The forward voltage
A conducting diode takes about 0.7 V and gives you the rest. That subtraction is why a battery project can stop working long before the battery is flat, why an LED needs a resistor beside it, and why nobody ever sets a diode's current by choosing a voltage.
A resistor divides. Put twice the voltage across it and twice the current goes through, every time, at any value. It is the friendliest part in electronics and it is why Ohm's law works.
A diode subtracts, and then gives up. Below its threshold it passes nothing; above it, the current runs away and something else in the circuit has to decide how much. Drag the slider and watch the two parts get the same volts.
The knee
The point where a diode changes its mind is called the knee, and it is the only number most people ever quote about a diode.
Four real parts, four datasheets. Three things on that curve are worth naming because they turn up in every datasheet you will ever open:
| On the curve | What it is called | Why you care |
|---|---|---|
| The knee, forwards | Forward voltage, VF | What the diode takes out of your supply |
| The flat bit, backwards | Reverse leakage, IR | Not zero. Matters on a battery, and nowhere else |
| The cliff, backwards | Peak inverse voltage, PIV | Go past it and the part is gone |
The families differ mostly in the first column:
| Family | Forward voltage | Notes |
|---|---|---|
| Silicon rectifier | 0.7–1.1 V | The cheap default. 1N4001–1N4007 |
| Silicon signal | ~0.7 V | Small and fast, low current. 1N4148 |
| Schottky | 0.2–0.5 V | Less drop, faster, leakier, lower PIV. 1N5819, SS34 |
| Red LED | ~1.8–2.0 V | Also a diode, with a much higher knee |
| Blue or white LED | ~3.0–3.4 V | Which is why they will not light from 3.3 V |
Germanium diodes sit lower still, at 0.2–0.4 V, and you will mostly meet them in crystal radios and in books written before 1980.
What the 0.7 V actually costs you
It comes out of your supply before anything else gets a share.
A protection diode on a 5 V rail leaves 4.3 V. On a lithium cell at 3.6 V it leaves 2.9 V — and a 3.3 V regulator that needed 3.6 V at its input has just stopped working, well before the cell is anywhere near empty. This is the single most common reason a project runs on USB and not on its battery.
The same subtraction, paid for on purpose, is why an LED circuit works at all: the LED insists on its ~2 V and the resistor takes whatever is left, which is what sets the current.
And it is heat. The diode burns its drop times the current flowing through it — 0.7 V at 500 mA is a third of a watt in a part the size of a grain of rice. That is the whole argument for a Schottky, and then for the ideal-diode chips that do the same job for a few tens of millivolts.
The part that catches people
Look again at how steep the knee is. Between "nothing at all" and "more current than the part is rated for" there is about a fifth of a volt, and the slope is exponential: every extra tenth of a volt multiplies the current by roughly ten.
So there is no supply voltage you can pick that gives a diode a sensible current. You do not choose the current with a voltage — you put a resistor, or a chip, or a motor in series, and let that decide how much flows while the diode takes its share and passes it through.
That single fact is behind three rules you will meet again:
- An LED always has a resistor with it.
- A diode is never connected straight across a supply.
- If you need an exact voltage out of a diode, you are using the wrong part forwards. A Zener is the diode built to hold a voltage, and it does it backwards.
Edit this page — content/fundamentals/electricity/the-forward-voltage.mdx
Discuss this article
Ask about this page. The answer stays here, on the page it belongs to, for whoever hits the same wall next.