Signs Of Bad Spindle Bearings?

wa5cab said:
....You cannot adjust the preload on a double-row bearing.
Click to expand...

As a Logan owner, you obviously know a lot more than me about this problem :)

I have an old manual for my Sheldon lathe, which is not dissimilar to a South Bend or Logan. In the manual, three different headstock configurations are shown:

  • sleeve bearings, oldest configuration
  • three ball bearings (one in the rear, two in front) transitional configuration
  • two tapered roller bearings (one in the rear, one in front) newer machines

The ball bearing configuration, at least on the spindle shown, appears to be adjustable by doughnut shims under the bearing caps at each end of the headstock. Note that I'm NOT claiming that the Logan can be adjusted in the same manner but it is possible. Removing the bearing cap from the rear of the headstock would be a good place to start - see if any shims are visible.

eeler1, too bad you moved, I would have been happy to take a look at your machine and could have brought my Millwright buddy along. He runs a job shop in Eureka and there's not much about machinery that he doesn't know !

The last test you made sounds a lot more realistic. When you say you got chatter as you moved further away from the chuck, do you mean that the end of the work was unsupported ? If that's the case - no tailstock center or steady rest - then that's just normal flexing.

A good rule of thumb if you don't use a center or steady is that the length of the work protruding from the workholder shouldn't be greater than about 3 times the diameter.

One last thought: Your 3-jaw chuck might be bell-mouthed. If that's the case, you'd get chatter when the end of the work is unsupported, probably even when cutting close to the chuck. You'll either need a new chuck or to grind the jaws true.
 
I've been told that spindle bearings on the vert mill would warrant adjustment if the fully extended spindle could be deflected by more than .001", was hoping for a similar rule of thumb for the lathe. (Although not adjustable in this case). I got a quote from Logan Actuator for my model and ser#, and mine apparently has one bearing fore and one aft, so two total. Parts diagram for the two bearing configuration shows a couple of bellevilles in the fore end, for pre load I would guess.

Anyway, I'll hold off and use the machine for a while, see how it performs on real work instead of just test pieces. I have other stuff besides bearings to spend money on.
 
eeler1 said:
I've been told that spindle bearings on the vert mill would warrant adjustment if the fully extended spindle could be deflected by more than .001", was hoping for a similar rule of thumb for the lathe. (Although not adjustable in this case). I got a quote from Logan Actuator for my model and ser#, and mine apparently has one bearing fore and one aft, so two total. Parts diagram for the two bearing configuration shows a couple of bellevilles in the fore end, for pre load I would guess.

Anyway, I'll hold off and use the machine for a while, see how it performs on real work instead of just test pieces. I have other stuff besides bearings to spend money on.
Click to expand...

Lots of folks say that the 2 x 4 test should barely move the needle on a .0005 indicator. My 1944 Sheldon flunks that test but still does good work and doesn't chatter under reasonable speed/feed conditions. I just re-measured spindle movement since I hadn't done it in a while. I saw DTI movement of .001 (cold), just like your measurement.

Hmmm, the description of the bearings suggests either angular contact ball bearings (good) or tapered roller bearings (best). Both are better than standard ball bearings. Belleville washers are a common means of pre-loading but I've not heard of them used on lathe spindles before. Oh well, whatever works, right ?

I think that your plan to use the machine for a bit makes perfect sense. And if you are still having problems, maybe try replacing the Bellevilles before springing for a new set of bearings. Might even add another Belleville if there's sufficient room to accommodate it :) And definitely check that 3-jaw for bell-mouthing !
 
Last edited:
Randy,

I actually own an Atlas 3996, not a Logan. My familiarity with ball bearings comes from another life, "long ago and far away".

His original report stated .025" on a 1" bar about 8" in front of the chuck. Not .001".

eeler1,

You should probably run one more test before you decide to live with it for a while. Remove the chuck. Get a piece of solid round or heavy wall tubing 18" to 24" long and of a diameter that is just small enough to easily slide through the spindle. Set up your indicator on the top of the spindle register. Slide the round bar into the spindle about full depth, push down on it firmly and zero the indicator. Pull up on the bar and read the indicator. The figure is close to the actual front bearing clearance. It isn't exact because (a) you are measuring outside of the bearing and (b) the rear bearing (at the other end of the spindle) may have clearance, too, which will allow the spindle to tilt. But if the figure is less than around .012", you may also have some movement in or of the chuck.

