If you have ever watched a learner’s two-wheeled Arduino robot crawl forward for eight minutes, shudder, and suddenly reboot every time the ultrasonic sensor triggers a turn, you have encountered the PP3 9V battery problem. The battery is not defective. It is simply being asked to do a job physics never intended it to do.

A standard rectangular 9V battery dies rapidly in school robotics projects because of three compounding factors: tiny internal energy capacity, high internal resistance that causes catastrophic voltage sag under motor loads, and the thermal waste of the Arduino’s on-board linear voltage regulator. Understanding the arithmetic behind these failures makes it easy to choose cheaper, more durable power architectures for the classroom.

What Is Inside a 9V Battery?

Peel the metal casing off a standard alkaline PP3 9V battery and you will find six tiny 1.5V cylindrical cells (often designated AAAA) wired in series. Because these cells are miniature, their total chemical capacity is remarkably low—typically between 400 milliamp-hours (mAh) and 550mAh when discharged very slowly.

Crucially, battery manufacturers rate that capacity under ideal conditions: a continuous load of only 15 to 25 milliamps (mA), typical of a smoke detector or a digital multimeter. When placed in an Arduino project driving two yellow TT geared DC motors (pulling 200mA to 400mA each) alongside a microcontroller, an ultrasonic sensor, and an H-bridge motor driver, the current draw easily exceeds 500mA to 800mA. Under that heavy load, the usable capacity of a 9V cell collapses to a fraction of its rated value.

The Math of Voltage Sag and Brownouts

Every battery has internal resistance ($R_{int}$). For a fresh alkaline 9V battery, internal resistance starts between $1.5\,\Omega$ and $2.0\,\Omega$, and climbs steeply as the battery discharges.

When a current ($I$) is drawn, the voltage delivered at the battery terminals ($V_{terminal}$) drops according to the formula:

$$V_{terminal} = V_{open} - (I \times R_{int})$$

Consider what happens when an Arduino rover starts its motors from a standstill. The momentary stall current of two small DC motors can easily hit 1.0A:

• Open-circuit voltage: 9.0V
• Voltage drop across internal resistance: $1.0\text{A} \times 2.0\,\Omega = 2.0\text{V}$
• Instantaneous terminal voltage: $9.0\text{V} - 2.0\text{V} = 7.0\text{V}$

As the battery discharges slightly over ten minutes and internal resistance climbs to $3.5\,\Omega$, that same 1.0A surge drops terminal voltage to $5.5\text{V}$. The Arduino Uno and Nano boards use on-board linear regulators (such as the AMS1117 or LM1117) to produce the stable 5V rail required by the ATmega328P microcontroller. These regulators require a "dropout voltage" overhead of around 1.0V to 1.2V. When the input voltage dips below approximately 6.2V, the 5V rail collapses, triggering the microcontroller’s brownout reset circuit. The robot resets, the motors stop, the load disappears, the voltage bounces back to 8V, the Arduino reboots, and the cycle repeats endlessly.

The Linear Regulator Penalty

Even before the battery sags into brownout territory, running an Arduino from a 9V battery through the VIN pin or barrel jack wastes nearly half the battery’s energy as pure heat.

A linear regulator steps down voltage by acting as a variable resistor. It dissipates the excess voltage ($V_{in} - V_{out}$) multiplied by the current flowing through it:

$$P_{loss} = (V_{in} - 5\text{V}) \times I$$

When stepping 9V down to 5V, you are discarding 4V out of every 9V supplied—an electrical efficiency of only 55%. The remaining 44% of your battery’s expensive energy is burned off as heat on the tiny surface-mount regulator chip.

Power Sources Compared for School Robotics

What to Use Instead in the Classroom

For primary and high school classrooms building projects under the gazetted Coding and Robotics curriculum, moving away from 9V snaps saves significant budget and eliminates classroom troubleshooting time.

1. The 4xAA or 6xAA Battery Holder

Standard AA cells have roughly four to five times the energy capacity of a 9V battery and significantly lower internal resistance. A 4xAA pack provides 6.0V (alkaline) or 4.8V (NiMH), which can power DC motor drivers directly while feeding a regulated logic rail. A 6xAA pack (7.2V–9.0V) provides sufficient overhead for the Arduino’s VIN pin without dipping into brownouts when motors turn on.

2. Dual-Rail Power (Isolating Motors from Microcontrollers)

The cleanest engineering practice is to separate logic power from motor power. Power the Arduino microcontroller via its USB port using an inexpensive, rechargeable single-cell 5V power bank. Power the motors and H-bridge driver from a dedicated 4xAA battery pack. Connect the ground (GND) lines of both sources together. When the motors stall or surge, the voltage drop occurs strictly on the motor battery, leaving the microcontroller’s 5V rail rock-solid.

3. 2S 18650 Li-ion Packs with a Step-Down Buck Converter

For secondary school robotics teams competing in open-hardware challenges like WRO RoboMission, standard 18650 lithium-ion cells in a 2S configuration (7.4V nominal) offer the highest power density. Instead of feeding 7.4V into the Arduino’s inefficient linear regulator, wire the battery through an inexpensive switching step-down converter (buck converter) set to 5.0V. Switching regulators operate at 85% to 95% efficiency and do not burn off excess voltage as heat.

Ensure that any 18650 battery holder used in schools includes built-in battery protection circuits (BMS) to prevent over-discharge, short-circuits, and reverse polarity hazards during learner handling.

The Economics of the Switch

In South Africa, a single alkaline 9V battery costs between R40 and R65. In a class of 30 learners working in pairs across a ten-week term, replacing dead 9V batteries every three practical periods easily costs a school upwards of R2,500 per term in recurring consumables. Equipping robot chassis with rechargeable NiMH AA cells or protected lithium-ion packs from the Sheen Robotics component catalogue pays for itself within two project cycles while eliminating the frustration of mid-demonstration robot reboots.