So. This has come up a few times recently. Querying the CD community for their thoughts on this as a project.
Disclosure: I have no formal training in laser optics. Google is my only friend in this regard. My real experience is in control theory and software architecture. I am thoroughly out of my comfort zone here.
Fiber Optic Gyroscopes use coils of fiber optics, along with coherent laser light to detect rotational motion. Due to the Sagnac Effect rotation of the coils sets up a standaning wave pattern, which (in the right configuration) modulates the intensity of the light beam measured.
My main questions, related to making a functional, <$500 one at home:
Am I completely off my rocker? If not:
Is there anything special about the fiber optic cable required to make one of these? Will the stuff they use for computer networking work (It’s cheap!)
Do you have to run the laser down both directions of the cable simultaneously? Or can you split the laser source into two coils (wound opposite directions), re-join the output of both cables, and measure the intensity?
Websites like this one and this one and this one seem to imply its possible… However, phrases like “mV difference is approximately proportional to RPM” make me a bit worried.
A Stanford lecture presentation seems to carry a few more details (“polarization-maintaining fiber”??), but full comprehension of it is a bit above my current paygrade.
So, before investing a ton of time and money, tossing this one out to the wise folks of FRC.
So, a FOG is basically a Michelson interferometer that couples one or both arms of the interferometer into some coils of optical fiber. I suspect that you could make it work by only sending light down the coil one direction, and using a small reference arm for the other half of the interferometer, but I think that would halve your sensitivity. It would also introduce fluctuations due to temperature etc, which are cancelled normally. Note that FOGs are usually implemented by sending both arms through the same coil in different directions (counter-propagating), rather than using two separate coils.
In order to implement the normal Michelson configuration, your fiber path length is pretty much limited by the coherence length of the laser; if the light isn’t very coherent after it exits the fiber, you won’t get a meaningful interference pattern. If I’m reading that Stanford presentation correctly, they seem to be proposing a kind of optical heterodyne technique that phase-modulates the light and looks for interference of that waveform, rather than waves of coherent light. You need some extra modulation hardware, but that allows you to overcome the coherence length limitation and use much longer fiber runs. And, of course, you could use a much cheaper light source with short coherence length.
I think you’d want to use bare optical fiber, rather than the jacketed stuff used for fiber optic telecommunications; FOGs have miles of fiber, and the jacketed coil would be way too bulky.
You might be able to put together a testbed FOG with off-the-shelf connectorized fiber beamsplitters and such, without having to do your own fiber cleaving and coupling. I’ll look into that; most of my optical experience has been freespace, and I’m only recently getting into fiber stuff.