Engineering Notation Explained: Why 4700 Ohms Reads as 4.7k

Open any datasheet, multimeter, or schematic and you will see numbers written a particular way: 47 kΩ, 4.7 µF, 1.5 GHz. You will almost never see 4.7 × 10⁴ Ω, even though that is the same resistance. The reason is not laziness or convention for its own sake. It is engineering notation, a specific number format that exists so the numbers electronics people write line up with the prefixes printed on every component they touch.

I spent years reading the difference between milli and micro off tiny surface-mount parts before I understood that the format was doing real work for me, not just being terse. Once it clicks, you stop counting zeros forever.

The one rule that defines engineering notation

Engineering notation writes a number as a mantissa times ten to a power, the same shape as scientific notation. The single difference is this: the exponent must be a multiple of three.

That is the whole rule. The exponent can be 3, 6, 9, 12, or going the other way −3, −6, −9, −12, but never 4 or −5 or 7. To keep the exponent on those steps, the mantissa is allowed to roam between 1 and 1000 instead of being pinned under 10.

This sounds like a small constraint. It is actually the entire point, because those multiple-of-three exponents map one-to-one onto the SI prefixes:

10³ → kilo (k)
10⁶ → mega (M)
10⁹ → giga (G)
10¹² → tera (T)
10⁻³ → milli (m)
10⁻⁶ → micro (µ)
10⁻⁹ → nano (n)
10⁻¹² → pico (p)

Every step of three on the exponent is a named prefix you can say out loud and print on a label. That is why 47 × 10³ ohms becomes 47 kΩ, and 4.7 × 10⁻⁶ farads becomes 4.7 µF. The prefix carries the magnitude so the mantissa stays a short, readable number.

Engineering notation versus scientific notation

Both formats write mantissa × 10ⁿ, so they look like cousins. The split is in what they constrain.

Scientific notation keeps the mantissa in the range 1 ≤ value < 10. It is built for compactness and for comparing orders of magnitude at a glance, which is exactly what you want in physics and astronomy. So 47000 becomes 4.7 × 10⁴, and 0.0000047 becomes 4.7 × 10⁻⁶.

Engineering notation keeps the exponent on multiples of three and lets the mantissa range up to 1000. So 47000 becomes 47 × 10³, and 0.0000047 becomes 4.7 × 10⁻⁶ — which, because −6 is already a multiple of three, happens to match here, while the 47000 case diverges.

Notice what the engineering form bought you: 47 × 10³ is one keystroke away from "47 kilo," and 4.7 × 10⁻⁶ is "4.7 micro." Try saying "four point seven times ten to the fourth ohms" at a workbench. Nobody does. They say "forty-seven k." The format that makes that possible is the one engineers reach for. If you mostly work in pure scientific form, the scientific notation converter handles the 1-to-10 mantissa style and can sit right next to this one.

A worked example: 4700 by hand

Walk through 4700 step by step, because the manual method shows exactly what the snap-to-three rule does.

Find the leading power of ten. 4700 sits between 10³ (1000) and 10⁴ (10000), so its natural leading exponent is 3.
Round down to the nearest multiple of three. Three is already a multiple of three, so nothing moves. (For 47000 the natural exponent would be 4, which rounds down to 3.)
Shift the mantissa to match. Divide the number by 10³: 4700 ÷ 1000 = 4.7.
Read off the result. 4.7 × 10³.
Attach the prefix. Exponent 3 means kilo, so 4700 ohms is 4.7 kΩ, written 4.7k.

So 4700 → 4.7 × 10³ → 4.7k. The same snap-down logic handles 47000 → 47 × 10³ → 47k and 470000 → 470 × 10³ → 470k, all three landing on the kilo step because their exponents (4, 5, and 6) round down to 3, 3, and 6 respectively. The mantissa absorbs the slack, which is why it is allowed up to 1000.

Reading and writing electronics values

This is where the format earns its keep daily. A resistor or capacitor value in engineering form is two pieces: a number and a prefix. To expand it, strip the prefix and put back the power of ten.

47 kΩ = 47 × 10³ Ω = 47000 Ω
4.7 µF = 4.7 × 10⁻⁶ F = 0.0000047 F
2.2 nF = 2.2 × 10⁻⁹ F
100 R = 100 Ω (the R stands in for the decimal point on parts that avoid the Ω symbol; 4R7 means 4.7 Ω)

The single most expensive mistake here is confusing milli and micro. The symbol m is milli (10⁻³) and µ is micro (10⁻⁶) — a thousand times smaller. Writing 4.7 mF when you mean 4.7 µF is a 1000× error that quietly wrecks a filter cutoff. The micro sign is also a typing trap: the proper character is µ, but most keyboards only offer the latin letter u, and pasted text sometimes carries the greek μ instead. A converter that accepts u, µ, and μ all as micro saves you from a parser that silently chokes on the wrong character.

Type any of these into the engineering notation converter and it runs both directions: 47k expands to 47000, and 4700 collapses to 4.7k with the kilo prefix named alongside, plus the scientific form shown for contrast so the multiple-of-three rule is visible rather than implied.

Where it shows up beyond resistors

The format is not electronics-only. Any field that spans many orders of magnitude leans on it:

Clock speeds and bandwidth. 1500000000 Hz is 1.5 GHz. A spec sheet full of raw decimal hertz looks amateur; the giga form scans in a single glance.
Currents and voltages. 0.0047 A is 4.7 mA. 0.0000012 V is 1.2 µV.
Mechanical tolerances. 0.000025 m is 25 µm, the way a machinist actually quotes it.
Data and storage. Sizes climb in kilo, mega, giga, tera steps, the same exponent ladder.

When you are normalizing a messy vendor feed where some frequencies arrive as 1.5G and others as 1500000000, converting everything to one plain decimal (and grabbing the code-friendly 1.5e9 form) before you store it beats hand-parsing prefixes scattered across a codebase. For the broader job of swapping between physical units rather than just magnitudes, pair this with the unit converter.

When numbers run past the prefix table

The SI prefix table is finite — it runs from quetta (10³⁰) down to quecto (10⁻³⁰). Engineering notation itself has no ceiling or floor, so a number like 5 × 10³³ still has a perfectly clean engineering form even though no named prefix covers it; the exponent stays a multiple of three and the mantissa stays under 1000. In practice, every value you meet in everyday electronics, mechanics, and physics sits comfortably inside k / M / G / T and m / µ / n / p, so the prefix column almost always has a name to show.

That comfort is the deeper reason the format won. A human-readable mantissa, a sayable prefix, and exponents that step in threes mean schematics, datasheets, bills of materials, and the multimeter in your hand all describe the same number the same way. Once you read 4700 as 4.7k without thinking, you are reading like an engineer.

Made by Toolora · Updated 2026-06-13