Also, in most cases, you can't add just one Belleville washer, you should add them in pairs.
 
Last edited:
Ah! I finally found this thread again. I knew I saw it somewhere, but I thought it was on the Logan lathe mailing list.

I've discussed my tribulations in this thread: Excessive spindle play fixed.

Short story is that a year or two ago I bit the bullet and bought (expensive) new bearings directly from Logan Actuator (probably unnecessarily). I've still got concerns with deflection, though, as I'm getting seriously bad chatter when parting.

With a stout drill bit held directly in the morse taper of the spindle, I now get about +/- 0.0004" of deflection at the spindle nose itself which doesn't seem too excessive. I'm applying about 10 to 20 lbs of force (guesstimate) on the drill bit, about 6" in front of the spindle nose.

When I put a chuck on, though, and repeat the test with a stout bar held in the chuck, I get about +/- 0.002" deflection at the chuck jaws with light pressure (5 to 10 lbs at a guess) and up to +/- 0.005" deflection with heavy pressure (about 20 lb).

This happens even with my lightest, most compact chuck, so I'm starting to wonder if the problem isn't with the bearings, lack of preload, or wallowing out of the bore in the headstock, but rather with the registration of the chucks onto the spindle itself.

I'm going to check the registration of the chuck onto the spindle today with some hi-spot blue markup. I'll post my findings shortly.
--
Rex
 
Rex,

If you get one deflection when you use a fixed length 3MT bar (in this case a drill bit large enough in diameter that its deflection is probably out of the equation) and a greater deflection with a similar diameter (but solid) bar mounted in any of your chucks, it is a safe assumption that either the bar is moving in the chuck jaws (bell-mouth problems) or the chuck is moving on the spindle nose.
 
Thanks, Robert.

My thoughts exactly. That's why I took some bluing to the spindle to see if the chucks were bearing on the reference shoulder fully. I didn't see anything obvious (they appeared to be registering all the way around).

I found something interesting when I moved the mag base from the saddle to the headstock casting, though. I got considerably less indicator movement (just a few tenths — sufficiently small that I no longer have much concern about internal or external clearance at the bearings).

I decided to be a lot more methodical/calibrated/repeatable with my investigation, so I set up several measurements using a setup like the following:

IMG_0064.jpg

The 3/4" drill has a MT2 shaft inserted into an adapter sleeve inserted into the spindle. The DTI is measuring as close as I can get to the reference shoulder on the spindle and has the mag base attached to the headstock. The luggage scale is about 8" away from the reference shoulder. I also tested with the original small 3-jaw Logan chuck, as well as a Gator 8" chuck holding some 1" drill stock (with the luggage scale still 8" from the spindle shoulder).

Here are my results with a 40 lb pull on the scale:

With the drill, measuring the spindle directly: 0.0003" (three tenths)

With the 3-jaw measuring at the outside rear of the chuck body itself: 0.0003"

With the 3-jaw at the outside front of the chuck (about 1 3/4" from the spindle shoulder): 0.0008"

With 1" drill rod in the 3-jaw measuring near the jaws: 0.0013"

With the heavy 6-jaw measuring at the outside rear of the chuck: 0.0004"

With the 6-jaw at the outside front (about 4 1/8" from the shoulder): 0.0017"

With 1" drill rod in the 6-jaw at the jaws: 0.0016"

For such an old, light-weight lathe I think these are pretty reasonable values. Over a 1 thou deflection at the chuck jaws is a sleight concern, but I suspect I'll have to live with it.

Now to figure out why there is such a large discrepancy when the indicator base references from the cross slide. All I can figure so far is that either the ways are actually twisting/deflecting a bit relative to the headstock (not much I can do about that other than buying a larger lathe) or the headstock is moving on the ways slightly. If you can think of any other possibilities, I'd like to hear them.

I think I've pretty much ruled out the chucks moving on the spindle nose — they are definitely pulled up tight on the registration shoulder, and the values from the rear of chuck bodies are pretty close to the direct spindle measurements. I don't think I've got bell-mouth problems, either, since the at the jaw figures are pretty close to the body front measurements.

When it comes down to it, though, the measurement that matters with respect to chatter is with the indicator at the jaws and the mag base on the cross-slide. I was seeing on the order of 5 thou of play when measuring this way.

I'll continue the investigation when I'm next able to get into the shop.

Thanks for the reply!
--
Rex
 
You must log in or register to reply here.
Back
Top