Excessive Spindle Play Fixed (newbie Error — I Didn't Need New Bearings)

[Rex Walters]

Embarrassing as this was, I'm posting this in the hopes that it might help someone else.

Tl;dr version:

My rear spindle nut was nowhere near tight enough, so the spindle assembly wasn't under sufficient tension. After making a dedicated tool to tighten the nut, I was able to reduce about +/- 0.005" of play at the front of the spindle to +/- 0.0004" (better than a 10x improvement).

Long version:

I bought my first (and only) lathe a couple years ago (11" Montgomery Wards/Logan PowerKraft 2130 circa 1941). As with any new hobby, there was a pretty significant learning curve, and I learned a ridiculous amount just correcting my various mistakes. I learned that nine times out of ten any problems I experienced were due to something I'd done wrong (dull tools, tool not at center, poor tool grinding angles for the material, poor work-holding, whatever).

Ten times out of ten I'd blame the tools first, of course.

Anyway, as I gained experience I was becoming increasingly frustrated with the amount of chatter when parting, boring, or taking heavy turning cuts. I became fairly skilled at taking light cuts and playing with speeds and feeds — I made plenty of parts accurately, if slowly, and with a fair amount of cursing along the way.

Parting was a particular headache. I spent a lot of time sharpening and honing, tightening gibs, minimizing tool and part hang-out, carefully squaring my tool, carefully checking that it's just a smidge below center height, and adjusting speeds and feeds. Still, most of the time I was getting excessive chatter, even with soft stuff like aluminum and bronze. I'd break out in a victory dance on the rare occasion that I was able to part smoothly.

The problem was particularly bad when using my large (heavy) 6-jaw chuck. Quite a while back I decided to replace the spindle bearings in the hope that that was the problem. In my defense, I relish every opportunity to take apart and learn more about my lathe (I'd never pulled the spindle before).

I ended up removing and replacing the spindle multiple times as I made various bone-headed mistakes (only a complete idiot would put the spindle back on before putting on the new serpentine belt he purchased expressly for the opportunity). I noticed that when I first took out the bearings, the front (tail-stock side) bearing took quite a bit of force to press out of the headstock, but on the final re-assembly (after two or three remove/re-install cycles) I could actually push it back into place with just finger force (significant finger pressure to be sure, but quite a bit looser than originally).

The new bearings did seem to help, but it was only a modest improvement. I still had to baby my cuts.

Finally I got smart and decided to measure things. I took off the chuck, and put a fairly hefty (3/4") drill with a morse taper shank directly into the MT3 spindle (I did need a MT2-to-MT3 adapter). Then I measured the deflection right at the tip of the spindle nose as I pushed and pulled on the drill bit (with a fair amount of force).

I could easily get plus or minus 0.005" deflection without undue pressure. This sure seemed like a lot, especially compared to tool and part stiffness, but with so little experience and such an old lathe I wasn't sure if it was excessive. I did know that chatter means something isn't sufficiently rigid.

Jim Sehr was kind enough to measure his lathe and he only saw a few tenths of deflection at the nose. Now I knew for certain that I had a problem, but I wasn't sure if it was something I could fix. I was worried that I had a clapped out lathe and that removing/inserting the spindle so many times scraped enough metal in the headstock to cause the play. If that was the problem, the only "solution" I could imagine was brazing on some more material into the headstock, and re-boring the holes for the bearings — a far larger task than I was comfortable tackling.

Fortunately, I read somewhere that the spindle is supposed to be in tension (I wish I could remember where I saw this — I'd like to thank whoever mentioned it).

The nut at the rear end of the spindle is just a ring with a hole for a pin-spanner:

I don't remember exactly how I removed that nut or re-tightened it when I first removed the spindle. I remember that I had to jury-rig something because the pin-spanners I owned all had a pin too large for that hole. Somebody on this forum or on the Logan lathe mailing list periodically suggests making a dedicated collar with a grub screw to remove/tighten this nut when removing the spindle. I don't remember whot it was, but thank you whoever you are!

It certainly seemed tight, but I wondered if I could tighten it further. So I built this tool out of some scrap aluminum:



I turned down the end of the grub screw so it fit into the pin-hole on the nut. The flats on the back of the tool let me use a big adjustable wrench to crank the nut tight. Aluminum is pretty soft for this purpose, but it sufficed for one-time use (the grub screw had wallowed out the tapped hole a bit after I used it).

Here it is on the nut, with the grub screw engaged:



I was surprised to discover that I could get a bit more than one full turn on the nut with this tool (I'm pretty sure it is now tighter than it was when I originally received the lathe, before I replaced the bearings). I was extremely happy to discover that tightening this nut reduced the amount of play at the nose quite significantly (by a factor of ten).

I haven't done a lot of work yet since making the adjustment (just a few bits of aluminum) but so far I've had no chatter whatsoever, even with very heavy boring-bar cuts. What an improvement!

