What's new
What's new

How much spindle nose runout is normal?

CompositesGuy

Aluminum
Joined
May 30, 2009
Location
AZ
I have a South Bend 13" lathe with the D1-4 camlock spindle. I purchased my first new chuck (a 6" set-tru, 6-jaw Bison) and was paying more attention to runout than I ever had during setup of this chuck. I measured the total indicated runouts on the spindle nose shown in the attached picture (under power at low RPM) and was surprised at how large the face runout was (0.001"). Is that normal? I'm not sure how one axis of runout could be so good, and the other one so off?

To set up my new chuck I suppose that I could take a skim cut on the front face of the chuck adapter plate to compensate for the spindle runout, then always install the adapter plate indexed to the same orientation on the spindle. That would keep the face of the chuck (and jaws) nicely perpendicular to the axis of rotation. Unfortunately, that solution wouldn't help my other chucks which are direct mount with no adapter plate (a Sjogren 5C collet chuck and an 8" 4-jaw chuck). I really like that Sjogren chuck and want it to run true also, now that I've noticed the issue. Does any cut/true-up their spindle nose? (Kinda seems like a risky proposition). Am I overthinking this?
 

Attachments

  • runout.jpg
    runout.jpg
    106.4 KB · Views: 41
  • Chuck.jpg
    Chuck.jpg
    452.5 KB · Views: 39
Proper fitting of an adapter plate requires facing as you describe. With a 4-jaw chuck it doesn't matter much as you're going to be indicating to zero anyway, so a direct-mount chuck is fine.
 
I measured the total indicated runouts on the spindle nose shown in the attached picture (under power at low RPM) and was surprised at how large the face runout was (0.001").
I'd probably never even know this, since I make a serious effort to do everything in one chucking but if it came to my attention, and was a thou, yeah. Go get the bench grinder, make a plate, mount the plate to the cross-slide, put a white wheel on the grinder, dress, and kiss the face of the spindle. .001" is more than I'd want to see.

I would not try hard-turning it. Those holes are not going to make an insert happy. Even if the insert lasts, the face won't be flat.

And every lathe I've ever had with camlock spindle had a mark at the #1 pin, so you'd always put it on at that position. Your six-jaw is adjust-tru but still. Doesn't hurt to always put stuff on in the same position.

This'll give you something to do over the weekend :)

(Check the front bearing on the lathe before you get too carried away, just in case).
 
What resolution is that indicator? Your first photo shows it on the taper. If the nose had a wobble, that would remain nearly nothing, while indicating off the face would show the wobble. A tenths indicator will give you a better idea of the taper running out vs the face.

How does that whole spindle nose assembly attach to the actual spindle? Have you had that apart? Checked the physical spindle itself? I have no idea how that comes apart but, from other internet pictures, it looks like it's not one piece with the physical spindle. That suggests something could be wrong in that mate (a burr, a chip, or?).
 
I'm pretty sure that the spindle is one piece with the mounting nose, both from observing the inner bore and from this old parts list. Assuming that the runout I see was not there at the factory, I'd think any crash-induced damage severe enough to cause my 0.001" axial wobble on the face would induce far worse radial runout on the tapered nose. I'm assuming that crash induced damage would treat that nose end of the spindle like a cantilever beam held at the bronze bearings. The absense of such radial runout has me scratching my head.

For what its worth, this lathe is from the 1980's and in fairly good shape otherwise. There is very little radial slop in the headstock bearings, and I've adjusted the axial (thrust bearing) slop to about 0.001. Both slops were measured with the lathe off at the nose, in the normal manner. I have not removed the spindle or had the headstock assembly apart. I don't see surface finish evidence that the nose has been recut in the past. This is really the first bothersome thing that I've discovered about the lathe in 15 years of ownership.
 

Attachments

  • partslist.jpg
    partslist.jpg
    138.3 KB · Views: 12
Last edited:
Been thinking. My main concern here is this: If I leave the spindle face this way, my direct mount Sjogren 5C chuck will turn slightly conical parts. This wouldn't be the case for fully turned parts cut in a single chucking setup. But if I turn the part around or partially machine some part where I'm counting on the axis of the 5c collet to be parallel to the axis of rotation, I'd have the issue.

As stated before, I can skim-cut the adapter plate to compensate for the spindle issue with my new 6 jaw. However, I'd like to really understand the issue, and perhaps 'fix' the spindle before cutting into the adapter plate. Even adapter plates are expensive these days.
 
As mentioned above it is customary to make a facing cut on a backing plate as standard procedure fitting a chuck. Clocking the chuck every time it is installed is probably also a good practice when runout needs to be minimized. A replacement spindle is almost unobtainable at this point and while a back plate isn’t inexpensive, they are still being manufactured.

That doesn’t help the direct mount chucks you have though. Are you sure it’s the face of the spindle nose that is not flat? Could there be crud in the thrust bearing that is randomly “pushing” the nose forward? Is the high spot always in the same location or does it move? If it’s moving, probably not the face.

What is the runout of the important surfaces of the 5C and 4 jaw chucks? I would measure those before grinding the spindle.

I would also use a tenths indicator as mentioned above.

Hope this helps

Ben
 
Try snugging down the axial thrust nut a bit. I've got a 10L with a d1-4 spindle. When I inspected it before buying it the spindle runout was as perfect as it gets. It was in a basement so everything came apart to get it out, and when I put it back together at home the spindle runout had become worse. I tightened the axial end play to less than .001" and the runout was fixed.
 
Thanks Hudstr. That was it. I had my axial spindle play set to ~0.001" (via the thrust bearing adjustment nut) and I have one of those modern drive belts that is a sort of plastic/fiber on the outside and some sort of synthetic leather-like material on the inside. Many years ago I had to glue the splice together on that belt and I didn't get it perfectly straight. That little wiggle in the belt was pulling the spindle slop left and right with each belt revolution. I tightened the thrust bearing up to about 0.0005" of axial spindle slop (with the lathe off), and my measured axial runout at the spindle nose (under power) improved to about 0.0002". I backed the axial slop back off to about 0.0008" and left it since South bend recommends leaving the slop at .001" or greater. Its nice to know that my spindle isn't bent like I feared. Maybe someday I'll get a new belt and glue it up straight... but really any cutting force during a facing operation would probably keep the slop pinned to the left.
 
Last edited:
I mounted the new adapter plate and took a facing skim cut, then centered the 6 jaw-chuck set-true feature and did a bunch of checks with different size ground pins. Everything seems to operate with Bison's advertised tolerance specs... so I guess I'm done fiddling for now.

In case someone reads this in the future, I was careful to find the index mark on the spindle (a very, very, faint horizontal short line) and line that up with the Bison badge on adapter plate and the pinion (chuck key location) marked with a "0" next to it on the chuck. I will always install the chuck onto the spindle in that orientation going forward, and always use the "0" pinion to tighten the work when accuracy matters. (I've seen elsewhere that the "0" pinion is the one that Bison uses when manufacturing the chuck.)
 
Last edited:








 
Back
Top