Spindle Bearing preload pm1340gt

[JBowlin]

I've been doing some gunsmithing with my new lathe and another smith ask me today if I had ever checked the spindle for end play...I replied that I hadn't since the lathe was fairly new I assumed this was set at the factory. He went on to say I should check it by chuck some small bar stock in the chuck and tightening the tail stock chuck on the same bar,put an indicator on the face of the check and run the quill in and out and note movement of indicator.

I did this with an interapid .0005 indicator as he said it should be less than .001 and boy was I surprised when I seen .0035 each direction in and out.

So following the instructions for tightening the spindle bearings I moved the lock ring about a quarter inch radially and let the lathe run at 1100 Rpms for 20 minutes. Bearings were at 76* F and room was 62*. Checked play again and it seems that I made zero progress in eliminating any of it.

I'm I on the right track or am I chasing my tail??

Thanks in advance.

Jb

 

[pdentrem]

In the shop at work, we let the two spindles in our Landis grinder run over an hour before starting anything. The surface grinder is turned on and left on all day. As the equipment warms up, the slop in the system is slowly reduced and/or eliminated.
I would run the lathe for an hour and keep checking the temperature as you do so. At some point peak temperature will occur, then check. Another thing is the bearings may not be completely settled in the casting or on the spindle and you may still be taking up that with your adjustments. If you tighten the adjustment too much, the bearings will run hot and the casting area where the bearings sit will naturally feel hot. Warm is good, hot is bad.

BTW new bearings will run hotter for a time and change the oil a couple times in those first few hours of use.

 

[JBowlin]

I've had the lathe for a little bit so it's not "brand new" but less than 6 months.

Total tonight I ran the lathe for an hour only stopping to remove rear bearing cover check the temp directly on the bearing , making adjustments and rechecking movement. I have now moved the nut over an inch radially and no improvement on play and hottest temp I got was 94*. I do have a correction on the speed though...thought I was at 1100 rpms but I was at 780 I think. I had it as fast as it will go with the belt in slowest position.

Speaking of Oil, I do need to change it again but can't remember for the life of me what oil to use.

 

[wrmiller]

I've had the lathe for a little bit so it's not "brand new" but less than 6 months.

Total tonight I ran the lathe for an hour only stopping to remove rear bearing cover check the temp directly on the bearing , making adjustments and rechecking movement. I have now moved the nut over an inch radially and no improvement on play and hottest temp I got was 94*. I do have a correction on the speed though...thought I was at 1100 rpms but I was at 780 I think. I had it as fast as it will go with the belt in slowest position.

Speaking of Oil, I do need to change it again but can't remember for the life of me what oil to use.

Mobil DTE, or equivalent, Heavy/Medium circulating oil (from the manual). 3.5 Quarts.

Just a suggestion, but if you are not running your lathe for an hour to warm it up prior to using it, then I wouldn't suggest doing so to take measurements. I go out, fire mine up and run it right away. I'm not going to stand around for a half hour or more to let the machine warm up. And if I let it sit for an hour before using it again I have to warm it up again. Nope. Not going to happen.

Run it cold, measure it cold. But that's just me.

 

[JBowlin]

Mobil DTE, or equivalent, Heavy/Medium circulating oil (from the manual). 3.5 Quarts.

Just a suggestion, but if you are not running your lathe for an hour to warm it up prior to using it, then I wouldn't suggest doing so to take measurements. I go out, fire mine up and run it right away. I'm not going to stand around for a half hour or more to let the machine warm up. And if I let it sit for an hour before using it again I have to warm it up again. Nope. Not going to happen.

Run it cold, measure it cold. But that's just me.

Remembered the oil shortly after I posted and already ordered it, but thank you.

I agree with your logic but so far checking cold and adjusting cold and then warm and adjusted warm ...neither have yielded any improvement

 

[Clock work]

It is my understanding that a surface grinder and it's high-precision spindle bearings are in a different precision class from those found on lathe spindles. If this is indeed the case, the thermal equilibration phase will do more for the surface grinder than the lathe. Thinking about what occurs as a spindle warms up, you get a slow movement over long time frames (minutes)... not things that cause impulsive (hundredths of a second to occur within a cycle of rotation. On the surface grinder, I can see this causing an error that is significant relative to the rated accuracy in the grind of a part that takes minutes to complete... the first several passes taking a different bite than the last passes. Relative to turning and facing operations on the lathe, and considering the accuracy of the lathe bearings, shorter window over which the error can grow through warmup and perhaps not significant when viewed thru the lens of the lathe bearing's rated accuracy. But setting thermal aside completely for the moment...

