# What is a gyrocompass and how might one be used by a planetary rover?

In comments and links therein below the question How does Curiosity know how to point and move it's high gain antenna in real time? there is mention of a gyrocompass.

I am aware that gyros can be used to measure changes in attitude, i.e. rotations, and that conventional compasses can be used to measure attitude with respect to an external magnetic field.

But what is a gyrocompass, by what principle does it function, and how might one be used in the context of a planetary rover?

Planets rotate. Place an accurate gyroscope on a table and you can readily observe earth's rotation.

A 3-axis gyroscope measures the axis of rotation and angular velocity around it. Angular velocity is rather unimportant as it is a known quantity anyway (earth: 360°/24h), but knowledge of the axis of rotation proves to be valuable:

• The horizontal component of the axis of rotation is the north-south direction.

• Its elevation (in degrees above/below horizon) is your latitude.

To use a gyroscope as a compass, the vehicle must stand still for several minutes and the planet must have sufficient rotation.

• So the term gyrocompass refers to a gyroscope with sufficient sensitivity to determine the direction of a planet's rotation. Its a subclass of gyroscope, not a distinctly different instrument? Your two bullets refer to references from other instruments to establish horizontal. If I had a gyrocompass mounted on a rover, what information could be obtained without referencing to other measurements as well?
– uhoh
Apr 15, 2018 at 14:10
• @uhoh - Yes, a gyrocompass is just a gyroscope mounted in a way it measures the rotation of a planet. If the horizontal plane is not known from other instruments, the gyrocompass still determines true north, but in body coordinates, which is not very useful. Apr 15, 2018 at 15:09
• Okay I see, it gives you one axis. So if you had a second axis, like the fix on a star, or "down" with an accelerometer plus a clock, or a fix on the Sun plus a clock, then you could get rover attitude and point the high-gain antenna as mentioned in the question.
– uhoh
Apr 15, 2018 at 20:06
• A gyrocompass doesn’t really measure the rotation: it uses the rotation to stabilize its axis pointing North, so the daily rotation isn’t torquing the gyroscope. Once a gyrocompass has self-aligned, you get no more info from it on rotation, just direction. Apr 15, 2018 at 20:56
• @RainerP. You’re probably right about modern non-rotating systems. Truly am feeling a bit old as I worked on the calibration of this (hq.nasa.gov/alsj/19740003321_1974003321.pdf) rotating one-axis gyro to see whether a lunar-North GC or Sol-aligned IM approach was better. For the short mission duration, IM was chosen as more robust though it requires multiple set up. Apr 15, 2018 at 21:50

A gyrocompass mounts a gyroscope in a particular way so that its axis (eventually) self-aligns with the axis of the planet, e.g. Earth’s, rotation. In this way you get a self-contained indication of true North.

It doesn’t depend on a magnetic field, which might be handy on e.g. Mars.

It works because the planet rotates and carries the mount along. If planet & gyrocompass axes are parallel already, great. If not, the daily rotation of the case torques the gyrocompass into alignment. On the Moon, with its long days hence slow rotation, this might not work well.

Note that this is different from the usual inertial gyro “points in a constant direction” idea: a gyrocompass deliberately doesn’t do that, but rather rotates it’s axis to be North facing.

• Can you find a source that explains how a gyroscope becomes a geocompass only if mounted so that it moves in a certain way? There are many kinds of gyroscopes, only some have a spinning mass; there are also gyrscopes based on light (usually but not always optical fiber based) and gyroscopes based on vibrating MEMS structures. All can measure rotation, but many do not "move" as you are describing.
– uhoh
Apr 16, 2018 at 3:26
• @uhoh Wikipedia has basic info on a couple approaches. en.m.wikipedia.org/wiki/Gyrocompass for nostalgia this description is hard to beat ed-thelen.org/SperryManual-05.pdf Part of the question is language: reasonable people may differ on whether non-rotating inertial platforms are “gyros” Apr 16, 2018 at 5:21
• Aha! I appreciate the distinction, even though some would claim that I don't count as "reasonable people" ;-) The Sperrry manual is a real gem, and an excellent resource to address "What is a gyrocompass...?" Would you consider expanding a bit on your answer? I think you have some insight here that other readers will appreciate as much as I will!
– uhoh
Apr 16, 2018 at 6:53

Gyrocompass is type of non magnetic compass. It is similar in principle of gyroscope but not same. The principle namely is gyroscopic precession.

One can definitely use the gyroscope/gyroscopic compass to get a coarse attitude.

If you are in a particular latitude and longitude one can calculate the direction of rotation vector one would measure in the local NEU frame. After that it is matter of finding thr rotation which will align the observed rotation direction vector to calculated one. This is your attitude in Mars Center Mars Fixed frame. This information must be combined with true absolute time(or sidereal angle) to get inertial attitude.

A Foucault Pendulum provides a nice demonstration of the effect of rotation which can be captured by gyros onboard.

Example of a Foucault Pendulum in the video Dartmouth Professor Discusses Foucault's Pendulum