Chuck Run Out

T

topcat41468

Registered
Registered
Joined
Sep 5, 2020
Messages
43
I just completed my 7301 rebuild. leveled it and am generally happy with it. I have a bunch to learn so step one was to just get something in it and see if it actually cut. However, the 4 jaw chuck has a ton of run out. I have checked runout at several places to try to diagnose and am concluding its the chuck.

1. I could see wobble when I turned it by hand.

2. I put a dial indicator on it at 12 oclock just behind the adjustment hole, zero'd it on the first jaw and rotated it in the normal operating direction of the lathe. I got +0.010 to -0.005 (0.015 tir).

3. I re-did 2, but indicated off the outer diameter of the backplate. Same result.

4. I redid 2 but indicated off the hub of the chuck I get 0.004 total.

5. Indicated off the hub on the spindle and get 0.001, slightly less. Though its all in one quadrant (75% of rotation is ~0 and 25% runs up to 001 and back to zero.

I had a notion to teardown the chuck clean it, and then put the backplate on the lathe, remeasure it and turn it true to the lathe and face it. then do the same to the chuck. But a bunch of googling never found doing such a thing and that's a clue that I might be thinking wrong.
 
Normally you put the backplate on the machine, then turn it to fit the chuck. This insures that the register surfaces of the backplate are concentric to the machine spindle.

If after doing that there is still runout in the chuck, then some work on the back of the chuck may be required.
 
if my backplate hub OD is not concentric with the bore and the outer edge is not concentric, machining those to be square and concentric would be step 1 i'd think. but the mounting holes for the chuck to the backplate are not moving in this process, so while the backplate will run true, the chuck once mounted will still mount in a non-concentric manner. so I would think I'd need to turn the OD of the chuck also while assembled.

so my thought was:

mount the backplate on the spindle, and face it. Then rotate it around such that the hub was facing out and face the back of the hub and turn the OD of the hub. Then remount it normally and turn the OD of the backplate.

Then assemble the chuck to the backplate and turn the OD of the chuck to create one assembly that runs true.

bear in mind this is a 50-70 year old chuck.
 
I've read your posts at least a dozen times and I'm still confused. It might help if you gave us more info so we can maybe make a diagnosis:
  • Are we talking about a 4 jaw independent chuck or a 4 jaw scroll chuck?
  • What is the run out at the spindle, inside the spindle taper? What is the run out of your spindle register? While they would ideally be the same, wear on the register can mess things up.
  • Once you know the run out on the register, take a light skim cut on the chuck back plate so you have a reasonably accurate reference point for that plate and measure the run out on the rim of the plate. Ideally, the run out on the plate should be the same as that of the register. If it is not then you need to fix that.
  • Measuring the outside of a chuck is fraught with error. Many chucks, especially older chucks, have wear or damage from years of use and cannot be relied upon to be accurate. This is especially true if the chuck was not mounted to the back plate accurately. Mounting chucks to back plates is not hard to do but it does take some care. Personally, I measure the machined bore of the chuck, inside the jaws. Remove the jaws and clean the bore well and indicate on that, then report the run out.
What you are trying to establish is which component is causing the excessive run out because you have a stack of tolerances to sort through.

If your chuck is a 4 jaw independent, does it allow you to dial a work piece in accurately? The chuck may run out but if you can still dial in a work piece to within a tenth or two then that may not matter a whole lot.
 
thank you for your patience and help here.

I have a 4 jaw independent chuck.

the spindle run out on the register (the "flange" immediately behind the threads) is less than 0.001. I have not measured inside the bore of the spindle. I will do that tonight. I would expect the same, but I will get that.

I verified that the chuck bottoms out completely on the spindle register. The run out on the backplate at the hub OD (the machined and ID threaded boss on the back) is not the same as the register. Then the larger OD of the backplate is worse yet. The chuck body is the same as the backplate OD.

I will re-measure the bore of the chuck. That makes better sense to me, as that is the "center" of concentric parts.
 
