Here is a fact that quietly sits underneath every schematic, datasheet and PCB you will ever work on: the current arrows point the wrong way. In an ordinary copper wire, the particles actually moving are electrons, and they travel in the opposite direction to the current drawn on the diagram.
This is not a mistake anybody made recently. It is a coin flip that Benjamin Franklin lost in the 1750s, and the entire discipline of electronics has been politely living with the result ever since.
Franklin had to pick a direction, and had no way to check
In the 1740s and 50s, Franklin was working on what became his single-fluid theory of electricity. The prevailing view held that there were two distinct electrical fluids. Franklin argued for one: a single subtle fluid that objects could have too much of or too little of. Rub a glass rod, and something moves from one body to the other. One ends up with a surplus, one with a deficit.
He needed labels, so he coined them — positive for the surplus, negative for the deficit. Those two words are Franklin's, and they are still the ones on every battery terminal and every pin in a connector datasheet.
But labelling the surplus "positive" also implied a direction of flow: from the body that has extra toward the body that lacks it, positive to negative. Franklin had no instrument that could tell him which way charge physically moved, and no concept of a charge carrier at all. It was a genuine 50/50 call, and he took it.
The electron showed up 140 years later, going the other way
In 1897, J.J. Thomson identified the electron at the Cavendish Laboratory, and by the early twentieth century it was clear that conduction in a metal is electrons drifting through a lattice of fixed positive ions. Electrons carry negative charge. They are attracted toward the positive terminal.
Which means the physical carriers in a wire move from the negative terminal to the positive one — precisely backwards from Franklin's convention, now called conventional current.
So why did nobody change it?
Because nothing breaks. This is the part that surprises people.
Charge moving one way is mathematically indistinguishable from opposite charge moving the other way. Ohm's law, Kirchhoff's current and voltage laws, power calculations, Thevenin equivalents, AC analysis — all of it produces identical answers whichever convention you adopt, as long as you are consistent. Changing the arrows would have invalidated a century of textbooks, drawings and component symbols to gain exactly zero predictive power.
Franklin's convention is also not as wrong as it first sounds. In a semiconductor, p-type material conducts by holes, which behave as genuine mobile positive charge carriers. In electrolytes, plasmas and ion channels, positive ions really do flow in the conventional direction. Metals are the case where the sign flips — they just happen to be the case most of us picture.
The convention is baked into the symbols on your screen. The triangle-and-bar of a diode, the arrow on an LED, the arrow on a BJT's emitter: they all point along conventional current, not electron flow.
Where this bites you in embedded and IoT work
For anyone doing embedded and hardware design, the convention is not trivia — it is the grammar of every datasheet you read.
It is why a microcontroller GPIO pin is described as sourcing current (conventional current out of the pin, into the load) or sinking it (into the pin, to ground), and why those two directions almost always have different maximum ratings on an ESP32 or STM32. Wire an LED expecting the pin to source when the design sinks, and it simply never lights.
It is why active-low logic is so common. Sinking current to ground is often the cheaper, more robust option in silicon, so resets, interrupts, chip-selects and I2C lines idle high through a pull-up resistor and assert by being pulled down. Every open-drain bus on your board is an artefact of the same current-direction reasoning.
And it is why current direction in a schematic review — through a flyback diode, a current-sense shunt, a MOSFET body diode, a ground pour — is always read Franklin's way. The convention is old, arbitrary and technically inverted for copper. It is also universal, which is worth far more than being right.
Building connected hardware and need the firmware, PCB and cloud side handled together? Talk to the Fluidwire team.
Top comments (0)