1

Is it possible to align a mems gyro to approximately true north with no external data input - ie nothing input by the user. I want to put a mems gyro/acc ship on a board and mount the board to a ceiling and have it recognize the approximate direction of north, ignoring mechanical misalignments.

JRE
  • 67,678
  • 8
  • 104
  • 179
Larry_C
  • 184
  • 3
  • 4
    No compass or something? No. – Eugene Sh. Dec 04 '17 at 17:41
  • Get one with a magnetometer in it, then you can do the "spin around 3 times" alignment procedure. – Ron Beyer Dec 04 '17 at 17:44
  • Do you want to put it in a stationary home? Just draw an arrow on the ceiling. – Eugene Sh. Dec 04 '17 at 17:47
  • 1
    @HarrySvensson I guess some context would be nice... Quite a few years ago I had a car with a compass in the dashboard that had to be calibrated by driving around in a circle 3 times in one direction, and then 3 times in the opposite direction. Even now, [Google recommends using a "figure 8" pattern"](https://support.google.com/maps/answer/2839911?co=GENIE.Platform%3DAndroid&hl=en) to calibrate the compass (align the magnetometers). – Ron Beyer Dec 04 '17 at 17:52
  • Go indoors, blindfold yourself, spin around until you fall down and puke, stand up. Now try pointing at north. Can you do that? No? Then a "*mems gyro*" can't do that either. – Harry Svensson Dec 04 '17 at 17:54
  • This question reminds me of Foucault's pendulum; spin acceleration DOES affect dynamic systems, and Earth's rotation is detectable even in a closed room. – Whit3rd Dec 04 '17 at 22:07
  • 1
    @Whit3rd Right. Now try to detect it with MEMS gyro *and* make the necessary calculations to find the north *and* do it in reasonable time... – Eugene Sh. Dec 04 '17 at 22:39
  • @Whit3rd "*..and Earth's rotation is detectable even in a closed room..*". Yeah... and how can I derive what's north if I know my latitude? Because that's more or less the only thing you can derive from Earth's rotation + gyro (assuming you know which direction "up" is). – Harry Svensson Dec 05 '17 at 00:12
  • @EugeneSh The pendulum is intrinsically vertical, and only has one sensor. 'MEMs gyro' with three orthogonal sensors is not so limited. Nail it to a spinning planet, give it a clock. Vertical is optional. – Whit3rd Dec 05 '17 at 08:57
  • @HarrySvensson: in a room that has three positions for a Foucault pendulum, where you can get three values of latitude, I can derive what's north. – Whit3rd Dec 05 '17 at 09:07

1 Answers1

3

MEMS gyros are not (yet) accurate (and stable) enough to do that, but laser-based gyros (FOGs and RLGs) are capable of directly measuring the 0.004166°/sec rotation of the Earth and determining the axis of that spin, thereby aligning themselves to true North.

If you'd like to learn more, search on the phrase "Strapdown Gyrocompass Alignment".

Dave Tweed
  • 168,369
  • 17
  • 228
  • 393
  • "*true north*". Is north where the south pole of earth's magnetic field converges? Or is north a point where earth's axis of spin is penetrating? - Maybe it doesn't matter.. It's just the first time I hear "*true* north", which would indicate that the other would be the "*false* north". I wonder if John Snow would be sad if he was King of the false north. – Harry Svensson Dec 05 '17 at 00:17
  • @HarrySvensson: I can't tell how serious you're being. True north is the Earth's spin axis, of course -- the only "north" that's relevant to navigation. Magnetic north is used as an approximation only because it's easy to measure with a compass. – Dave Tweed Dec 05 '17 at 04:51
  • No, I was being serious - Woaw, I never knew that. Now when you're saying it, it kind of makes sense. – Harry Svensson Dec 05 '17 at 05:11
  • I thought he also asked "in jest". Google "compass deviation" and you'll see the difference may be 5-6 degrees. – Larry_C Dec 05 '17 at 15:20
  • With an RLG or FOG, does heading only alignment require a maneuver, or rotation of the platform for alignment? – Larry_C Dec 05 '17 at 15:22
  • Where I live (New England USA), magnetic deviation (declination) is about 17 degrees! If you want to know more about that, check out the [World Magnetic Model](https://en.m.wikipedia.org/wiki/World_Magnetic_Model) maintained by the U.S. National Geophysical Data Center. And no, a laser gyro does not need a maneuver of any sort -- it measures the spin directly. It just needs a few minutes of integration time to get a good answer. – Dave Tweed Dec 05 '17 at 15:44