SI prefixes, significant figures, tolerance stack-up
Three small tools that come up on every project: converting prefixes, rounding results properly and checking whether dimensions will fit.
Result
- In the base unit
- 0.0047 F
| Prefix | Value |
|---|---|
| T (tera) | 0 TF |
| G (giga) | 4.7 × 10⁻¹² GF |
| M (mega) | 4.7 × 10⁻⁹ MF |
| k (kilo) | 0.0000047 kF |
| — | 0.0047 F |
| m (milli) | 4.7 mF |
| µ (micro) | 4,700 µF |
| n (nano) | 4,700,000 nF |
| p (pico) | 4,700,000,000 pF |
SI prefixes
Prefixes shorten very large and very small numbers: 4700 µF is 4.7 mF and 0.000 001 m is 1 µm. Four new ones were added in 2022: ronna (10²⁷), quetta (10³⁰), ronto (10⁻²⁷) and quecto (10⁻³⁰).
Significant figures
Significant figures are all digits except leading zeros. A result shouldn’t have more significant figures than the least precise input: 2.5 m × 3.14159 m gives 7.9 m², not 7.853975 m².
Tolerance stack-up
Worst case adds all tolerances, so every combination of parts fits. RSS (root sum square) assumes extreme sizes rarely occur together and gives a narrower range.
Common questions
How do I enter an asymmetric tolerance?
As +upper/-lower, e.g. 25 +0.1/-0.05. Put a minus before a subtracted dimension: -25 +0.1/-0.05.
How many significant figures does 0.00500 have?
Three: the 5 and the two trailing zeros. Leading zeros only place the decimal point.