Re: [CAD_CAM_EDM_DRO] Re: noob laser question

Cristi wrote:

> Hi Jon,
>
> thanks for the fast answer. The problem with angular sensors "piezo"
> or smth like this is they are not gyroscopes ( gyroscope = keep your
> spin direction in the old greek lang ) but they are angular rate
> sensors. That is, they have drift in such a way that in 15 min let say
> you will get report that you're spinning with 2 deg / minute. You
> start to compute and after 10 min. you think you've spin 20 deg but
> you haven't at all. That's why mechanical gyros have been used for
> navigation while angular MEMS or piezo sensors haven't and will never
> be. Today they have laser and optical gyros, as angular sensors but
> with much much less drift.

As I said in my last post, all real stable platforms (not
including steady-cams) use rate sensor gyros that correct the
platform orientation through amplifiers and torque motors.
I am quite certain that if you strap down 3 gyros to a
gimbal-mounted platform, you will have HORRIBLE drift. After
weeks of tweaking to balance the thing, you will maybe be able
to get the drift down to a couple degrees a minute, even with
really GOOD gyros. It is the amplifiers that improve the drift
problem by many orders of magnitude, by isolating the gyros from
the imbalances on the platform.

> 1-2 degrees per hour is not such a big problem, although I would hope
> for better.

You will never get there without rate gyros and amplifiers.
Better? Serious inertial nav systems costing several hundred
thousand $ don't do a whole lot better than this!

> As for the platform, it won't work for more than an hour. You have
> spherical platforms suspended in fluid bearings to avoid gimbals lock
> and increase precision.
> So, going back to balancing. I will put the disc on a fixed small
> table via ball bearings, and put an accelerometer on that table. My
> disc is from steel and conduct electricity. I think I will connect it
> to 5V via 2 small brushes. In a certain point, where one of the
> brushes touches the disc, I will put a small spot of glue ( CyanoA or
> smth ) so it will cut the 5V anytime the brush passes through there (
> while the rotor will spin ). What do you think?

You are going to have a 2" disc spinning at 30000 RPM, with
brushes riding on it? How much heat will that produce? What
motor will be able to keep the gyro spinning with that drag?
The glue will handle the brush going over it 1.8 million times
(one hour at 30K RPM)?

> The second approach will be an optical sensor - but I don't know what
> type of optical sensor should I use.

If you are a tinkerer, you've seen the optical slot sensors in
old floppy drives and such, an object gets between emitter and
detector and breaks the light beam. There are very similar
devices that work by reflection, look in the Digi-Key catalog
under opto-electronics. What do you need a sensor on the
rotational position of the gyro for, anyway?

Jon