Model Forum / General / Models / July 2007

 Chucks wear unevenly.

 All the reconditioning that anyone can throw at one, can at best, make
it relatively accurate for one particular diameter bar. For the other
diameters, not so much.

 If you can get hold of a couple different diameters of accurate bars,
you can get a better idea of what diameters are more accurate.

 Look at how the chuck seats on the spindle. A little swarf at the
register will result in rather a lot of runout at the chuck face.

 Look for damage as well as for solid seating. The burr raised by a
trapped chip can cause some difficulty long after the chip was cleared out.

 The chuck should thread on smoothly, with little play once the
register (the cylindrical area behind the threads) comes into the
backplate or the chuck body.

 Spend less time with the engineers level, and more time taking test
cuts. Stop fiddling with the leg screws when the lathe turns a
reasonably parallel cylinder, rather than a cone. Precision levels are
pretty much overused, and underutilized. Quite useful if you need every
last drop of coolant to run down the drain, but often taken far too much
as the final authority when setting a machine up. How it cuts matters
rather more than how level it is, though level, or near to it, is a good
place to begin. For the most part, a decent carpenters level will do.
 Look on the net for "Rollie's Dad's Method" for setup and adjusting of
feet. Pretty much just a way to read a test cut and understand the cause
and effect of adjustments.

 Now would be a good time to get used to dialing in a 4 jaw chuck. For
whatever reason, the process seems to scare grown men as much as having
to cut threads with a singlepoint tool. Both are easy processes that get
easier with practice. I like to set my dial up with the stylus on the
center of the work, on the back side of the work from me. I can then
adjust the jaws on the front side, and see the movement on the dial at
the same time. A mag base mount on the cross slide is the goods for that
job! Allows you to adjust the position relative to the work, without
having to deal with the friction joints of the mount each time.

 Also! Don't get too hung up on having to achieve perfection before you
can use the lathe. It's supposed to be recreation. If you want a dead
nuts perfect lathe as a hobby, you can have one. If you want a lathe
that cuts accurately enough, you probably already have one of those, too.

 Cheers
  Trevor Jones

Brad Hi, firstly don't panic this is only the first of the
"interesting" issues that using a lathe will throw at you just relax
and enjoy the challenge!

Lots of good advice already but a couple of things spring to mind, .
005" on a three jaw chuck is not good but also not unheard of, they
are not the most accurate method of workholding. You will get more
accurate results with your 4 jaw even over the most accurate three
jaw.

However, to try and improve the three jaw; is it mounted on
"backplate" or directly threaded? If it has a backplate then loosen
the mounting bolts a little and try tapping the chuck to reduce the
error, if you can reduce it then retighten the bolts. If the chuck
will not move on the backplate then you need to take the chuck off the
backplate, fit the backplate on its own and check on the register
diameter where the chuck body fits onto it. If there is any movement
on this small flat face then you will need to machine the register
again whilst fitted to your own machine as well as taking a small cut
of the large flat face. If it is directly threaded then correction is
almost impossible unless we can find something wrong with the mounting
register or thread.

The vertical surface you should be checking is not the one right at
the front but the one just behind the smooth portion (register) on the
spindle (ie close to the bearings). If the register diameter in the
chuck is slightly oversize (very common) then this vertical surface
actually positions the chuck when tight, there should be no movement
at all. Make sure it is absolutely clean and with no burrs etc. Check
the vertical surface within the chuck mounting that fits against it;
again it should be absolutely clean as just a speck can throw things
out. One thing I have had with some chucks is that the very front of
the spindle thread is actually jamming against the chuck body, make
sure there is clearance for the very front of the spindle inside the
chuck body when mounted on the backplate.

When you mount the work try it by tightening with it held gently
against each of the three jaws in turn using different key positions,
you will normally find that one is better than the others. This may
already be marked with a 0 or a centre punch mark. Look at the inside
of each jaw to ensure there is no small burr or swarf trapped between
the jaw and work. Most three jaws when older, even if reasonable, will
only repeat to about .003" (just in front of the jaws) so don't worry
too much if you can't get it absolutely perfect.

Best regards

Keith

I would start with the test bar, one that fits the spindle internal
taper, as this should be the most accurate datum. Then you can set the
headstock up for vertical and horizontal alignment with the saddle /
bed, using a dial gauge. Of course, the spindle bearing needs to have
zero meaureable play first to get accurate / consistent results. Next,
put a center in the tailstock end, bar between and use the same method
to align the tailstock. Then put the chuck on and see how accurate that
is. Of course, the spindle itself will have a runout and coning motion
on its internal taper, but this should be very small and there's not
much you can do about that anyway.

If you start with a 3 jaw chuck fitted, it just introduces another
variable and you don't really know if the spindle or chuck is in error.
Chuck jaws wear unevenly, tend to bell mouth and (most relevant here)
the error will be different every time you tighten the chuck up.

It's important to get fun out of all this, but I think you have the
right approach, especially with a machine received in an unknown state...

Chris

Brad, ignore the chuck for a minute. If Myford said 0.002" runout they
were probably correct.

Take the chuck off. Stick a block of wood in the gap on the bed then
get a broom handle between this and the spindle nose, and see if you
can lever up the spindle, and measure any deflection with your clock.

If may be that the spindle has some wear, but this can usually be
adjusted out. Underneath the caps that hold the bearings on is a
laminated shim - it looks like one piece but is actually a stack of
shims, each one 0.002" thick.
These can be peeled off using the the point of a scalpel or similar to
seperate them, and will take up any clearance caused by wear.

There is a stack on each side of the spindle.

Peter