If you want to know the spindle runout, just measure the spindle if at all possible. There's significant opportunity for other effects to be creeping in to your measurement... a cascade of error sources are potentially being attributed to the spindle: spindle error + chuck error + work piece shape error (is it actually round?) + work piece registration error including whip.

I check my spindle by removing the chuck and indicating the spindle directly. The outer surface of my chuck adds more error to this. When I chuck up a piece of stock or drill rod, I see more, though these sources are within my control to some degree. Nearly every piece of stock I put in my 3-jaw needs to be convinced to run out minimally... multiple minutes of indicating the piece and ensuring it's chucked up parallel and concentric to the axis of rotation.

Perhaps an interesting analogy... in my real life, I own a company that sells WICKED precise electronic measurement tools. With 1000's of customers, and multiple decades of being in that business, I have met at MOST one guy that had the know-how to hit/use the asymptotic repeatability of our tools... i.e. operator error frequently dominates the total error term. Good luck!

 

[JBowlin]

It is my understanding that a surface grinder and it's high-precision spindle bearings are in a different precision class from those found on lathe spindles. If this is indeed the case, the thermal equilibration phase will do more for the surface grinder than the lathe. Thinking about what occurs as a spindle warms up, you get a slow movement over long time frames (minutes)... not things that cause impulsive (hundredths of a second to occur within a cycle of rotation. On the surface grinder, I can see this causing an error that is significant relative to the rated accuracy in the grind of a part that takes minutes to complete... the first several passes taking a different bite than the last passes. Relative to turning and facing operations on the lathe, and considering the accuracy of the lathe bearings, shorter window over which the error can grow through warmup and perhaps not significant when viewed thru the lens of the lathe bearing's rated accuracy. But setting thermal aside completely for the moment...

If you want to know the spindle runout, just measure the spindle if at all possible. There's significant opportunity for other effects to be creeping in to your measurement... a cascade of error sources are potentially being attributed to the spindle: spindle error + chuck error + work piece shape error (is it actually round?) + work piece registration error including whip.

I check my spindle by removing the chuck and indicating the spindle directly. The outer surface of my chuck adds more error to this. When I chuck up a piece of stock or drill rod, I see more, though these sources are within my control to some degree. Nearly every piece of stock I put in my 3-jaw needs to be convinced to run out minimally... multiple minutes of indicating the piece and ensuring it's chucked up parallel and concentric to the axis of rotation.

Perhaps an interesting analogy... in my real life, I own a company that sells WICKED precise electronic measurement tools. With 1000's of customers, and multiple decades of being in that business, I have met at MOST one guy that had the know-how to hit/use the asymptotic repeatability of our tools... i.e. operator error frequently dominates the total error term. Good luck!

Thanks for your reply and some very good points, but I'm not checking spindle runout. What I'm checking is movement in and out of the chuck within the bearings it rides in. The runout was checked on the spindle the day I set the lathe up and I can't remember the number but I do remember being impressed..lol

 

[jbolt]

Just for giggles do the same test but place the indicator on the face of the head stock at the same level as it was on the chuck.

You could also mount a TDI to the chuck with a mag base and measure off the face of the head stock while you do the push/pull test.

 

[Bob Korves]

I check my spindle by removing the chuck and indicating the spindle directly. The outer surface of my chuck adds more error to this. When I chuck up a piece of stock or drill rod, I see more, though these sources are within my control to some degree. Nearly every piece of stock I put in my 3-jaw needs to be convinced to run out minimally... multiple minutes of indicating the piece and ensuring it's chucked up parallel and concentric to the axis of rotation.

Agreed. Try to do your best to check the bearings, not the flex of other stuff between the bearings and your indicator. You can not check the bearings directly, but you can check the spindle which is directly connected to the bearings. As jbolt pointed out, mount the indicator to the headstock, not a train of stuff sliding on the ways. Eliminate all the middle men, as much as possible...

 

[Mitch Alsup]

What I'm checking is movement in and out of the chuck within the bearings it rides in. The runout was checked on the spindle the day I set the lathe up and I can't remember the number but I do remember being impressed..lol

The Grizzly manual for my 12-36 lathe specifies that you are to remove the chuck(s) and measure the plate that holds the chuck.
http://cdn2.grizzly.com/manuals/g4003g_m.pdf
page 75
Once you get all of the play out, add 1/16" (3-odd degrees) turn to set preload. Install chuck, run for a while, and check again.