"maxlynn" wrote in message
news:j90Lb.103199$pY.54961@fed1read04...
Vic:
I was also trying to stay on topic. And I completely understand the
function of the accelerometer which you reference, as well as a number of
other similar instruments. I would similarly invite you to reconsider my
question regarding a zero-g environment for such an instrument. You state
that the accelerometer is built to function in a one-g environment. I
submit to you that the instrument is built to function in a +/- 2 g
environment. If you place this instrument's two sensitive axes in a local
horizzontal plane in your laboratory, neither axis will see a
gravitational
input, and therefore will provide no output. Now, rotate the instriment
about one axis. The opposite axis will begin to sense a component of
gravity as a function of the angle which you have rotated about the other
axis. Vic will say AHA, I have sensed a rotation. But now, let me simply
slide the accelerometer across the table, moving along that same axis
which
just provided the output previously, but not rotating the instrument about
either axis. As you can see, the indication from the instrument is the
same
in either case. In both cases, all you have "sensed" is linear
acceleration, in one case, the amount proportional to the component of
gravity, and in the other case, the amount proportional to the force
exerted
on the case to provide the linear motion.
Picture the accelerometer on a boat in turbulent conditions with the input
axes similarly aligned(both horizontal), and consider how you would remove
the component of acceleration due to a wave passing under a boat and
lifting
the boat several feet and then allowing it to sink several feet,
simultaneously causing both roll and pitch motion. Consider, please, that
the wave may cause vertical and horizontal displacement of the boat, as
well
as rotational motion about the pitch and roll axes. Then please explain
to
me how you plan to discriminate on an instantaneous basis between the two
totally independent types of motion using only the two outputs from your
accelerometer. I really wish you would stop and consider my original post
where I pointed out that an accelerometer, and particularly the Analog
Devices ADXL 202(not made by AMD as your URL directed us) is an instrument
built to sense LINEAR acceleration along an axis, NOT ROTATION about an
axis. Gyros are used to sense rotational motion, not accelerometers.
Pendulums(sometimes also called inclinometers) are also used for limited
rotational sensors, provided the motion is not too large and/or the
frequency of the rotational motion is not too high. I did not mean to get
so wrapped up, so please excuse my fervor.(rant??)
"Vic Fraenckel" wrote in message
...
Max Lynn wrote:
| As an exercise, consider how you would perform your suggested
computation
in
| an orbiting spacecraft(i.e., zero g).
Do you understand the concept of how the ADXL202 works? They are
designed
to
work in a one g environment. The gravitational force is exerted on a
solid
state (silicon) beam and the beam deflection produces the output which
is
a
function of the sensed gravitational field which is essentially 1g when
the
beam is perpindicular to the gravity field and 0g when the beam is
parallel
to the field. Read the data sheets and the application notes that you
can
easily find at the AMD site ( http://www.amd.com/us-en/ ). I urge you to
do
so.
Very few boats will ever find temselves in a zero g environment. My
comments
are directed to a query about a pitch/roll sensor for a sailboat. Just
trying to stay on topic.
HTH
Vic
Excellent explanation of the problem, and thats why we used two dirty great
big Datawell Hippi's filled with glycerin or some such thick fluid.
garry