I don't quite understand the physics of why this should make such a difference. The rear (left, headstock-end) bearing on a Logan is a close-tolerance slip-fit. The front bearing is a press-fit. All I can figure is that cranking down on the nut locks the rear bearing to the rest of the spindle better. I suppose it's like bolt-on legs on a table, unless you really crank down on the bolts all the little micro-slippages add up to a wobbly table. Welded on legs are infinitely more rigid than bolt-on legs no matter how much you tighten the bolts.

If anyone has a better explanation, I'd love to hear it.

Anyway, I hope this might help someone else in the future.
--
Rex

 

[Redlineman]

Har;

Your candor is embarrassing, funny, and resonant with every noob who has walked the path.

My project lathe don't quite run yet, but I think I'll go down and lean on my spindle nut a bit.

 

[Rex Walters]

Heh. Well, the funniest part to me was when I referred to it as "my first lathe."

I introduce my wife that way sometimes.... She never seems to be amused.
--
Rex

 

[wa5cab]

If I hadn't been told a year or two ago that you couldn't do that on a Logan, I would think that you had preloaded the spindle bearings.

 

[Rex Walters]

If I hadn't been told a year or two ago that you couldn't do that on a Logan, I would think that you had preloaded the spindle bearings.

Okay, here's where I really show my ignorance. I know almost nothing about bearings, and barely understand (probably don't understand) what "preload" means. I thought it was mostly about taking up radial clearance around the bearing — I still don't quite get what axial pre-load is for.

Here's a chicken-scratch drawing of my spindle assembly (minus a few spacers, etc.). The spindle is inserted into the headstock from the right (tailstock) side. The front bearing has to be pressed onto the spindle, and seats against a wider diameter "collar" turned on the spindle shaft. The rear bearing is a (tight) slip fit but has nothing to locate it axially on the spindle or within the headstock housing.



The front (right, tailstock-side) bearing locates axially with a retaining clip-ring that surrounds the bearing (the outer race of the bearing has a groove for the clip-ring). That ring mechanically seats on the outside of the headstock housing. The belleville washer presses against the bearing as the bearing-cap screws are tightened. When the bearing-cap screws are fully tightened, the bearing cap seats against the housing, the belleville washer flattens, and the front bearing is pushed axially to the left with the clip-ring seating against the headstock housing.

I think the belleville washer is "pre-loading" the front bearing both axially (pushing to the left) and radially (pushing the outer race outward, taking up some slack in the bored hole in the headstock housing).

Correct me if I'm wrong, but I don't think the take-up nut has any affect on the front bearing "pre-load." The front bearing is trapped between the headstock casting on the left and the bearing cap on the right. Tightening the take up nut does seem to put the spindle in tension (it takes all the slack out between the various spacers, cone pulley, and gears along the shaft) but if the bearing cap were removed with the take-up nut still fully tightened, the whole assembly could still shift to the right in the headstock.

I'm thinking that tightening the take-up nut effectively "pre-loaded" the rear bearing somehow.

My guess is that when there were small gaps between the various sliding parts along the spindle, the rear bearing was able to slip and move a bit (about +/- 0.005") — probably because it was just a slip fit on the shaft so it had room to cant left and right slightly. Once I snugged up the take-up nut and eliminated all the gaps, the rear bearing was better retained in a perpendicular orientation to the spindle (no longer able to cant left or right).

At least that's the only explanation I can come up with. It wouldn't surprise me at all if I'm utterly misunderstanding something, though. I'd love to hear any corrections or better explanations.

Regards,
--
Rex

 

[wa5cab]

First, the Belleville Washer as drawn doesn't do anything that couldn't just as well have been done by the bearing cap if properly machined. The only thing that it accomplishes is to allow an axial force against the bearing outer race as you tighten the screws holding the cap onto the headstock. The washer cannot exert any significant radial force on the outer race. And turned the way that it is shown in your sketch, it cannot exert any axial force on the bearing inner race. The only Operator and Parts manual that we have is one on the 200 Series. It clearly shows a shoulder on the spindle close to the left end but inside of the headstock. A spacer abutts this and keeps the step pulley more or less from having axial motion on the spindle. The left bearing pulls up tight against the spacer. And then the gear and nut bear against the bearing.

 

[Rex Walters]

Well, I plead ignorance. My engineering education was mostly in imaginary things like semiconductors and software. This mechanical stuff I'm figuring out as I go along. ;-)

My memory is also even worse than my drawing skills. I tried to draw the assembly from memory (it's been a while since I pulled the spindle). The only major error was that I thought the front bearing pressed on from the right (the short end). I showed the shoulder on the spindle on the wrong side of the front bearing (the front bearing presses on from the left — the long side). I also left out the take-up nut for the front bearing (to the left of the bearing).

[I suspect it might be possible to tighten that front take-up nut further as well, but with the press fit I doubt it would make much difference, and I don't want to pull the spindle again to find out.]

For the record, here is the relevant diagram from the manual I ordered for $25 from Logan Actuator company:


That's a terribly confusing diagram, though, as it doesn't actually show how things are assembled (the line of parts in the middle are interspersed among the line of parts on top).

So here's my corrected drawing (if I ever pull the spindle again, I'm going to make a real measured drawing and maybe even a sketchup model).



