inclinometers

[crime dog]

I work on a machine that has several inclinometers. Pendulum Pots have been used in the past, but we are changing to inclinometers that are ALL ELECTRONIC. No moving parts, no pendulums. How is this possible? What electronic component/circuits can sense what angle it is on?

[philba]

You can use an accelerometer going into an ADC to measure the acceleration of gravity. look at this page http://www.analog.com/en/subCat/0,2879, ... 5F,00.html


If you place the accelerometer's axis perpindicular to the gravity field (er, co-planar with the G field normal) it will register 0 G. Tilt it one way, it will register increasingly larger acceleration until the device is parallel to the grav field (er, perpindicular to the G field normal) where it will register 1G. Tilt it the other way and it will go down to -1G. It takes a little math to convert the acceleration value to angle.

[MicroRem]

OK, I have to ask... take a control line type airplane, which flies in a circle around the operator...can an excelerometer tell when it is in level flight, vertical going up, inverted, or vertical going down? Could it be slaved through a uP to control throttle to keep a constant speed through all the flight regimes?

just wondering what ya'll think


Tom

[philba]

yes but...

Accelerometers are pretty sensitive and will pick up all sorts of vibrations and small movements. They are usually set up with a reasonably low frequency (10s of hz) but you would still need to do some sort of filtering - maybe a simple impulse filter or just an average. With a 2 axis acc, you might be able to use both axises (axes?) to cancel out non-attitude accelerations. not sure how I'd make that work...

I've played around with them and even just holding one in your hand causes a fair number "noise" accelerations.

I kind of doubt it could be used effectively for the throttle control you mention. consider the case of a wind gust. It might last a second and induce some odd accelerations and cause the controller to think the wrong thing is happening. I also wonder how much you would need to dampen the engine vibrations.

<small>[ December 01, 2005, 10:27 PM: Message edited by: philba ]</small>

[Chris Smith]

Same with piezo gyros.

[Engineer1138]

The accelerometer only measures acceleration, so it can be "fooled" since the signal that looks like a tilted accelerometer can be due to thrust from a motor or because of rotation. To avoid a thrust force appearing to be a tilt, you need at least a two-axis accelerometer.

It's customary to combine accelerometers with rate gyros to differentiate between inclination and force from rotation. IIRC, it involves a Kalman Filter for processing the data. There used to be a bunch of links at www.rotomotion.com Those guys build autonomous helicopters and have a lot of experience with this kind of thing.

[philba]

good points engineer. a minor quibble, thrust can cause acceleration, that's what gets measured. If the vehicle is traveling at a constant velocity, constant direction, there would be no acceleration. However, the centrifugal force would be seen as an acceleration.

A gyro is a much better solution for that application.

[MicroRem]

I have tried to imagine all the forces the sensor would be affected by, and it makes my head hurt. Maybe i could use a small pitot tube and a pressure sensor like a MAP sensor from a car....

Wow, now my head hurts again,

thanks for all the replies, folks


Tom

[jwax]

The original post was how an electronic inclinometer works.
Google "Capacitive Inclinometers".

[MicroRem]

Good point Philba, I apologize for "hijacking" the thread


Tom

[Mike6158]

Eagle Tree Systems for all your RC telemetry needs

[crime dog]

Thank you for the info.