[LMB] Gyroscopes (WAS OT: Sorry, Cordelia...)

James M. BRYANT G4CLF james at jbryant.eu
Mon Sep 27 14:33:41 BST 2021


Marc understands the ones in cellphones to be *accelerometers*
rather than gyroscopes, but observes that with smart software
accelerometers can do the same job, and wonders if the almost-
free ones are as accurate as the ones I used to work with in
my youth.

Matija (of the chef's kiss and exploding head) remembers correctly
that the precision gyroscope problem was resolved not by making
better bearings and better precision components, but by gyroscopes
using laser light going in loops and creating interference patterns
when there is movement.

They're both right BUT today the motion sensors in many (most?)
cellphones incorporate six orthogonal sensors, three each for
rotation (gyroscopes) and three for linear acceleration - they
have moving parts and are all built on the surface of a single
silicon chip. The moving parts of the accelerometers are stationary
and deflect under acceleration, the moving parts of the gyroscopes
vibrate rather than rotate (so no bearings are necessary) and in
both cases the deflection due to acceleration (linear or rotational)
is measured by capacitance changes. They are quite accurate and
stable (and because a cellphone is a powerful computer they can
self-calibrate to high accuracy). The technology is known as
microelectromechanical systems (MEMS) and is discussed at
https://en.wikipedia.org/wiki/Microelectromechanical_systems
and elsewhere.

MEMS gyroscopes are more than adequate for short-term navigation,
but I am not sure if they're capable of being used in inertial
navigation systems with journey times of many hours without
reference to GPS positioning. But laser gyroscopes surely are,
but they're larger and, being made in much smaller quantities,
much more expensive.

James - who used to advise people on how to incorporate MEMS
devices into their gadgets 


More information about the Lois-Bujold mailing list