10F spindle runout/bearing preload

[cdhknives]

While I had the indicator in the toolpost checking my ram, I indicated all sorts of spots on the lathe to see how worn it is. I particular I was seeing about .003-.004" of wobble in a brand new MT3 dead center in the spindle...with similar movement with hand applied firm pressure.

I did a little research on the spindle and learned there is an adjustment for the bearing preload. I don't have an original manual, so I did some research and found info suggesting: tighten and torque an additional 1/16 turn...or the same rotation as 2 teeth on the spindle gear.

I didn't find any info on what the initial tightening should be.

After a 20 minute warm up running on medium speed with the idler locked in: I set it firm, to the point the hand pressure would only cause about .001" deflection, and then added the suggested 1/16th turn. This was about 1/8th turn of the nut from its previous position.

Unloaded spindle indicated movement is less than .001" and firm hand pressure causes about .001" movement.

Running for 10 minutes causes no noticeable warmth in the bearings or inside the spindle.

Should I leave well enough alone or can that last .001" runout be adjusted out?

 

[wa5cab]

In Downloads, you will find the Atlas Technical Bulletin on disassembly and reassembly of the Atlas 10" headstock. The initial assembly is to zero endfloat in the spindle. Then tighten 2 teeth. Run 20 minutes. Check the runout. Mark the high spot. Eject the center, rotate it 180 deg., and repeat the runout check. If you still have only 0.001" runout, leave it alone. New bearings might improve it but they might not, too.

Robert D.

 

[cdhknives]

The indicator needle at the spindle barely quivers with axial loads, at least of a magnitude as I can provide with my 2 hands and a good lean. Perpendicular deflection is running just under .001 indicated with similar force. No sign if heating from the bearings in a 60 degree garage/shop.


Bed width measurement for wear indication was a max-min difference of .0015" on my digital calipers (no micrometer big enough). I was really surprised by this number being this small.


Here is a test cut on a piece of annealed 7/8" W1 rod. .008" advance, 1/64" radius carbide bit, "Interstate" brand from Enco, .0044" feedrate, no lube. Silver chips and springs, some of the springs 2'+ long.





I'm thinking it's about as good as it's going to get. Time for some serious work!

 

[wa5cab]

Using a much larger radius cutter for finishing will give a much better finish (usually).

Robert D.

 

[Dranreb]

You might like to try mounting your indicator directly onto the headstock, less chance of any error that way.

Bernard

 

[Kernbigo]

I used to have a 10" and ran the spindle bearings with .0005 lift, using a bar in the head stock to check the lift (about 18") bar

 

[cdhknives]

Using a much larger radius cutter for finishing will give a much better finish (usually).

Robert D.

I know I need some cutters with larger radius, but finances and all, trying to buy 10 packs, and could only afford one. I could in theory thread large material with these. Given the amount of gouging and chatter present 6 months ago when doing the same thing, those straight, smooth cuts are beautiful to my eyes!

Some 320 grit with light oil and a hard backer would polish this in a minute flat.

 

[Kernbigo]

I respectfully disagree a small lathe using carbide a !/64-1/32 is perfect, you have something else going on. This is the insert i use and used it on my 10" atlas and my 9" south bend samdvik insert smt 32.51-mf 3(2.5)09t304-mf 1025 ebay $4.99