Basic question on small bearings.

[Highspeed]

In a current high speed motor design,we are using 2x 623ZZ (OD=10mm, ID=3mm, H=4mm) 15% grease filled bearings and a mechanical guy (gone) designed something that we can probably improve .
Bearings :
are hold in housings with a 4-8 µm play
are tight fit on M6 3mm shafts
are 6N preloaded using a spring washer inside the housing.

Those bearings are oversized regarding axial/radial dynamic/static loads

The detail we need to improve is this play between housing and bearings that is still gives a little noise but is needed for the preload to apply.
We are considering to get rid of the spring washer preload
and will try to apply a deadweigh load while the assembly will cure.
Any suggestions will be appreciated
Cheers

 

[ccw]

Highspeed,
Your cost savings approach may not provide the same function as the spring under all operating conditions. The spring essentially guarantees a near constant preload regardless of the wear and thermal/dimensional excursions of the balls, races, armature, and housing. Loss of preload can result in increases in radial compliance. This can lead to acute and fatal consequences in some high speed motor applications. Those bearings are probably oversized for a reason.

The 4-8 um noise may be eliminated by seating one of the bearing outer races in a tough elastic/plastic compound (for instance an anerobic curing acrylic) in a manner that would not interfere with the spring-preload function.

At least, before committing to this change in production, I would send a few "prototypes" out for field evaluation by my customer(s).

One question: What is the recommended preload range for the needed service life of the 623ZZ?

 

[Highspeed]

Many thanks ccw,
It is not a cost saving approach, just trying to get rid of this play that has a noise impact and keeping the preload, I'll try tour anerobic approach that sounds more rigourous.
Actually I tried the approach described in the initial post it actually works regarding the play but is too sensitive to thermal effect (preload disappears @45°C).

We are not in production yet so little changes are still OK but we will launch a whole set of V&V.

The recommended preload is 3 N for our 623ZZ --- up to 6 we will still be OK for expected lifetime.
Cheers

 

[Hippo41]

It's quite common to preload miniature angular contact bearing sets with a dead-weight and glue in missile gyro's.

Have you tried a higher preload with a different spring? This may be the simplest solution.

Lester Milton
Telford, Shropshire, UK

 

[EnglishMuffin]

Hippo41:
I didn't know that. There was a long discussion on one thread (thread821-64411) about whether it is ever acceptable to loctite bearings in position - that would have been a good example to bring up, and I don't think anyone did.

 

[Highspeed]

Finally got it to work, and I have 2 options under deeper testing.
Keeping the original setting and gluing 1/2 bearing with Loctite 603
Gluing the 2 bearings with the deadweight while curing
Both are running so far according to expectation.
I think I'll keep solution 1 as I have to guarantee 10000 hours MTBF not meaning solution 2 is not going to work but will need more reliability testing
in solution 1 if the glue breaks, device will still operate and will only be a bit noisy
Solution 2 is cost effective and I'll consider it in next designs.
Regards and many thanks for your contribution.

 

[EdDanzer]

You might try using an elastomer for the preload, an o-ring is an example.