detecting linear motion

[raceeend]

Hi,

I am currently designing the electronics for a large scale RC helicopter. This is more an electronics project than mechanical since i want the helicopter to be ultra easy to fly. That means it has to hover stationary regardless of wind or other altering conditions. This means the on board chips will have to compensate for all forms of motion.

I have already been able to compensate for spin, forward tilt and sideways tilt. This is simply done by using gyro's, an electronic compass(hall effect sensor) and accellerometers. (i use the accelerometers to determine "down")

The accelerometers however are not accurate enough to see small linear motion (up/down, left/right, etc) AKA drift. Does anyone have an idea how i could properly detect this linear motion?

 

[crutschow]

I don't know of any way to detect linear motion in free space other than an accelerometer. That's what spacecraft and aircraft inertia guidance systems use.

How sensitive are the ones your are using? What is the part brand and number?

Do you do any zero compensation at startup?

 

[raceeend]

i used this one:

https://www.electro-tech-online.com/custompdfs/2011/06/DE-ACCM3D.pdf

it has an analog output. i then use a adc i had laying around to convert it to a 16 bit signal.

https://www.electro-tech-online.com/custompdfs/2011/06/AD1376_1377.pdf

i can see motions from about 0.0005m/s^2. below that is invisible.

 

[4pyros]

TRY a GPS system.

 

[Ian Rogers]

Most people just use the 9DOF .. https://proto-pic.co.uk/products/9-Degrees-of-Freedom-%252d-Sensor-Stick.html

You can get better ones.. It might be more accurate than separate items