PDA

View Full Version : What would cause this?



SMG
07-14-2010, 09:02 PM
What would cause my SB 9 lathe to turn a 0.010" taper over a 7inch piece of bar stock using a live center? http://i1041.photobucket.com/albums/b412/smgraph/th_DSCF2950.jpg?t=1279155517

Thanks Steve

PeteF
07-14-2010, 09:07 PM
Offset tailstock.

reggie_obe
07-14-2010, 09:14 PM
Twist in the bed.

Willy
07-14-2010, 09:14 PM
As PeteF said offset tailstock.
Or perhaps loose bearings in your live center.

Dr Stan
07-14-2010, 09:24 PM
all of the above plus an off center center hole, a very dull tool, or a too loose live center.

To help diagnose the problem you need to turn a piece between centers.

PeteF
07-14-2010, 09:47 PM
To help diagnose the problem you need to turn a piece between centers.

Isn't that what he has just done?

The easiest way to check the tailstock is correctly centred is to chuck a rod with a DTI in the chuck. Bring the tailstock up to the DTI and ensure the tailstock is locked to the bed and the friction clamp on the quill is locked. Swing the DTI around the tailstock quill. The DTI should read precisely the same left to right, but will read slightly different top to bottom even on a perfectly adjusted lathe. It is the side to side reading you should be concerned about at this point. If the 2 readings aren't exactly the same you will turn a taper between centres. To adjust, unlock the base and adjust the screws on the base of the tailstock. Re-lock and check. This is an easy check and the most common reason for turning a taper between centres.

Pete

SMG
07-14-2010, 09:49 PM
Isn't that what he has just done?

The easiest way to check the tailstock is correctly centred is to chuck a rod with a DTI in the chuck. Bring the tailstock up to the DTI and ensure the tailstock is locked to the bed and the friction clamp on the quill is locked. Swing the DTI around the tailstock quill. The DTI should read precisely the same left to right, but will read slightly different top to bottom even on a perfectly adjusted lathe. It is the side to side reading you should be concerned about at this point. If the 2 readings aren't exactly the same you will turn a taper between centres. To adjust, unlock the base and adjust the screws on the base of the tailstock. Re-lock and check.

Pete


Thanks Pete, I'll give it a try.

SMG
07-14-2010, 10:01 PM
Pete, I think your right. The tail stock seems to be out by about 0.008. I'll try to fine tune it and see if thats the issue. thanks again. Steve

PeteF
07-14-2010, 10:25 PM
Let us know how you go. I'd suggest some of the suggestions above won't in fact cause a taper, but one that may is loose bearings in the live centre. There is a certain gap in bearings, albeit small, that allow them to turn. The pressure of the tool will cause the piece to be forced back and cause a taper (again small unless the bearings are totally crap) just as if the tailstock was slightly off. If you're after ultimate precision you probably want to turn between 2 dead centres and drive the work with a dog. Personally I think it's nit-picking, but the way the lathe geometry is set up that's just the way it works.

Pete

darryl
07-14-2010, 11:03 PM
-offset headstock-

Evan
07-14-2010, 11:11 PM
Make sure the grub screws for the top slide rotation are tight. The first models of the SB9 only had one screw on the right rear and it wasn't enough to securely hold the compound. There is another on the left back side and if it isn't secure the tool can move during turning.

Also make sure that the rear under carriage gib is properly adjusted so the carriage is securely held down on the Vee ways.

Take off the top slide and you will find a bronze insert behind the compound socket. In that hole is a slotted screw. Tighten it so that the cross feed nut isn't loose in the cross feed casting.

When testing for taper lock the cross slide gib so it cannot move at all. Do the same for the top slide after selecting the depth of cut.

The headstock cannot be offset. It is scraped in to the same vee ways as the tailstock and clamped down by a couple of bolts. It isn't adjustable.

dp
07-14-2010, 11:12 PM
Isn't that what he has just done?

Nuh uh. He turned between a chuck and a live center.

Evan
07-14-2010, 11:17 PM
The tail stock seems to be out by about 0.008. I'll try to fine tune it and see if thats the issue.

Put a centre in both ends and bring the tailstock up to the head stock. Place a piece of feeler guage stock between the points after you get them close to being in line. The slightest deviation from dead on will make the feeler guage stock pivot very clearly away from perpendicular to the long axis.

PeteF
07-15-2010, 12:47 AM
Nuh uh. He turned between a chuck and a live center.

Ah yes, sorry quite right. However what difference will it make to determining a taper whether a centre or chuck is used at the headstock end? Provided of course the work isn't removed and replaced or slipping in the chuck (something I was fighting with a few days ago).

Evan I use that method too, but think the DTI is a lot more accurate, especially if the centres aren't ground quite concentric or seated correctly in the bore.

fciron
07-19-2010, 10:05 PM
However what difference will it make to determining a taper whether a centre or chuck is used at the headstock end?

If the chuck jaws are not aligned perfectly then the axis of the work piece is no longer in line with the spindle axis. This can flex the work, the tailstock, or the center in the tailstock.

Turning between centers eliminates this problem. Even if your headstock center was inaccurate the work would still rotate around the same axis as the spindle.

darryl
07-19-2010, 11:44 PM
Of course. Once you turn off some of the OD, the workpiece becomes concentric with the spindle axis. Doesn't matter if the chuck jaws are off center, or they're holding the workpiece off center. As long as the jaws don't shift during the turning, the workpiece will become centered automatically. Supporting the outboard end in a tailstock center would automatically flex that end off center if the tailstock wasn't aligned properly, or if the bed was twisted.

Turning two collars without the support of the tailstock would generally indicate whether the headstock is out, or if the bed was twisted. This procedure is not without error- if the workpiece flexes away from the cutting edge, or digs towards it, this test would lose some validity.

I don't know that lathe, but as Evan has suggested, the headstock is scraped in and there's no adjustment- in most any case that would mean that the headstock is aligned, and short of having sustained physical damage, it would be fair to just assume that it's good. My order of procedures then would be to frist check and adjust for bed twist, then move on to the tailstock adjustment. First get it rotated so its axis is parallel to the spindle axis in the horizontal plane, then shim left or right as required to make it parallel in the vertical plane. Part of this shimming also gets it to the correct height. After that, move it forwards or backwards to bring the tailstock axis to be the same line as the spindle axis. Now when you run the ram in and out, the center should remain on-axis.

Once you have the bed twist adjusted out, you can turn a point on some scrap and consider that to be the spindle axis indicator to measure against. If you also turn two or three collars and measure them, this is a tool to help test the straightness of the bed. Others have suggested placing a disc or a rule- whatever- between two points to show in which directions the discrepancy between the two axis exists. Works great, use a fresh spot on the disc each time you pinch it between centers for a test.

I guess the big thing with me here is if you try to force the workpiece into an alignment using the two centers, it will be constantly flexing the workpiece as it rotates, and in some cases will force the workpiece to become loose in the jaws. This is not the right thing to do, even if it does mean you can make a shaft less tapered. Solve the 'turning a taper' problem first (headstock alignment and bed twist), then align the tailstock to a freshly turned point at the chuck end.

Notice that I have not included any discussion of bed wear problems. This can easily make the other adjustments futile to a large extent. You need to know that your carriage can ride the bed without going through convolutions.