Last edited by a moderator:
topcat41468 said:
I verified that the chuck bottoms out completely on the spindle register. The run out on the backplate at the hub OD (the machined and ID threaded boss on the back) is not the same as the register. Then the larger OD of the backplate is worse yet. The chuck body is the same as the backplate OD.
Click to expand...

I'm beginning to wonder if the female socket in the back plate that contacts the male register of the spindle is actually fitted well. It should be a precise match. When you install the chuck, wiggle the chuck as you thread it on; it will be loose until it hits the register. Just before it locks down there should be very little, if any, play between the register and the chuck. Once you lock it down there should be zero play. At that point, run out on the rim of the back plate should be zero if the rim of the plate was machined. If it is NOT zero then either the plate is not accurately seating on the register OR the rim of the back plate needs to be machined very lightly and accurately so you can measure run out on it.

When we install a back plate, the idea is to transfer the accuracy of the spindle to the back plate. Then we machine the snout of the back plate to fit into the female socket of the chuck with the accuracy we choose. Some prefer to machine the snout for a light press fit (0.005 - 0.001"), while others prefer to leave about 0.002" of play in that joint so they can tap the chuck true. Regardless, the key goal when installing the back plate is to get it as close to the accuracy of the spindle as possible.
 
Last edited by a moderator:
ok I have measured the spindle at several spots. Its easier to show the pictures than to describe. In all pictures I have zero'd out at one extreme and taken the picture at the other end.

image0.jpeg

Measured down the bore of the taper. got .006.

image1.jpeg

The face of the register. 0.0015

image2.jpeg

Top of the register 0.0015

image3.jpeg

The flat behind the spindle threads 0.007. This was hard to get clean reading. The DTI tip would catch and jump on the thread end a little. But 007 is pretty close to what I think its out.

These were all taken while hand turning ccw (toward me).

I think tis tells me that my taper is kind of crapy, but my spindle over all is 0015. Not sure what the last one means, it kind of surprised me being right next to the register and so different. I prob need to understand this before I understand what the chuck is doing.
 
Not sure but I'm guessing that your spindle uses the flat face of the spindle just behind the threads as the register, in which case the run out on that register is 0.0015". All else being perfect, this is the best you can expect from your lathe. That does not mean it cannot do better work; it means that's the best it will measure out.

Just to give you an idea of what a good lathe will measure at, generally accepted run out on a spindle is anywhere from zero to about 0.0002" or so. With that said, a lot of old lathes have run out much larger than this and can still do good work if the operator knows how to get that from his machine.

At this point, what I would do is to mount the back plate and take a very light skim cut on the rim of the plate to give you a reliable reference surface, then see what the run out is on the plate. This tells you how well the plate is mated to the spindle. If it is 0.0015" TIR then it is fitted as well as you can expect. If it is greater than 0.0015" then you need to surface the inside of the plate where it touches the register. You may need another lathe to do this.

Once you get the plate as accurate as you can, remount the chuck and measure the run out in the center of the chuck's bore. That will tell you how accurately it is mounted to the back plate. You may be able to loosen the mounting screws and tap it true if the fitment between the back plate and chuck allows. If not then you may need to machine the snout on the back plate to allow you a tiny amount of movement. Then align it as best you can and you have what you have. A 4 jaw independent should allow you to dial out 0.0015", I suspect.

To be very honest, I would try to do as much of your work with a 3 jaw chuck. As long as you don't take the work out of the chuck, the lathe will turn true to the spindle centerline, even with 0.0015" TIR. Try it and see.
 
And the similar measurements with the backplate on:


image2.jpeg
The register on the front of the backplate 007

image3.jpeg

The face of the backplate 014

image4.jpeg

the bore of the backplate. 006. This is not a finished surface, there are a couple burrs on the edge.

image0.jpeg
The hub of the backplate 010. I can see a height difference as the spindle is turned between the surface of the register and the hub. The bore of the hub is not concentric to the OD of the hub.
 
You must log in or register to reply here.
Back
Top