bridge circuit to sense acceleration and this is how they work.

    The mass is seeing the one G acceleration force due to gavity
    and is deflected downward. That means the spacing between the plates
    of C1 gets larger and the spacing of C2 gets smaller.

    As the spacing of the plates of a capacitor get larger
    the capacitance is reduced. When the spacing gets smaller
    the capacitance is increased.

    Think of a capacitor as a variable resisters for waves.

    There is an oscillator that supplies a constant amplitude  sine wave
 to the
    device.

    The amplitude of the wave coming out of this circuit will
    then be proportional to the number of G's the mechanical circuit is
    subjected
    too.

    Very high frequencies are used as these capacitors are tiny to say the
    least.
    The basic capacitance is on the order of only one pico farad.

    The wave amplitude coming out can be rectified or converted to DC and
    then fed to an A to D for computer input. Use four of them
    mounted equally spaced in a circle and you have a angular rate sensor.

    This is an extremely simple idea and it is easy to make.
    One can make their own, in any shop, using nothing but hand tools,
    vastly scaled up from what one finds in a MEMS chip.

    Paul Lamar


    It seems that you are discounting the sensors based on vibrating beam
 and
    the Coriolis effect.  Not sure why..  These are the sensors that
 actually
    act like gyroscopes.


 http://m.sensorsmag.com/sensors/acceleration-vibration/an-overview-mems-iner
 tial-sensing-technology-970


    Regards,

    Matt-

    Again they are misusing the term "gyroscope" for marketing purposes
 IMHO. A
    real gyroscope is a 3D device. A change
    in the Z axis of the real gyroscope results in a force in the X or Y
 axis.
    You get both roll and
    pitch information from one device.

    "As a result, the gyroscope[???] provides accurate angular rate
 sensing for
    rate changes from very small
    to very large, with minimal errors from short- or long-term drift."

    These are plainer devices hence they are angular rate sensors. It
 takes two
    and computer
    integration to get roll and pitch information. The beam deflection
 they are
    talking about
    here can be sensed by the same bridge capacitor electronics.

    http://en.wikipedia.org/wiki/Coriolis_effect

    "In physics, the Coriolis effect is a deflection of moving objects when
   they
    are viewed in a rotating reference frame. In a reference frame with
    clockwise rotation, the deflection is to the left of the motion of the
    object; in one with counter-clockwise rotation, the deflection is to
 the
    right."

    Note: no mention is made of a 3D system.

    Also they are not telling you how they sense the deflection of the
 sensing
    beam.


    Paul Lamar

    http://en.wikipedia.org/wiki/Vibrating_structure_gyroscope


    Matt-

    And: http://en.wikipedia.org/wiki/Halteres


    Matt-

   Agree that a vibrating structure gyro is not an accelerometer. It
   should measure absolute change in heading. It's closer to this:

   http://en.wikipedia.org/wiki/Foucault_pendulum

   -Chris

   No I don't agree.
   It is not a gyro!!!!!

   A pendulum senses the vertical. Is it a gyro. Hell no!!!!


   Paul Lamar

   Those links explain it pretty well. A vibrating structure that is
   vibrating in planes other than the vertical will display the same
   effect. This effect is completely independent of gravity, however the
   pendulum is simply a convenient way to illustrate it. Definitely not
   an accelerometer. Functions like a gyro, though I can see why the term
   is disputed.

   -Chris



  Hi, this seems useful:
 http://electronicdesign.com/article/analog-and-mixed-signal/This-Year-s-Best
 -Analog-Design-Will-Find-You-In-The-Dark.aspx?cid=ed_powernewsletter&NL=1&YM
 _RID=
 Murry I Rozansky

 "In June, Analog Devices released the ADIS16407 iSensor inertial measurement
 unit (IMU), which integrates a tri-axis gyroscope, tri-axis accelerometer,
 tri-axis magnetometer, and pressure sensor, with associated signal
 conditioning, in a single package (Fig. 1). That?s 10 degrees of freedom.
 Every IMU is factory-calibrated for sensitivity, bias, alignment, and
 temperature. Each sensor has its own dynamic compensation formulas,
 maximizing accuracy of sensor measurements."

 "The pressure sensor and the tri-axis magnetometer, which adds the ability
 to
 sense the Earth?s magnetic field relative to the device, is the major
 advance over the company?s ADIS16334, increasing the degrees of freedom from
 six to 10."

 "Navigation technology used in first responder or unmanned vehicles not only
 requires multiple axes of sensing, precisely aligned, but also
 cross-integrated to discern tracking/location in dynamic environments,? says
 Bob Scannell, business development manager at Analog Devices. ?Where no one
 single sensor provides the required precision, the solution involves merging
 multiple sensor types with a deep knowledge of the dynamics of the
 application environment."

Paul,

I think one challenge with using just a set of accelerometer for
measuring and displaying attitude is that they aren't measuring rates
directly.  Accelerometers must be paired with a processor which will
use the sensor to detect when attitude has begun to change and when it
stops changing.  Imagine rolling into a turn..  Assuming an ideal,
coordinated turn, first, the airplane starts to roll.  The
accelerometers detect this.  Once established at a steady roll rate,
the input from these sensors goes to zero (neutral).  The airplane
continues to bank.  As the correct bank angle is approached, the pilot
reverses the roll input, and the accelerometers are able to detect
this.  During the rolling maneuvers, the computer must estimate the
roll rate.  Once the steady state roll is reached, the computer is
timing the roll, and changing the display to show the roll rate.  Then
the roll is stopped, the computer detects this, and displays it.
There are small amounts of error in each of these steps.

To improve this, an instrument that measures attitude changes directly
is very helpful.  The vibrating beam sensor does this.  As attitude
changes (even a steady rate of change), the vibrating beam is
deflected, directly indicating the change.  An accelerometer can't do
this directly.

I believe the vibrating beam bridges this gap because the beam is in
motion - like a rotating wheel is in motion.  The vibrating beam even
experiences precession.

I agree with your point that the situation in a real flying airplane
can be made easier by using logic as well brute-force sensor
measurement..  Knowing that altitude isn't changing is a big clue.  Or
if you add GPS tracking, you can have a pretty good indication of yaw
rate.  But again, these are derived sources of information about
attitude change.  More direct sensing is useful - rate of change.

The common loss of control scenario in IFR flight is the "death
spiral" (dumb name)..  It involves a small, undetected, roll rate that
leads to decreasing altitude, increasing airspeed, and increasing
airframe G load, to the point of airframe failure or impact with the
ground.  The pilot doesn't think he's turning because he doesn't feel
it.  Directly detect the roll makes displaying attitude easier.

I admit that you can keep the wings upright by using just GPS track,
but the flying isn't going to be done very precisely, and in
turbulence, the deviation from straight and level can be considerable.


Regards,

Matt-

One needs an array of 4 accelerometers  in a diamond or X pattern.
Not only can it detect the rate of roll but it can detect a continuous roll
situation as an airplane in a coordinated turn develops a higher G loading.

It detects the the steady state  amplitude of the vertical acceleration.

It is all in the software.

I am not saying it is better than a vibrating beam.

Paul Lamar