Centering Rotary Table

[Christianstark]

And the plot thickens.
I indicated the center hole as I spun the table. Couldn't see any movement in the needle at all.
Frustrated, I even made a new longer dowel. Didn't help.
Now, I'm starting to think that maybe the hole is tapered just slightly, so that I'm hitting it in some wonky way.

 

[Christianstark]

Do you have a coaxial indicator?

 

[markba633csi]

I wouldn't start cutting on the RT- your measurement or setup technique is suspect unless proven otherwise

 

[GT-6 Racer]

Certainly agree. Don’t cut on your table until you are positive about your measurements. It’s unlikely that the center hole is damaged unless visibly so. To check fit for your shop made center do a blue check. Should be even band of contact.
A DTI can be used on the surface plate to check concentricity of center hole and outer edge of the table rotating it with the crank. Same for parallelism of top surface as it’s rotated. This eliminates tram and other uncertainties from putting it on the mill. If the outer edge of the table is suspect, which it may be, take a large bearing race and toe clamp to the table and use the DTI to center it with rotation of the table while on the surface plate or bench. Now you know you have a true circular part concentric with table rotation. Put it on the mill. Sweep the spindle mounted DTI to center the bearing race on the spindle, now you can sweep the center hole to see what’s going on with it. Write all your
Deviations on the location on the rotary table with a sharpie to keep track. Sounds tedious but should be too bad.

Good luck sorting it all out !! Hope it works out that the table is in good shape

 

[homebrewed]

If your dowel is cocked over the amount of runout will vary as you move the indicator up and down along it. The RT doesn't even need to be centered in order to do this test.

Running a DTI up & down the center hole will tell you if you have much of a taper in there. Bluing up the dowel would be an even more sensitive test, and would be independent of an out-of-tram condition.

 

[Shotgun]

BINGO! @homebrewed. The dowel was indeed cocking to the side, because it wasn't long enough.
I made another dowel. This one about twice as long. The other one only reached about halfway into the collet. I also used a flat stone to "grind" the dowel to take out any ridges.
I got the table to less than a thou on the first try with the new dowel.
I want to thank all you gentlemen from talking me off the ledge. I'd be real sad right now if I had bored a new, out of round hole.

 

[Shotgun]

Lesson learned: An ER-32 collet needs the part it is holding to bottom out. A half full collet is a door for error.

 

[GT-6 Racer]

Glad you got it figured out!

 

[WobblyHand]

Lesson learned: An ER-32 collet needs the part it is holding to bottom out. A half full collet is a door for error.

If I need to turn a short piece in an ER collet, I will use a section of rod that's the same diameter and stuff it in the back of the collet. I try to use the same piece of rod as the original piece. The filler piece can be up to a few inches long, so you can use cutoffs. This seems to work for me. I mike the diameters so they match up well. Hope that helps.

 

[Christianstark]

If I need to turn a short piece in an ER collet, I will use a section of rod that's the same diameter and stuff it in the back of the collet. I try to use the same piece of rod as the original piece. The filler piece can be up to a few inches long, so you can use cutoffs. This seems to work for me. I mike the diameters so they match up well. Hope that helps.

Same!