[mmurray70]

I have a little backlash in all axes of my fadal 4020, trying to improve this a little if possible. I had the way covers off yesterday to fix an oil line and i noticed if i try and twist the X ballscrew with my hand you could easily see the coupling flexing without much force. Servo end was staying put while the rubber spider was flexing. Is this normal or are my coupling worn? Or is it just poor design to start with? My couplings have 3 lobes, i only see 2 lobe and 4 lobe for sale as replacements. Would these be any better? Would a new spider work for mine?

Any suggestions on determining how much backlash is from coupling and how much is from ballscrew? Thrust bearings appear good on the X axis. Tried indicating the groove of ballscrew from table to check ballscrew but i have over 0.001 runout in screw so its hard to say.

[jbranting]

As long as the OAL and the bores are the same the lobes don't matter. Couplers are definitely a good place to start when finding the source of backlash. I replaced all of them on both of my 4020s last year, Still had to do thrust bearings on most of the axes but it helped determine where they were needed. I also added a set screw to them to make sure there was no slippage.

[CNCMAN172]

I would start by replacing the spiders, they are rubber and wear over time, the new replacements should just squeeze into the couplers on both sides. This is pretty stiff rubber so they actually work pretty well. There are lots of cheap replacements out there from China, but if I were you I would stick with the original type as Fadal put on high quality couplers. You should be able to find the spiders by the original coupler manufacturer. You can find quality units from Rotex, Ruland, A+W, and you will find Chinese knockoffs that look similar but not made to same tight tolerances and the high end manufacturers.

Russ

[Al_The_Man]

I never use spider type coupling on new designs, but you can get quality replacements from Misumi USA.
I see you are in Canada, Misumi will ship with all customs clearance taken care of.
Al.

[mmurray70]

Just did a little more testing, may not be as bad as i thought, at least on X axis anyway. I lightly clamped a pair of vice grips on the ball screw end (hex) and indicated the handle. Looks like im only loosing about the equivalent of a half a tenths worth of rotation with normal movement which is not too bad at all. Twisting the ballscrew by hand and flexing the rubber shows about the same movement on indicator as a 0.0005" move and it springs back to center within about a half a tenth again.

My Y axis is much worse, but looks like ballscrew is tight which is great. I seem to have 0.001" slop in thrust bearing and another 0.001 lost in coupling. You can clearly rotate screw by hand and notice slop, doesn't return back to same place. Table moves about 0.001 when doing this. Perhaps this means im actually loosing 0.002" from lost rotation considering im loosing 0.001 from thrust bearing and table still moves 0.001 when rocking back and forth. Not sure if its spider or couplings just loose on shaft. Y axis is hard to service so it will have to wait a few days now. I will post results when i get at it.

Wheres the best place to buy thrust bearings?

[jbranting]

I get all my parts from FadalCNC.com

[Al_The_Man]

Don't know where you are in Canada, but most large cities have bearing suppliers'
Check them first, many bring it in in one day if needed.
The No. will be on the bearing.
Al.

[mmurray70]

Im in Newfoundland, we do have a decent bearing supply store here. But i thought these bearings were matched or something to give the proper preload? Are you sure any old bearings will do?

[DareBee]

Thrust bearings are a common replacement part on the Y, especially if you have the cool power ballscrews (which I bypassed to avoid this)
I agree with JB for parts ordering. I prefer FADAL Parts Experts - KMAC-Parts, Inc. slightly over fadalcnc

[mmurray70]

I do have the liquid cooled ballscrews. Probably a good time for me to do this as well. So to bypass i just cut the lines and connect spindle only right? I guess just blow out the old lines would be ok to drain remaining fluid?

For the Y axis, will I need to align the Ballscrew after removing the bearing block? How do you manage to do this considering where it is? Cant really do it from the front with the cover over the ballscrew.

