Rate-gyro accuracy

Seems to be around $600.

John

Il 26/09/2010 22:17, John Larkin ha scritto:

This is very interesting! Do you think it maintains the accuracy with tons of iron all around few meters away?

Marco

Il 26/09/2010 22:04, Rich Webb ha scritto:

Using two gyros, one upside down, may compensate something? They will be at the same temperature so the drift should be equal. But it should sum in opposite directions.

I used VRUs on ships. For this projects I think they are too big and expensive.

Thanks to you too Marco

What are the potential liability costs if a pointing error causes Something Really Bad to happen? The up-front cost of a proper FOG or RLG IMU or AHRS with good drift numbers may be a small fraction of that.

So, can you, while setting it up, use a separate gyro or compass on the service boat, and turn the platform, take readings, compare, and compensate for the iron with some kind of look-up table or something? The iron isn't expected to change, is it?

Thanks, Rich

A gyro shouldn't be affected by metal at all - it doesn't rely on any magnetic field, just the gyroscopic effect. Of course, then you'd have to account for drift and precession - the gyro will point the same way relative to the Universe, even as Earth rotates on its axis and revolves around the Sun.

But they seem to have figured out how to correct for that on, say, airliners. :-)

Good Luck! Rich

Un bel giorno Marco Trapanese digitò:

Thermostat the sensor; most of the offset and sensitivity error of MEMS sensors is due to temperature drift. You won't probably get down to 2° per hour though; we are talking about fiber optic gyro performances here, and if it was enough to thermostat a

Gyrocompasses use the rotation of the Earth to point at true north, not magnetic north. They take a while to settle, but don't drift.

Airplanes use inertial guidance systems, which do drift a little, not enough to be a problem over the duration of a flight. I suppose thay'll all be GPS soon.

John

That depends - to some extent - on how good you think your thermostat can be.

What you can buy as a thermmostat for a water-bath never seemed to do better than +/-0.1K.

The best performance in the literature is +/-3.5 microdegrees; basically a water-bath completely embedded in a foot thick layer of plastic foam insulation, and a thermal time constant of about six hours. +/-1 millidegree can be managed with shorter time constant packaging - you can't completely embed most optical systems, and single-mode laser diodes need that kind of temperature stability.

-- Bill Sloman, Nijmegen

Going back to GPS,

How wide is the platform? Two sensors (or three to define a plane) will give orientation. If it's not wide enough but can tolerate extra width, outriggers provide more "moment" on the measurement.

Tim

-- Deep Friar: a very philosophical monk. Website:

According to OP, that was the first thing he dismissed, as the unit is under water - no reception.

Cheers! Rich

Yabbut, how close to the surface? If it's not moving around, how hard could it be to put an arm on it, or something?...

Tim

The error in the yaw axis will be roughly proportional to 1 - the cosine of the pitch and roll, if you're disregarding them. So maybe you could get away with just one gyro.

But if you're running a vehicle around, you're going to get pitch and roll far higher than a couple of degrees, and the "1-cosine" error will climb rapidly as the angle increases.

Il 26/09/2010 23:01, Rich Grise ha scritto:

This exactly what I usually do in similar situations.

The iron on the platform itself doesn't change. But the platform will go very close to *big* iron pylons. I already verified that a digital compass saturates in a while.

Marco

Il 27/09/2010 07:27, Tim Wescott ha scritto:

In fact, for 2 degrees of roll I have an error on yaw less than 0.001

Nope. I said the pitch and roll is *always* limited to that values.

Marco

Il 27/09/2010 06:51, Tim Williams ha scritto:

Sorry, I can't provide too much details but there is no way to use GPS, also for depth. It was the first and straight idea.

Marco

Il 26/09/2010 23:23, dalai lamah ha scritto:

Ok, I like it. And I could give it a try quite easily.

Marco

PTC

Then it's running on a track or an arm of some sort. Why can't you measure the position against that?