Quad encoder problems

[SteveB]

Having a little experiance with this from the operator end, Darrel is right, at any flight speed. It is inconcievable that you will only be flying in straight, unaccelerated flight, and this is the only time your concept would be accurate. Even at slight bank angles and accelerations, your "gravity" based system would not give you useful or accurate infomation.

Doing a little searching, found an adequate explaination of why this the the case.

Steve

[Darrel Taylor]

Darn 40 posts per page.

Smitty, if you missed it, go back to post #40, page 1.
Steve and I both hit it at the same minute
<br>

[smitty505000]

Crud! You guys are convincing me that this will not work. I didnt think the forces would be that bad. I guess it is like walking and carrying a glass of water, and we all know how it sloshes around.

Ok so a Parallax dual accelerometer shows where down is? Is it effected by inertia?

What about a pair of Mem's inclinometers from Digital ADIS16203?

I need to do some more reading I guess.

Smitty

[Darrel Taylor]

Is it effected by inertia?

Yes it is. That's why it can't give an artificial horizon.

What about a pair of Mem's inclinometers from Digital ADIS16203?

At $48 a piece for a single axis, It's makes the dual axis for $29.95 look a lot better.
<br>

[smitty505000]

I am a little confused. Will inertia effect the ADIS16203 digital inclinometer as well? Dosnt it work off the earths magnetic field?

Is there any way I can make this artificial horizon work without a PHD in astrophysics?

Smitty

[sayzer]

....

Is there any way I can make this artificial horizon work without a PHD in astrophysics?
..

I was about the say that this issue is getting too complicated for me to follow.

[SteveB]

Here is a quote from the datasheet of the ADIS16203:

The ADIS16203 is a calibrated digital inclinometer that provides a full 360° of measurement range in any rotational plane that is parallel to the earth’s gravity. A dual-axis accelerometer provides the base-sensing function, which resolves the earth’s gravity into two orthogonal vectors, as displayed in Figure 23. A power-efficient approach to a common trigonometric identity converts these orthogonal vectors into an incline-angle measurement.

It is an accelerometer. Thus, in flight, will only be accurate when the plane is in wings level, unaccelerated flight. In other conditions, it will be giving you the resulting sum of all accelerations on the plane, which will certianly not be the same as the acceleration of the earth's gravity.

Steve

[sayzer]

Does that mean that if the speed in any direction is constant (meaning not increasing or decreasing) then you get zero?



--------------

[Luciano]

Hi,

See these links.

Vibrating structure gyroscope

http://www.invensense.com/shared/pdf/MEMSGyroComp.pdf
http://www.analog.com/library/analog...37-03/gyro.pdf
http://www.nec-tokin.com/english/pro...ramicgyro.html
http://www.nec-tokinamerica.com/prod...ts/pd-057e.pdf

Best regards,

Luciano

[SteveB]

Does that mean that if the speed in any direction is constant (meaning not increasing or decreasing) then you get zero?

An aircraft can experiance both linear and angular acceleration in 3 axis (6 in total). If there is NO acceleration in any axis, the answer to your question is yes (and maybe no-see next paragraph). This is not "0 g". In this stabilized flight, the aircraft (and occupants) would experiance "1 g", just like what you feel sitting in front of your computer screen (unless you are really exicted playing the lastest and greatest video game )). Accelerometers would read "0", since no acceleration is occuring.

The main problem with accelerometers is that they will always be referenced to the aircraft itself. I can fly my airplane upside-down, and still make it "feel" to the accelerometer as if It was experiancing the same accelerations as level flight (but not for long). This makes them unable to provide useful attitude infomation.

A Gyro, on the otherhand, can be "spun-up" and aligned with the surface of the earth. The better the gyro, the longer it will remain aligned with it's initial reference. This will then provide the independent reference needed for attitude information.

Don't know if this helped, or just made the water muddier.

Steve

[SteveB]

Smitty,
Here is an interesting link to a site on a homebrew UAV. It deals with the problems associated with accurately calculating attitude of an aerospace vehicle. You may get some ideas you can implement on a smaller/limited scale to achieve your objectives.


Better link, check out items under "Design:". Particularly AHRS and Sensors.

HTH,
Steve

EDIT: One final note, there is a reason the off the self systems are very expensive. It is not a trivial task!

[smitty505000]

Thanks for all the info guys!! I have been doing some reading and I see what you mean. At least I know how to read encoders now.

I do have an idea how to make the artificial horizon. I need to do some tests but the basic idea is to use a couple or RC heading hold gyros. I have a futaba gy401 and have already done some tests on it. It outputs a pwm signal to move a servo. The down side is It only tracks about 30-35 degrees of movment. Now here is the interesting part. It can be reset to 0 via another pwm signal! So here is my idea. Have it set up to read movement and output to the pic. Then in the program have it set to reset when it reaches its limit. A counter will keep track of which set of 30-35 degrees it is on. This should be able to keep track with a certain amount of drift or error. Depending on how much error there is, I thought it could also be reset when level to the ground. To do this I can use a futaba pa1 pilot assist unit. This unit detects the light to dark difference optically and knows when it is level. It outputs 2 pwm channels that can reset both gyros! Hope this makes sense, I am in a hurry.

What do you think?

Smitty