Acceptable lathe runout?

[Joe Pack]

Tonight, I came home from a trip to Rockler toting a 3" Don Pencil face plate. I put it on the lathe and spun it up. Knowing the face plate was CNC milled, I was curious as to just how "perfect" it really was, so I pulled out my magnetic base dial indicator and checked it out. To my surprise, I found a variation of .008" as I rotated the spindle by hand. Thinking the problem was in the face plate, I pulled out a decent caliper and found absolutely zero error...the face plate was perfectly round.

So, I took off the face plate and set up the dial indicator to measure run-out on the lathe itself. I found a "wobble" of .003" at the point where the spindle enters the headstock, just behind the starting point of the thread. According to my caliper, the spindle itself is milled round...no variation at any point, but there is .003" of wobble from something.

Obviously, I am not terribly sharp with tolerances on tools. Is this .003" wobble typical and acceptable? Remember, it translates to .008 on a 3" face plate, and would be even more on a larger face plate or chuck.

Will this cause me problems? If so, I want to get it taken care of while the lathe is new. If not, I need to look for another excuse for crappy turning.

[Bill Howatt]

Joe,

If I understand you correctly, you are measuring the wobble with the gauge against the circumference of the faceplate rather than the face?

If so, I don't see how this is scaled up by the diameter of the plate. However, if you are measuring on the face then the diameter definitely scales the wobble. A 12" diameter object would wobble 0.018" or about 1/64".

My General 26020 has virtually undetectable wobble at the spindle on either surface (the circumference or the face the faceplate butts against). With the General 3" faceplate, the circumference wobble is less than 0.0005".

You might try putting a blank on the plate and remeasuring to see if loading the bearings takes any slack out. Also, you are measuring statically, and perhaps under spinning it "goes away"?

Bill

[Joe Pack]

Joe,

If I understand you correctly, you are measuring the wobble with the gauge against the circumference of the faceplate rather than the face?

If so, I don't see how this is scaled up by the diameter of the plate. However, if you are measuring on the face then the diameter definitely scales the wobble. A 12" diameter object would wobble 0.018" or about 1/64".

Thanks for the reply Bill.

I didn't think to measure against the face. This is new stuff to me. I will do that before I go to bed.

I measured against the circumference of the spindle, then against the circumference of the face plate. My thinking was that the scaled up measurement (.003 to .008 to ???) would be the result of carrying an angle out over a longer distance. If I imagine a "V," the farther I get from the bottom of the V, the farther apart the sides get. The angle remains the same, but, the distance between the sides increases with the distance from the intersection of the lines.

Maybe this isn't even a problem. Again, this is a new area for me and I may be overthinking a minor thing.

Thanks again for the input.

[Bill Howatt]

I measured against the circumference of the spindle, then against the circumference of the face plate. My thinking was that the scaled up measurement (.003 to .008 to ???) would be the result of carrying an angle out over a longer distance. If I imagine a "V," the farther I get from the bottom of the V, the farther apart the sides get. The angle remains the same, but, the distance between the sides increases with the distance from the intersection of the lines.

Maybe this isn't even a problem. Again, this is a new area for me and I may be overthinking a minor thing.

Thanks again for the input.

I agree with your V analogy applied to the face of the plate and this certanly scales up the error measured on the surface of the faceplate. I may be wrong (has happened before) but I see the 0.003 measured on the circumference of the spindle still being 0.003 when measured on the edge of a perfectly round and centered faceplate's circumference measured on a line perpendicular to the spindle. If you have a stick and move one end 0.003 then the other end moves 0.003; it doesn't matter if the stick is 1" or 10" long.

I don't think the circumference tolerance is the critical one. The face wobble is considerably more critical IMO since, if I am right, it scales up with the diameter whereas the circumference error doesn't.

I wonder if if the 0.008" error you measured isn't due to the spindle error (0.003) plus an error in the faceplate as mounted to the lathe. I am not saying the faceplate is out the extra 0.005" by itself but perhaps it isn't locating itself exactly on the center of the spindle.

Bill

[Chuck Wintle]

Could it be the headstock bearing has become slightly worn thus allowing a run out condition? Why not take your new face plate to a machine shop and have the run out checked.

[Joe Pack]

Could it be the headstock bearing has become slightly worn thus allowing a run out condition? Why not take your new face plate to a machine shop and have the run out checked.

Hi Charles.

The lathe is brand new, as am I to using it. The second part of that sentence is why I'm curious.

[Joe Pack]

You may be right about the face plate surface being most important. Before I went to bed last night, I set up the dial indicator against the face of the face plate instead of the circumference and read a variance of .003".

Your stick analogy made me think a bit, which I try to avoid when possible. :D My logic went more along the line of a wobbling shaft that has a fixed point at one end, much like a moving line in a marching band...if a straight line is turning around a fixed point (the bearings?) the person beside the fixed point moves only a little bit in the turn, while the person on the farthest end from the fixed point moves a far larger distance in order to keep the line straight...

Lordy, conceptual discussions tire me out easily! :D

Anyhow, I'm probably making much ado about nothing...sometimes a little knowledge is dangerous in newbie hands. Thanks for being willing to help.

Joe

[Chuck Wintle]

Sorry I missed the part about the lathe being new :o I was just wondering what brand your new lathe is? Maybe they can replace the spindle with one that will run true.