Hi all,

I have a rotating disk and I want to detect if it starts to oscillate.
The only solution to me seems to be a wireless accelerometer, but the problem here is the battery.. In my application I cannot change the battery every month, it should work for some years..I was not able to find any wireless accelerometer with these characteristics..

Or maybe I need a way to transfer power to the rotating disk, without slip-rings... maybe I can use magnetic induction in some way to create power on the rotating part to power the sensor?

Or, if you have any other idea to detect oscillations of the profile, you are welcome

thank you

 

Are the oscillations caused by out-of-balance conditions? or somthing else?

It's hard to visualize what kind of 'oscillations' you are talking about- please elaborate.

 

yes sorry.
If an heavy weight is placed on the rotating disk in the wrong way, it doesn't rotate anymore on a straigth plane (perpendicular to its axis), but it starts to oscillate. As you said, is a sort of out-of-balance oscillation...

 

The typical solution is a 'vibration sensor' mounted on the bearing housing, no need to spin the sensor.
Out-of-balance conditions create big forces on the bearings, which transmit through to the never-totally-rigid support structure, causing easily detectable vibrations.

 

Aircraft Dynamic Propeller Balancing

 

The oscillations are not so big.. maybe some centimeters of displacement.. I'm not sure that all the structure could receive useful vibrations...

 

how thick is the disk? was thinking along the lines of a photo interrupter

 

The oscillations are not so big.. maybe some centimeters of displacement.. I'm not sure that all the structure could receive useful vibrations...

The aircraft prop balancers I referred you to do not measure displacement; they measure acceleration.

 

no the disk is not flat..

 

Put a strain gauge on the axle and do FFT.

 

Give us pictures of the disk and what is being placed on it.

 

what is the deisk made of? if it is a ferrous metal, use a coil to pickup induced voltages from oacilations. if non ferrous conductive metal, use a coil wrapped around a magnet to pick up the oscilations.

 

it's ferrous metal.
Good idea... but how to discriminate between voltage induced by the normal rotation and the one due to the oscillations?

 

Give us pictures of the disk and what is being placed on it.

+1
This would answer so many speculations.

 

Would it be possible to mount a code band around the perimeter of disc, read by two photo reflective detectors? Multiple strips with alternating white & black markings, with each strip from center with decreasing spacing so that if disc is flat, sensors would read a constant frequency. With a little wobble sensors would read F1, F2+, F1, F2-. With increasing wobble more frequencies would be read. ??

 

your idea is brilliant Bernard, but all the optical solutions could not be used, because of the possible dirt on the sensors/disk..

The idea of the coils could be interesting, but I don't understand how to detect voltage induced by the displacement and not only the one induced by the normal use rotation...

 

Hello,

Can you make a drawing of the setup and how and where you expect the inbalance?
Could you use a magnet in the naberhood and a hall sensor?
In the axis would move in front of the magnet and the hall sensor is behind the axis, changes in movement of the axis should be seen by the hall sensor.

You could have a look in this PDF for more information on the hall sensors and their possibilities:
http://educypedia.karadimov.info/library/hallbook.pdf

Bertus

 

here it is the drawing.

I think that near the axis there is not so much space, so I prefer to stay away from it. Even because I think that the displacement are bigger at the edge of the disk and so easier to detect..

 

Hello,

What meterial is the disk made off?
When it is magnetic sensitive, you could place the hallsensor with a small magnet behind it on the corner of the base:

Bertus

 

yes, it's ferromagnetic material..

But if I put the hall sensor on the base, it detects both the movement due to the usual rotation and the ones due to the oscillations, right? How can I discriminate them?