Radial Bearing Backlash?

[LazyMan]

Does anyone have any clue on the amount of radial backlash their is in typical ABEC 1 bearing. Im designing a gearbox for a rack and pinion setup and am debating on using a ball bearing or a sleeve bearing (bushing) for the "pivot shaft." It will hardly rotate but is very important that there is no backlash.

This was probably a dumb question but is the radial runout rating on the bearing essentiall the radial backlash?

[Bill C]

The radial runout of a single row ball bearing must be designed into whatever spindle systems use is intended. Unless your query is misunderstood, the use of a double row ball bearing or externally preloaded tapered roller bearings, or a pair of angular contact ball bearings would meet your needs.... preload the bearing system for precision control.

[jcc3inc]

Sir,

Likely the sleeve bearing will be OK. We used them as pivot shafts on all the drives we made. I never found it to be a problem. IF you want to use ball bearings, a slight end-wise preload will be OK. For more substantial end loads, use angular contact ball bearings or tapered rollers. I don't think you'll need ball bearings, however.

Regards,
Jack C.

Edit: Likely you will have other inaccuracies greater than the small backlach found in plain sleeve bearings. For the ultimate no-backlash solution, use some sort of leaf spring arrangement wherein the flex of the spring(s) accommodate the necessary pivoting. WEe used a similar arrangement for encoder mounts.

JC

[bbutcher85710]

The difference between ABEC 1 and ABEC 7 bearings are in the tolerances used in manufacturing. In general higher ABEC numbers will be less sloppy, but usually it is not a good practice to just install the bearing without preload since the balls may slide instead of rolling, causing rapid wear. Higher ABEC bearings also should have less runout if that is important. I think what you need to do is design the assembly with a suitable preload to get good bearing life. You could use a thrust bearing to remove backlash, but that works well in one direction only. If the forces are low, then you can use a spring to push (preload) a single ball bearing to one side and take out the slack. This is usually done with a belleville or similar wave washer that does the job with very short length and is cheap. Most motors that I have seen use a ball bearing on each end of the rotor with a wave washer pushing the bearings toward each other to take out the slack. Again the works better in one direction than the other since it is possible to push against the spring and get movement. Conventional ball bearings are designed for equal force capability in both directions. For high thrust applications special "angular contact" bearings are made that are designed with the bearing races angled so they take larger forces in one direction without damage to the balls. In high load applications, like a lead screw, where thrust loads in both directions must be dealt with, these bearings are mounted in pairs, with rigid spacers between both the center and outer edges of the bearing that are cut to the exact length required to force the balls against the races with the proper force. It is possible to purchase special angular contact bearings in matched pairs that either use equal length spacers for both positions, or can be mounted touching each other. These have the edges ground to the exact dimension to provide the correct preload when clamped tightly together, and are quite expensive.

[LazyMan]

Alright thanks for the responses guys, the main reason im opting for a sleeve bearing is to save money. Im likely to encouter atleast .001 of backlash with a sleeve bearing which is ok. Im just trying to minimize backlash as much as possible because it will accumulate with backlash in the rack and pinion and so forth. Maybe ill just not be lazy and measure the backlash with tenths indicator some time.

[LazyMan]

Alright, I decided to take a shot of my design so others can see what I was asking more clearly.

Heres a pic of the design at the moment.