Tilt sensor

[Roy L. Fuchs]

I'd go for pins from the top and a strip at teh bottom. The strip at the
bottom is ground; measure teh conductifity from the two other pins that
are more-or-less inserted in water (or other liquid).

Ever heard of capillary attraction? Surface tension? How about
miniscus? Wetting? All four can pose problems in such a system.

 

[Roy L. Fuchs]

I think you will run into huge errors with this method when the angles
are small due to noise and inaccuracies in the accelerometers.

You're nuts. They are far more accurate and less noise sensitive than
those of years past. The guy wants a 2 degree resolve. These can
detect a 0.1 degree tilt. The only problem is the support circuitry
and readout method implementation.

Where is your head at?

 

[[email protected]]

Just a simple solution that wouldnt take much circuitry would be to use
a mercury tube like older thermostats.

John

 

[John_H]

Nice item, but it also requires a lot of support circuitry, as well
as software. Hell a G force resolver would work, if you want to get
into advanced technical solutions.

I liked the parts so much I bought a couple analog parts in the little
8-pin LCC package from Digikey yesterday to play around with (along with
a couple 2-axis devices from Analog Devices). The ST digital devices
are neat, taking you out of the analog realm, but the 28-pin leadless
pack is less simple to deal with for tinkering.

Would you expect any tilt meter to need no support circuitry? I'll need
to manipulate analog signals and do some vector math but tilt should be
simple to extract.

 

[Nico Coesel]

You're nuts. They are far more accurate and less noise sensitive than
those of years past. The guy wants a 2 degree resolve. These can
detect a 0.1 degree tilt. The only problem is the support circuitry
and readout method implementation.

Where is your head at?

Support circuitry and readout method... Besides that you'll have added
noise coming from the floor, air flows, sounds, doors slamming, trucks
passing by, etc, etc. The datasheet may specify a 0.1 degree
resolution, but you should bare in mind that this is probrably a
hypothetical figure based on math rather than practical situations.

 

[John_H]

Support circuitry and readout method... Besides that you'll have added
noise coming from the floor, air flows, sounds, doors slamming, trucks
passing by, etc, etc. The datasheet may specify a 0.1 degree
resolution, but you should bare in mind that this is probrably a
hypothetical figure based on math rather than practical situations.

You're funny.

Any tilt system... ANY tilt system is sensitive to outside forces that
disturb the gravitational vector such as those items in your list above.
You just need appropriate filtering to get out the DC (or the limited
bandwidth you want).

ANY system that needs accuracy to 2 degress in the range of +/- 60 degrees
will need the support circuitry and a readout. The original request was not
for a mechanical system where the readout can be visual.

 

[Rich Grise]

Hi,

Some of my students need to use a tilt sensor for their project.
Unfortunatley they need a sensor with a continuous output to measure
the angle of tile over a wide range ( +/- 60).

Can anybody suggest a sensor that measure tilt yet it is still cost
effective or maybe a homebrewed method that gives similar results.
Accuracy is not an issue here, 2 degrees tolerance is allowed.

Any idea will help!

Support a ball-type mouse by its edges, and glue some kind of bob
to the mouse ball. Then write some S/W that displays a calibrated
grid, and the mouse moves a crosshair cursor.

Good luck!
Rich

 

[Nico Coesel]

You're funny.

No, I'm getting old and should indeed be looking where my head is at
. You're right; you'll need to filter either mechanically or
electronically.

 

[Roy L. Fuchs]

I liked the parts so much I bought a couple analog parts in the little
8-pin LCC package from Digikey yesterday to play around with (along with
a couple 2-axis devices from Analog Devices). The ST digital devices
are neat, taking you out of the analog realm, but the 28-pin leadless
pack is less simple to deal with for tinkering.

Would you expect any tilt meter to need no support circuitry? I'll need
to manipulate analog signals and do some vector math but tilt should be
simple to extract.

No, I would expect it to need some, but I think the OP was looking
for a much simpler solution, though I am not sure what his needs
actually are, nor the degree or depth of circuitry he wants to
incorporate into the thing.

 

[Roy L. Fuchs]

Just a simple solution that wouldnt take much circuitry would be to use
a mercury tube like older thermostats.

John

That would only create a tilt switch.

He wants a tilt sensor that provides a readout in degrees with a 2
degree resolution. An entire cascade of those switches, placed in 2
degree pre tilted increments would be possible to wire up to a circuit
that could read how many of the switches were closed, but not a single
switch.

 

[Roy L. Fuchs]

Support a ball-type mouse by its edges, and glue some kind of bob
to the mouse ball. Then write some S/W that displays a calibrated
grid, and the mouse moves a crosshair cursor.

It would have to be a fairly sensitive mouse, and also, they only
have the ball in suspension when it is pushed up into the ball cavity.
Hanging a bob under a mouse ball will lock most of them up on the dust
scraper (at the ball opening of nearly all mice).