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

Thank you, John! Fantastically helpful info there (and you guys are eventually going to shame me into painting my lathe!). Both of you will end up with machines almost too pretty to use.

I'm new to the forum and I'm still catching up on all the great content here. I've just spent some time reading posts you mentioned: MBFrontier's rehab thread, your rescue thread, and JST's posts on the practicalmachinist site (especially this one).

[Off topic, but from the photo in comment #157 on your thread it appears you have flat, scraped ways? My Wards 2130 has vee-ways as well as flat that don't appear to have ever been scraped. Did you do the scraping?]

Earlier today, I spent some time carefully measuring the radial movement when I pull on the spindle (or on a chuck with the spindle) with a reasonably measurable amount of force (I used a hook type luggage weight scale to pull on the bar). I'd post photos, but I just loaned my son my phone with the pictures on them (he's visiting from overseas).

Unsurprisingly, there was more movement with a chuck than with a morse-taper drill bit stuck directly into the spindle. With about 40 lbs of pulling on a 3/4" bar about six inches from the spindle (a pretty heavy lean backwards, imagine hefting a suitcase weighing just under most airline's weight limits) I could get up to 0.0008" of deflection at the spindle. Interestingly, I noticed that with my 6-jaw chuck I got a lot more deflection when I measured the bar deflection right at the jaws (about 0.005") than when I measured on the outside of the chuck itself (which showed the same 0.0008" deflection as at the spindle). Clearly there is some internal movement of the jaws at work — this was enlightening.

BUT I did this measuring with the mag base on the cross-slide (hey, I have all this nice flat cast iron back there now, I want to use it!). I just read that the proper way to measure this deflection is with the mag base on the headstock casting itself, because the ways themselves can actually twist/warp slightly with this kind of pulling (makes sense to me).

Also, I'm unsure, but it sounds as if the new front bearing I bought from Logan Actuator is actually a "a double row internally preloaded ball bearing." Per JST, "[internally preloaded] means it was made with zero to negative 'clearance', and there was no 'rattle space' as there often is in a standard ball bearing." Yet more reason why the Belleville washer at the front is unnecessary.

Amazing how much I'm learning on this site.

More updates soon as I continue my investigation, but I'm increasingly confident that I'll get things stiff enough to eliminate the chatter, even if I have to resort to JST's pre-load mods (which looks to be a pretty big job). From his posts, by the way, JST definitely sounds like someone I'd enjoy spending a few afternoons with — he really seems to know his stuff.

Since I only have one lathe, it's kinda like re-building a plane as you're flying along. Without another lathe to refer to, I'm unsure what's "normal" and I'm loathe to tear apart my lathe again because I can at least turn parts on it now, chatter or no!

Regards,
--
Rex
 
Ha!

What you lack in direct Logan experience you more than make up for in candor and bravery!! Most of us are in the same boat on here; just groping along with no real experience to draw on. We get by on what we read elsewhere, and the hard won wisdom we share here, to a large extent. I have no real machine experience, but a lot of peripheral related experience with mechanical things, metal, welding, fabricating, etc. Others have definite technical skills that I do not. Some are just bold and brave. All tolled, between the lot of us, we might add up to one great machinist if we're lucky!
 
Rex,

I think that all of your most recent comments WRT preload are correct. I will add that the concept of taking out axial and radial play by preloading a pair of ball bearings is somewhat iffy as you drawings with exaggerated clearances illustrate. But with two opposing ball bearings deliberately preloaded, you do reduce the possible axial and radial float.

However, as the drawings (especially those produced by Redlineman) show, you cannot reliably preload the spindle bearings in at least the 10" logan (the only one that we have seen drawings of). The left hand spindle bearing's inner race is in fact locked in place on the spindle. But the only thing resisting axial movement of the outer race is friction between the bearing outer race and the bore in the headstock. There are no shoulders for the outer race to bear against.

Back to the Belleville washer that was added to the 200 prior to 1947, the only that that I can see it doing is taking out the clearance between the snap ring (circlip) and the groove it fits in on the bearing outer race. However, this a several thousandths improvement in limiting axial movement. It has no effect on radial movement, though.

