Instrument bearing noise problem 1

[gaufridus]

I have a shaft running in two ball bearings. The shaft is vertical and has an eccentric at one end and driving pulley at the other. The eccentric moves a mass of 132 grams which is biased against a bidirectional spring of stiffness 17 N/mm. (yes its all quite small). The mass is constrained to move in a linear motion. The shaft itself is 10mm diameter and runs at up to 6000 rpm (more if possible). Eccentricity is adjustable up to 0.75 mm (ie total travel of 1.5mm. The problem is that the mechanical noise from the bearings is too large for the sensitivity of the whole instrument. The instrument is sensitive to noise above a few microns level. Contamination from oil/grease etc would be a serious problem (medical use). Anyone have any ideas for a replacement bearing?

 

[alexit]

The mechanical noise in bearings is function of size, preload, external excitation, and quality. You measure noise as function of vibration in mm/sec yes? What frequency range causes problem (running speed, cage noise, ball pass, etc.?)

If your equipment is sensitive within a dominant frequency, change bearing frequency is possible (cannot dampen all bearing noise.) It is more expense but instrument air bearings or magnetic bearings also available. Does device operate in sterile field (air may not be allowed?)

 

[Barry1961]

With that light of a load you should be able to use a plane bearing.

Are you preloading the bearings with wave springs? That might reduce the noise.

Barry1961

 

[MikeHalloran]

Dump the shaft, the eccentric the bearings, the pulleys, the belt, and the motor.

Suspend the mass on cantilever leaf springs and drive it with a voice coil and a servo.



Mike Halloran
Pembroke Pines, FL, USA

 

[gaufridus]

Well spotted Mike!
We do in fact have two prototypes - both types have the mass suspended on leaf springs. One is motor driven and the other is driven by a voice coil. So far we have only the motor driven version running. I was hoping that someone could help with the refinement of the bearings in the motor driven version.
As far as conventional plain bearings are concerned I avoided them because of the rapid load reversals, although I was hoping that someone might have some experience of harder (say ceramic) plain bearings.

 

[ccw]

Hi gaufridus[\b],

Flexures and linear coil drives like Mike describes are hard to beat for your application. Flexure bearings(if you can call them that) are practically zero hystereses, friction free, and no lost motion.

If you want to pursue ball bearings, try Barden (now part of FAG) super precision, ABEC 7 or better, angular contact spindle bearings in matched sets. Follow the set up instructions very carefully. They have metal and ceramic ball versions. If you are successful there, then you will probably discover that the belt drive and motor input are too noisy.

 

[StefanHamminga]

If you are looking in trying ceramic bearings:

http://www.cerobear.de/

on the noise issue I think ccw has a good point!


Stefan Hamminga
Mesken BV
2005 Certified SolidWorks Professional
Mechanical designer/AI student