[rodney247]

If you want to check your thrust bearings there is a very easy check. Just put an indicator in the tee-slot if your checking the Y axis or off the vise or mag base if you want to check the X. And then just push and pull by hand the table and you will see the indicator move should be only a couple tenths. The kits from Fadal part suppliers like Independent Technology Service come with the proper shim which sets the preload. Pretty easy to change the bearings.

[Tedger]

I do have the liquid cooled ballscrews. Probably a good time for me to do this as well. So to bypass i just cut the lines and connect spindle only right? I guess just blow out the old lines would be ok to drain remaining fluid?
For the Y axis, will I need to align the Ballscrew after removing the bearing block? How do you manage to do this considering where it is? Cant really do it from the front with the cover over the ballscrew.

Just remove the 2 lines from the bearing block housing and reconnect with a plastic 3/8" push-in type connector.

I leave the bolts slightly loose on the bearing block then jog the table back to allow the ballscrew to center up. Tighten the bolts in this position.

[mmurray70]

ok great thanks. Ill get the parts ordered on monday.

Would you guys recommend just sliding off the bearing block or removing the whole ballscrew and ballnut from saddle? I called Kmac to check on shipping today and they suggested removing ballscrew and all. I was thinking i would just be able to slide off the block.

[Tedger]

ok great thanks. Ill get the parts ordered on monday.

Would you guys recommend just sliding off the bearing block or removing the whole ballscrew and ballnut from saddle? I called Kmac to check on shipping today and they suggested removing ballscrew and all. I was thinking i would just be able to slide off the block.

I always remove the ballscrew. It's easier in the long run.

[carbidecraters]

Ok first of all couplers can develop grooves from the key...that can cause huge backlash. Enough to want to sell the machine. The other thing is that with just about any machine with boxed ways you should steer the hell clear of the clamp style that doesn't use a keyway...we have had those slip quite a bit.

[mmurray70]

Hi guys, i bought the complete kit from Kmac so i wouldnt have to try and figure out what size shim but something does not look right. I received two NSK 7304 angular contact bearings and a 0.001" shim. With bearings laid against one another and the shim in place i have almost no preload. I can push outer races together and there is no gap, impossible to see any light between the two. My old bearings are rough but actually seem tighter. When assembled the same way there is cleariy a light gap between them and you definatly cant push outer races together with shim in place.

Should i run both shims? Is a full thou of preload too much?

[rodney247]

the bearings with the shim should be correct....................when you pulled the bearings was there a thou shim in there too? i think your good to go

[mmurray70]

i installed the bearings in the block and installed a dummy shaft and theres no movement by hand with the 0.001" shim but I fear i have little to no preload. Tried again with two shims and doesnt feel bad. Just slightly tighter but can still easily turn it by hand. Just to be on the safe side i ordered some 0.0015 shim and i think i will go with that.

[mmurray70]

Got some 0.0015 shim from mcmaster today and used this. Thrust bearings seem perfect with this. Just a little preload can be felt on shaft now but still turns very easily.

I decided to do Z axis first and after i disassembled i noticed that my ball screw had no preload and about 0.0005" play. It looked to be in great shape, just a little loose so i thought maybe i could reball it. Did some research and measured my old balls at 0.1871 so i thought i would try nominal 3/16 balls since that seemed about the size i needed and they are dirt cheap compared to oversize balls. These worked fairly well. Definitely tightened up the ballscrew. No measurable play in ballnut now, and very smooth. Still down a touch on preload but its a big improvement for sure. Reloading the ballnut was very easy, i was surprised. I counted the old balls to know how many to put back. I put the screw on v-blocks and fed a couple of balls in at a time and rotated until i could put a couple more in and continued until i could see balls about to come out of the opposite port. Stuffed the remaining balls in the recirculating tube and put a dab of grease over the ends to keep them from falling out and put it back on the nut. Only took about 15 mins or so to load them all in.

Got it all reassembled now and im down to about a half a thou backlash on Z, was over 2 thou before so big improvement. Still might have a little slop left in ballnut under load and i guess everything flexes so much too. I think I can live with this now. Will tackle Y axis when i catch up with work. Thanks guys.