Sensor for detecting piston axial position

[Andrea Vironda]

Hi,
I have a steel piston running below a 3-4 mm thick steel wall.
I'd need to know the piston position below the wall without drilling the wall. I'm thinking to an inductive sensor but I think it will not work because it will detect all the time something in front of him (the fixed wall) despite the fast the piston is moving.
Which other technology could I use to fix this problem? Piston diameter is about 40 mm.

 

[Baluncore]

What is the temperature of the piston ?
How fast and far does the piston move ?
Or does the piston oscillate at some frequency ?

 

[Andrea Vironda]

Hi, the temperature stands between 20-120 °C, the stroke is about 30-40 mm and the speed is not so high (i think around 1" to perform the complete stroke).
There are no oscillations, the stroke is performed around 10 times a hour, but the system is placed receive vibrations from machining. (let's say 30-40 Hz, but few amplitude)

 

[Baluncore]

Maybe you could fit a strong permanent magnet to the piston.
The position can then be sensed through the wall of the tube with a Hall effect sensor, or a magnetic compass.

 

[Andrea Vironda]

Is it possible to calculate how much a wall of steel dampers the magnetic field of a magnet?

 

[Baluncore]

Is it possible to calculate how much a wall of steel dampers the magnetic field of a magnet?

Yes, but it might be easier to do an experiment with a magnet and a compass.

 

[Svein]

Here is a description of a research experiment: Fix a microphone of some sort to the wall (close to the piston). Log the output of the microphone over several piston strokes. Experiment with filters until you have something that indicates "Piston is passing".

 

[Andrea Vironda]

Experiment with filters

Thanks for suggestion. Can you be more clear here?

 

[Svein]

Thanks for suggestion. Can you be more clear here?

The whole point of the suggestion is to make you think "outside the box". When you have recorded the noise look if you can see something that tells you "yes, this is where the piston passed". Then try to characterize that noise - is it amplitude, frequency or something else? Then write a specification detailing what to discover. Then ask around how to solve it.

 

[Andrea Vironda]

Interesting, I was thinking to use an Arduino and a Hall sensor plus a magnet I can buy easily online

 

[Svein]

Interesting, I was thinking to use an Arduino and a Hall sensor plus a magnet I can buy easily online

Yes - that is what is "inside the box".

 

[Andrea Vironda]

I'm not native english, what do you mean for "inside the box"? Thinking like a robot?

 

[Svein]

I'm not native english, what do you mean for "inside the box"? Thinking like a robot?

Neither am I. "Thinking outside the box" means an unconventional approach. So "thinking inside the box" has to mean using a conventional (or obvious) approach.

 

[Andrea Vironda]

Hi,
I made the proof and I notice the piston behaviour is not the same on the outward and inward travel.
For example the damping time changes.
How can I convert this information in a tension [V] on a machine?

 

[Svein]

Better than I expected! The obvious approach would be to set up a comparator to detect the transients - this will tell you when the piston passes. Then, if you want more information, use an A/D converter to digitize the pulse shape and do some conclusions from that.