At least I think that's correct. The front bearing presses from the left onto a slightly larger diameter portion of the spindle until it butts up against the shoulder. (The right face of the shoulder is actually what the chuck registers against). Then a take up nut prevents the front bearing from moving to the left along the spindle.

Between the press-fit, shoulder to the right, and take-up nut to the left, the front bearing is effectively welded to the spindle at that location. It can't move axially along the spindle left or right at all.

All other parts, including the rear bearing, are slip fits and can slide left and right along the spindle until they hit either of the take up nuts (or the headstock casting). The bull gear, of course, transfers

First, the Belleville Washer as drawn doesn't do anything that couldn't just as well have been done by the bearing cap if properly machined. The only thing that it accomplishes is to allow an axial force against the bearing outer race as you tighten the screws holding the cap onto the headstock. The washer cannot exert any significant radial force on the outer race. And turned the way that it is shown in your sketch, it cannot exert any axial force on the bearing inner race.

It's possible I put the belleville washer in backwards when I reassembled, but I still think I'm showing it in the correct orientation.

My (apparently faulty) thinking was that since the diameter of a belleville washer increases as it flattens, I thought it effectively provides a force vector that pushes outward slightly along the outer edge — so I thought it was supposed to bear on the outer race. The only way that any outward radial force could be transferred is through friction, though, so my thinking appears to be pretty fuzzy. I now think a belleville washer is just a glorified spring and is only intended to apply an axial force.

You seem to be saying that the belleville washer should be oriented the other way and bear on the inner race, but it isn't possible to bear on the inner race — it's blocked by the shoulder on the spindle. The inside diameter of the washer also registers on a lip on the inside face of the bearing cap. If the washer were oriented the other way, it would be almost impossible to get the registration right during assembly. All indications appear to be that I've shown the orientation correctly.

That front bearing isn't going anywhere along the spindle for the reasons I described above, and the entire spindle is prevented from moving left by the headstock casting and the clip ring. So I'm a bit puzzled as to the purpose of the belleville washer.

My only guess is that it's there to provide constant pressure (forcing the spindle assembly into the headstock) but still allow some axial movement due to heat expansion in the spindle assembly itself. That is, if the bearing cap was bearing directly on the bearing, then any expansion due to heat might cause distortion and inaccuracy. The belleville washer provides enough force to keep the entire spindle assembly located against the headstock, but with sufficient give to allow heat expansion.

At least that's currently my best guess, but I could be completely off base. What do you think? How hot do the bearings and spindle get in heavy operation?

The only Operator and Parts manual that we have is one on the 200 Series. It clearly shows a shoulder on the spindle close to the left end but inside of the headstock. A spacer abutts this and keeps the step pulley more or less from having axial motion on the spindle. The left bearing pulls up tight against the spacer. And then the gear and nut bear against the bearing.

You're right. I didn't have my spindle out, so I forgot about the step on the left end of the spindle. I searched for a photo of a disassembled Logan spindle, and RedlineMan had a good one with the front bearing still in place: see the fourth photo in comment #4 of this thread.

The inner spacer bears against this step. So when I tightened the rear take-up nut, I squeezed the spindle-gear, outside spacer, rear-bearing, and inside spacer (in that order) against that step. That makes much more sense than squeezing the entire series of parts along the shaft together, since the cone-pulley has to be able to rotate on the shaft when using back-gears.

But I'm back to my original question and confusion. How did tightening the rear take-up nut remove radial spindle play for me? I don't see any way that the rear nut could affect anything at the front bearing. It could affect the rear bearing (snugging it up between the two spacers on either side) but I'm not at all clear on how that cleared up my radial play unless it prevented left/right canting of the rear bearing as I suggested earlier.

I'm going with that explanation unless someone has a better theory.

Regards,
--
Rex

 

[hvontres]

I think what is going on here is that by tightening up the rear take up nut, you removed the remaining INTERAL clearance in the bearing. I was going to do a long writeup, but I figured the bearing guys are better at explaining this kind of stuff:
http://www.nmbtc.com/bearings/engineering/preload/
http://www.bardenbearings.co.uk/ind...//www.nmbtc.com/bearings/engineering/preload/
http://www.engineerlive.com/content/23694
http://www.nskamericas.com/cps/rde/xbcr/na_en/Preload.pdf

Hope this helps to clear things up a little.

 

[Rex Walters]

Hope this helps to clear things up a little.

Thanks, Henry! Very helpful. That's a lot of content. I'll spend some time trying to understand it. I'm still unclear on precisely what is flexing/stiffening and in what way such that it removes clearance (internal or external). Hopefully those docs will clear it up for me. I saw diagrams there. Diagrams are good. I need gorilla logic. ;-)

Cheers!
--
Rex

 

[Redlineman]

Hey;

Here is the image you are looking for.



You would assume that the tightening of that nut added no preload to any part of the spindle. Whether it stabilized, in some fashion, the rear bearing (drive side) or not by jamming it more tightly between the spacers and into the first shoulder machined in the spindle is an interesting notion, and perhaps covered in those linked documents. If I feel brave, I might dip a toe in. Tech stuff tends to make my head spin.