power a NIXIE display.

Of course, I’ve simplified things a bit here. In the world of SMPS, we can also find buck-boost, Cuk and a number of others. For example, a buck-boost supply can deliver output voltages above or below the input voltage.

Design Objectives

My original objective was to design an efficient boost regulator to give a reliable supply for a small microprocessor design. The microprocessor required a power supply between 4V and 5V at a current of no more than 20mA. I wanted to use either one or two AA batteries. The microprocessor I chose could not operate below 3.5V so using the microprocessor itself to drive the boost regulator was unfortunately not possible. I also wanted the regulator to use easy to find discrete parts to make it easy to duplicate.

Commercial Alternative?

As is typical of these things, no sooner than I had completed this design (which is documented elsewhere on this website) than I came across the unit pictured at the top of this page. It’s actually a charger for a cellular phone powered by a single AA battery. It claims to deliver enough power to the phone to provide up to 2 hours of talk time or 24 hours of standby operation. Since most phones require about 5V to charge, this design looked like it might be appropriate as an alternative to my discrete design.

There are some real advantages to using this type of product. They are made in quantities numbering in the tens of millions. With those sorts of volumes, they tend to be cheaper than the cost of the individual parts to you and me. The one I bought cost me about $US5, but I’ve since seen them at half this price. It is made from rugged anodized machined aluminium with a tough clear plastic cap. It also came complete with an AA battery and five different cables to suit a variety of popular celphones. Even if it didn’t work as advertised, I probably had $5 worth of connectors in that package.

But, it did work. The charger does not show any signs of operation until it is plugged into the phone, or some form of load. At that point, the LEDs on the charger light up, and it begins to charge the mobile.

However, as I discovered after drawing out the circuit, the unit is drawing current from the battery even when not operating. I measured the no-load no-LEDs glowing idle current at just under 2mA. In fact, with that no-load current, the battery will be discharged in less than 6 weeks. So, while this is not mentioned anywhere on the label, the moral of this measurement is : Do not leave the battery in this sort of charger when it’s not being used!

The Design

Here’s the schematic of the unit I bought. I suspect others follow a very similar design.

aa battery boost converter schematic

The heart of the unit is a three-pin chip marked IC1. This generates a 100kHz square wave to transistor Q1. Q1, D1 and L1 form the classical boost charger, delivering up to 5.8V out of the charger. I have not been able to accurately identify either of the SMD devices Q1 and IC1, but IC1 is probably one of those very low cost LED driver ICs. These have an internal PWM oscillator and are intended to drive white LEDs with only a single external component, the inductor. An example of this type of device is the Zetex ZXSC380.

These chips usually have an internal voltage comparator which turn off the internal oscillator when the battery voltage falls below 0.9V. In this charger, this voltage is actually the DC voltage generated by the boost converter. Schottky diode D1 rectifies the high voltage pulses generated by Q1 and L1, rectifying these into a DC voltage smoothed by C1, probably a 1uF or 2.2uF capacitor. D1 has a forward voltage drop of about 0.2V, and the LED driver chip (IC1) stops the boost charger operating when the battery voltage drops under load to its end of life voltage of about 1V. This stops the square wave drive to Q1, halting the boost SMPS regulator, and the LEDs will then turn off.

top side viewIn fact, due to the high current drain on the AA battery, when IC1 turns off, the battery voltage will immediately rise by 100 – 200mV. That will likely turn IC1 back on again, and, moments after, the falling battery voltage under load will turn the charger off once more. This possibly means the charger may appear to be a little erratic as the end of battery life is reached, and this functionality is confirmed by some of the review comments seen on the internet about this type of charger.

The two LEDs which indicate charger operation are connected in series with D2 and several passive components. When the boost SMPS (Q1, L1, D1) is operating, the voltage on the collector of Q1 rises during voltage peaks sufficiently to cause the LEDs to conduct. C2 ensures the DC no-load current of the charger is minimised while its impedance at 150kHz, much lower than the parallel 5k1 resistor, results in a high peak LED current (and bright LEDs) when the boost charger is operating.

While the LEDs will no longer be visible when IC1 is off, the AA battery will actually continue to be discharged into the load (via L1 and D1). Even so, it is unlikely that any cellular phone battery which is left connected will still be charging at such low output voltages.

