Brainstorming how to cause, and measure displacement.

Hello. I have very limited experience with arduino and have done some googling and have yet to figure out the best way to go about this project I am working on.

What I want to do: I want to measure the wobble of a metal bar in a slot.

I want to be able to shift (wobble) the bar linearly, back and forth and measure that distance. If they can be done without human intervention would be best. But having to physically move the bar wont be a bad option either.

Wobble won't be more than 2", but because it may be very small movements I need something fairly accurate.

Some Ideas:

  • I've though about using the same setup as one would use for a light switch, using a servo motor, combined with a sensor for distance. Similar to this http://www.oomlout.com/a/products/sesw/The only problem with that is the wobble distance is unpredictable so guessing and stopping the sliding distance of the servo would be hard to predict.
  • I also thought about using a linear potentiometer to move the bar using a servo. But how would you go about translating the sliding motion -> the servo -> and accurate distance measurement

I have access to a rapid prototyping machine if there is a need to manufacture specific components.

I'm open to any suggestions! Thanks in advance.

accelerometer ?

A magnetic pickup or prox sensor might work.

Gut a DVD drive and use the head actuator?

How fast is the bar moving, accelerating? Quantify fairly accurate. If you tell a servo to move it, you should know the position within the accuracy of the servo.

I want to measure the wobble of a metal bar in a slot.

Are we allowed to ask why ? (what's the purpose or objective of all this ?)

groundfungus:
How fast is the bar moving, accelerating? Quantify fairly accurate. If you tell a servo to move it, you should know the position within the accuracy of the servo.

Not fast at all. Total linear movement can be as little as .25cm so around +.05cm would be fine.

raschemmel:

I want to measure the wobble of a metal bar in a slot.

Are we allowed to ask why ? (what's the purpose or objective of all this ?)

It's just to verify the fit. The less wobble the better and more reliable the fit. Were talking pencil size scale.

I'm not completely sure I understand your setup, but what I'm envisaging is something to push on a part and see how far it can be moved. I'd try something like this:

Spring the part so that it is held lightly at one end of the available travel.

Build a position sensor consisting of a potentiometer operated by a pushrod (e.g. rotary pot/crank/pushrod).

Build an actuator which will push the free end of the pushrod lightly against the part to be moved, with enough force to keep it in contact but not enough to move the part. The intent is that if the part moves, it moves the potentiometer. The actuator could be a servo in series with a soft spring.

Build an actuator which will push the part to the far end of the available travel. This could be a servo/crank/pushrod arrangement;if necessary to protect the servo fit a stiff spring in series with the pushrod.

The measuring sequence would be:

Move the position sensor into contact. Read the initial position.
Use the actuator to push the part. Read the final position.
Retract the actuator.
Retract the position sensor.
Rinse and repeat.

PeterH:
I'm not completely sure I understand your setup, but what I'm envisaging is something to push on a part and see how far it can be moved. I'd try something like this:

Spring the part so that it is held lightly at one end of the available travel.

Build a position sensor consisting of a potentiometer operated by a pushrod (e.g. rotary pot/crank/pushrod).

Build an actuator which will push the free end of the pushrod lightly against the part to be moved, with enough force to keep it in contact but not enough to move the part. The intent is that if the part moves, it moves the potentiometer. The actuator could be a servo in series with a soft spring.

Build an actuator which will push the part to the far end of the available travel. This could be a servo/crank/pushrod arrangement;if necessary to protect the servo fit a stiff spring in series with the pushrod.

The measuring sequence would be:

Move the position sensor into contact. Read the initial position.
Use the actuator to push the part. Read the final position.
Retract the actuator.
Retract the position sensor.
Rinse and repeat.

Sounds promising! Ill look into doing this

sounds like you want to measure the axial movement. not sure if wobble is the correct term.

lots of options.
I like steppers, so i would use a stepper with the shaft as a lead screw.

http://www.anaheimautomation.com/products/stepper/stepper-linear-actuator-item.php?sID=378&pt=i&tID=80&cID=52

mount the stepper on a plate that can move. connect a switch to the plate. the stepper will rotate with count of steps until the plate moves and the switch is made. count steps.

this probably has flaws, but just tossing an idea on the table.