On the right spindle bearing, this is a double row bearing. I have read quite a few discussions on the subject of whether or not the current bearings as supplied by Logan still have the built in preload between the two rows. Apparently vintage Logan advertising claimed built-in preload between the two races. Or at least so it has been written here.
 
Thanks, Robert. That's helpful and makes total sense to me. I hadn't thought about the (axial) clearance in the snap ring groove.

The only thing I didn't understand was:

wa5cab said:
I will add that the concept of taking out axial and radial play by preloading a pair of ball bearings is somewhat iffy as your drawings with exaggerated clearances illustrate. But with two opposing ball bearings deliberately preloaded, you do reduce the possible axial and radial float.
Click to expand...

I'm not following. When I read that it sounds like the second sentence says the exact opposite of the first. Could you clarify?

Thanks!
--
Rex
 
Last edited by a moderator:
Rex,

What I meant was that if you are dealing with two radial contact (the most common type) ball bearings, pushing the outer races toward each other (or away from each other) will eliminate essentially all of the end float but only reduce the radial float The reason is that the inner diameter of the outer race and the inner diameter of the inner race do not overlap. So if you drew a line through the contact points between a ball and its inner and outer race, it would not be far enough off of a line parallel to (in this specific case) the spindle axis. So it is relatively stable parallel to the spindle axis but unstable at right angles to that. For bearing preload of a ball bearing pair to approach the effectiveness of a tapered roller bearing pair, you have to be using a special type of ball bearing called an angular contact type. In these, the contact angle is typically 45 degrees, so they are stable (resist motion) in both the axial and the radial directions. See crude drawing (made from your drawing because I can't draw circles worth a flip).

Angular Contact Bearing.jpg

Two of these, back to back and properly pre-loaded against each other, are stable. For relatively light duty applications, they are a reasonable substitute for a pair of Timken bearings.
 
In my correspondence with JST, the OP of the rear bearing mod that I used, I learned that he had converted his lathe to 3 phase power in addition to the rear bearing mod and asked him the following question:

If you were to rate the improvement of the rear bearing mod and converting to three phase power what percentage improvement would you assign to each of these two changes to eliminate chatter?

His response was:

For me, the bearing preload mod made it useful again. As good as it was before. I don't know how you rate a change from "Scrap that POS", to "Hey, I can USE this".


The 3 phase made it much MORE useful, it made it fully "industrial" for me. Chatter is totally manageable by standard remedies, the belts slip less, and a slip does NOT end with the belt coming off as it ALWAYS did before.

In neither case was there any comparison between before and after.... it was miles better.

After my rear bearing mod there was a noticeable improvement. I could take a .020 cut on mild steel without the lathe chattering and got a nice finish. However, parting is still an impossible task on this lathe. I have been able to make several parts that turned out with a nice finish but use my band saw for cutting and face the part on the lathe.

I have never had an issue with the belt coming off.

I have been considering converting my lathe to 3 phase power using these items from Automation Direct:

GS2-21P0
GS2 1.0 HP AC DRIVE 230V 1/3 PH IN 3 PH OUT

GS-21P0-FKIT-1P
FUSE KIT FOR GS1/GS2/GS3-21P0 1-PHASE (FUSES INCL)

MTR-P75-3BD18
AC MOTOR 0.75HP 1800RPM 56C 208-230/ 460VAC 3-PH ROLLED STEEL

LR-21P0
LINE REACTOR 230V 1HP 3PH DRIVE INPUT OR OUTPUT, 3% IMPEDANCE

GS-CBL2-3L
CABLE GS KEYPAD TO DRIVE 3m (9.9ft) FOR REMOTE KEYPAD MOUNT

The above items total $425.00 so I've been staying away from my lathe for a while as I consider whether to spend more money on this project. I may have to do it to see how much better 3 phase vs. single phase is on my machine. However, I know it won't transform it into a one ton lathe.

Anyway, I think I'll get away from this computer before my wallet gets $425.00 lighter.
 
wa5cab said:
In these, the contact angle is typically 45 degrees, so they are stable (resist motion) in both the axial and the radial directions.
Click to expand...