Estimating Performance

A typical AA alkaline battery is rated at about 2000mAh with a typical discharge current of about 50mA. At the higher load currents of this charger, AA alkaline battery capacity falls to about 1300mAh according to the specifications of well-known battery manufacturers.

Important: This battery rating is at room temperature! If it is cold, the AA alkaline cell capacity may fall to as little as 300mAh at 0C!! That’s a fraction of what you expect. If you are caught in the cold and you need to recharge your celphone this way, warm the AA battery next to your body first, then charge the phone.
But here’s a curious thing. It’s best to store alkaline batteries in a cold place. Storage life will be doubled if they are stored at 0C instead of 40C. While I don’t recommend storing batteries in the fridge, keeping them on a cool shelf is clearly a good idea.

underside viewAssuming the charger is as efficient at the packaging claims, over 80%, then the charger might deliver as much as 1000mAh to the cellular phone. Then, allowing for a further 80% efficient power transfer between the charger connector of the cellular phone and the internal celphone battery, it might finally actually receive up to 800mAh. A typical Lithium-Ion battery inside a typical small celphone (e.g. Nokia BL-4C Li-Ion battery) is rated at close to 800mAh. Since this battery powers the phone for up to 2 – 4 hours of talk time, and for up to 48 hours of standby time, the AA charger manufacturer’s claims appear reasonable.

However, elsewhere on the packaging, it mentioned a maximum cellular phone battery recharge of 240mAh from a single alkaline AA cell. That suggests only a bit over 10% of the alkaline cell capacity finally reaches the phone battery. Not particularly good, but it should be enough for perhaps a few short calls.

The package also claims the AA charger’s maximum output current reaches 80mA, while other manufacturers of similar devices claim output currents between 200 to 280mA. This implies an average AA battery current of well over 1/2A! That’s a fairly impressive current being handled by such a little unassuming innocent AA battery and such a tiny circuit.

The more likely result is that the AA battery current will very rapidly fall from 400mA to 50mA (or less) within an hour or so. It’s unlikely that an 800mAh celphone battery would be fully recharged in that time, confirming that the energy transferred is limited, as outlined above, and enough for a few brief emergency calls and perhaps a few texts.

Yet another package I saw with what appeared to be an identical unit stated:

Input voltage: DC 1.5V
Charging time: At 280mA - 10min
Output voltage: 6.2V + 0.1V/-0.5V
Output current: 196mA – 280mA, falling to 100mA, more than 30min.

These numbers seem to confirm my initial calculations and overall conclusions.

Testing the Boost Converter

I decided to do some measurements on the unit I had purchased to check output voltage, current capabilities, and charger efficiency. I used a series of resistors to load the charger with a range of output currents.
Without any load connected, I measured 11.5mA current from the 1.5V AA battery.

Here are the other results:

As the output current increased by using lower value resistors as the load, the internal resistance of the 1.5V AA battery resulted in the battery (input) voltage falling slightly, from 1.6V without any load to 1.4V under heavy load.

Vin (V)
Iin (mA)
Vout (V) Iout (mA) Load (Ohms)
Pin (mW)
Pout (mW)
Efficiency (%)
No load

With the 56 ohm load, the input current was measured to be over 500mA(!!), but efficiency was still being maintained at over 60%. At these battery currents, good low-resistance connections were essential to reach these efficiencies. I suspect the actual efficiency might be higher because I was dropping a little power across the input current test meter during these measurements.


Is this charger of any use as a boost supply to power a microprocessor from a single AA battery?

Most microprocessors have a maximum supply voltage rating of 6V. These measurements suggest this to be a workable solution, although efficiencies are a little lower than the circuit I designed earlier.

For those wanting to limit the voltage on their circuit to something closer to 5V, connecting the output of this regulator via a series silicon diode (1N4148 etc) to the circuit will drop the voltage closer to 5V for load currents less than 50mA. A 5.6V 400mW protection zener diode fitted across the supply terminals of the microprocessor will probably also be fine. So, yes, this circuit appears to provide a useful solution. Of course, efficiency will be reduced with the addition of these diodes.

Is this solution practical?

The answer to this depends on the circuit you are powering. If your circuit only requires power for a few seconds, or perhaps even tens of seconds, then this arrangement will probably work well.

However, if power is required for longer periods, then the AA battery will quickly become discharged. For such cases, a more conventional buck charger with a higher voltage battery of greater capacity would probably be a better choice.


Want to go back to the main page? Click here to return directly.