http://www.androidzoom.com/android_applications/productivity/aircraft-horizon_zcuz.html


   Robin Bergland.

   That would be worth buying a small Android tab just for that.


http://www.everbuying.com/smlclass854.html?gclid=CJr9qfbE7awCFUWMtgodigzI9A

   Paul Lamar


   I know many phones and tablets can sense which way is down to
   determine portrait or landscape orientation of the display but that
   is a single axis sensor.  How does he get roll AND pitch?  i am
  suspicious!

   SD

   I suspect the sensors are 3 axis.

   There is an ap out there where you can look at the night sky and
identify
   star names by merely tilting the tab back and fourth pointing it in
   the direction of interest. It also uses the GPS to locate one on the
   planet.

   Paul Lamar



  It can do all of that stuff with just 3-axis accel, compass, and
GPS. To
  get "real" horizon info and not just a toy that works while you're
standing
  on earth, you need a gyro. I believe some phones may have a MEMS
gyro, but
  that software only claims to use the accelerometer, which is a flawed
  method.

  -Chris W

  It sure keeps my toy helicopter flying straight.
  A lot depends on the A to D used. A 16 bit A to D is one part in
  65,000 more or less. In terms of percentage that is .0000153% I think.

  Having said that I don't know what the A to D is but 12 bits are very
  common and cheap. One part in 4096 or .000244%.  Modern CPU chips
found in
  desk tops and tabs have enormous computing power.


  Paul Lamar

Your toy helicopter uses a gyro and not an accelerometer, as far as I
know.
At least my toy helicopter does. :)

-Chris Wood

What is the difference?

Paul Lamar

A gyroscope  provides a reference from which you can directly measure
angular displacement and rate.  Note that you need at least 2 gyros
for full
spatial measurement, that is why airplanes have an artificial horizon
AND a
gyro compass.

An accelerometer measures single axis, linear acceleration.  Even with 3
orthogonal accelerometers you still can only measure acceleration.
With an
array of 6 suitably oriented and displaced accelerometers and appropriate
integrators for each you could theoretically calculate angular
displacement
but the drift might be far worse than gyroscopic precession.

Like someone said earlier, if it was easy, your inner ear would be
sufficient for IFR!

As for inertial navigation systems that the airlines and the military
used
before GPS, note that they were based on ring laser GYROS, not just
accelerometers.



SD


IMHO:

An accelerometer coupled with a computer equals an attitude gyro.

This law of physics (see the jpg) for constant acceleration gives a
distance or angle.
If the acceleration varies in a known mathematical curve it can be
integrated
using calculus. If the acceleration varies in a non mathematical way a
$30
computer can accumulate all the little bits of acceleration to
determine a total distance
or angle. It is called integration in calculus speak.

For that reason cheap computers have replace calculus :)
Mathematicians that teach calculus have been replaced by
teachers teaching programming :)

I think a ring laser gyro is a rate gyro or angular accelerometer and
not an attitude gyro.
Four simple accelerometers (mass between two springs or a mass at the
end of beam) equally
displaced from a central point can replace a ring laser gyro with the
help of a computer.
Perhaps not as accurate.

Here is a blown up picture of a MEMS. IMHO they are tiny masses on the
ends of tiny
beams. The beams are instrumented to measure bending stress due to
acceleration
of the tiny mass. My guess.

Paul Lamar

It is true that acceleration can be integrated to give displacement; the
problem is in sensitivity. The signal-to-noise ratio of current MEMS
accelerometers does not allow adequate sensitivity to provide long-term
accuracy of heading or attitude. All the available systems using MEMS
components utilize an additional input from either GPS or 3-axis
magnetic sensors to provide long-term stability.

Victor Roberts

One feature of the 4 bar chip is it can detect horzontal orientation
to one part in 4000. If the beams on the right and left are both
deflecting downward  the same amount the unit is horizontal. It is both a vertical
and lateral accelerometer and a rate gyro. If the beams are allowed
to deflect fore and aft instead of only in one plain it is a complete 3 dimensional
stable table sensor. Now I have no idea how this is interfaced to a tablet.
On the lowest level asm language it has tremendous potential. Almost unlimited.
On the Java level I have not a clue.

What is the source of the noise? Is it the strain gage molecular amplifier noise?

If it is aircraft vibration that can be filtered out with a low pass filter.

If it is strain gage amplifier noise that can be reduced by making it larger I suspect.

Paul Lamar