Awesome. I finally understand what an angular contact bearing is (and your drawing was what let the penny drop for me).

Am I correct in assuming they should be oriented on the shaft correctly (not flipped left-for-right)? Seems like the single bearing you've drawn would be better able to resist axial spindle thrust to the left than it would thrust to the right. I can see why a pre-loaded back-to-back pair would be stable and resist axial thrust in either direction.

Remember I started off completely clueless about bearings! :) Are "Timken bearings" shorthand for tapered roller bearing (like "Kleenex" for tissues) or does that manufacturer just make particularly high quality bearings? Timken appears to manufacture all kinds of different bearings.
 
MBfrontier said:
I think I'll get away from this computer before my wallet gets $425.00 lighter.
Click to expand...

Thanks, Mike. I understand that concern! <laugh>

3-phase motors are nice, but if you've got a reasonably quiet and vibration free single phase motor, I wouldn't change it. I'm probably missing something, but for a belt driven lathe I can't see it making much difference (at least with respect to chatter, but even for power transfer it seems like belt-driven is belt-driven). In my case, changing from a leather clip-lace belt to an automotive serpentine belt was a huge win (quieter and far less slippage) that was a lot cheaper (and I believe more meaningful) than a motor change. (That said, I did replace my motor with a new one but mostly due to filth and concerns with the rotted out wiring.)

I think JST's spindle play was much worse than mine before he made his mod, so his lathe was completely unusable without it. It sounds like he got replacement bearings that absolutely needed to be pre-loaded to provide any kind of accuracy, so his mod was necessary. It appears the new bearings I got from Logan Actuator don't have that problem, so I've no plans to make the mod.

Regards,
--
Rex
 
Rex.

Like Redlineman, my lathe was made in 1942 and there was no Belleville washer utilized in that vintage by the factory. I spoke with Scott Logan before I ordered my new bearings and he said the bearing cap should have .001 to .002 clearance between the spindle cap and headstock after the spindle cap is tightened to insure that axial play is eliminated. I achieved that by using shim stock instead of milling my spindle cap.

After looking at your drawing which shows your Belleville washer I was thinking I could have tried to preload the front bearing by reversing the washer orientation and installing it between the take-up-nut and the bearing. Perhaps multiple washers (if there is room) would have accomplished the same outcome on my lathe without adding the extra bearing. I'm not sure there is enough room on my spindle to do that and I won't be taking the spindle out to find out.

I'll have to remember these ideas if I ever restore another Logan 200 Lathe but I'm thinking once was enough. It has been fun though.

Good luck with working on your lathe and I'll be interested to see how you make out with yours as you progress.
 
Rex Walters said:
Awesome. I finally understand what an angular contact bearing is (and your drawing was what let the penny drop for me).

Am I correct in assuming they should be oriented on the shaft correctly (not flipped left-for-right)? Seems like the single bearing you've drawn would be better able to resist axial spindle thrust to the left than it would thrust to the right. I can see why a pre-loaded back-to-back pair would be stable and resist axial thrust in either direction.

Remember I started off completely clueless about bearings! :) Are "Timken bearings" shorthand for tapered roller bearing (like "Kleenex" for tissues) or does that manufacturer just make particularly high quality bearings? Timken appears to manufacture all kinds of different bearings.
Click to expand...

Rex,

The bearing that I crudely drew would in fact only work as one of a pair, with the second one flipped over and installed to the left of the one drawn. The one drawn will take axial thrust from the right but not from the left. Unlike radial contact (or just "radial") ball bearings, it is possible to manufacture an angular contact ball bearing with the inner and outer races separable. I don't recall ever using any and don't know whether the balls would normally remain in the inner or the outer race but would guess the inner.

Yes, "Timken" was used like "Kleenex" to mean tapered roller bearings. Timken has for nearly a century manufactured high quality tapered roller bearings. And in fact I can't ever recall seeing or using any bearing made by them that wasn't a tapered roller type. In any case, that's where they made their reputation. They don't (or at least claim not to) sell ANSI-ABMA Class 4 bearings.
 
You must log in or register to reply here.
Back
Top