making the lathe more rigid
Stops & locks
A carriage lock is built in to the SC4
lathe. It is engaged by tightening an M5/0.8 socket-head cap screw (Sieg part #220) on the
carriage. Tightening this screw pulls the Carriage Lock Block (Sieg part #231) up
against the underside of the bed. Depending on the setup and position of
the compound slide, access to the screw may be difficult.
The space in which a comfortable handle might be added to the screw is
limited. Also, although not much rotation is needed to bring the screw
from (loose enough) to (tight enough), a handle not capable of 360° rotation
would need to include a screw of exactly the right length, with
adjustments presumably necessary from time to time to account for wear.
Drawing on some ideas that I found in this
forum, I replaced the carriage-lock screw with a new screw that carries a
sliding T handle. The Sketchup designs are here. and
the finished part in situ looks like this:
Trying to move the carriage against the
resistance of the carriage lock is not a good idea, and one might conclude that
if loose is good, looser must be better. This is a mistake. If the
Carriage Lock Block falls off the screw, it may vanish into the depths, requiring radical
disassembly and reassembly of the saddle. To reattach the block to the
screw, one must first remove the two M4/0.7 SHCSs (not numbered) just to the right
of the carriage-lock screw; this will free the two-hole Rear
Clamp (not numbered) whose right-hand face is seen at the right side of the saddle.
Use a magnet to remove the Rear Clamp, and use the same (long) magnet to guide the Carriage Lock Block into position for
its screw.
If the carriage-lock screw is tight but the
carriage is still easily mobile, the carriage-lock block may have turned itself
sideways. This can't happen if the screw is never made too loose.
Designs for
carriage stops are easily found on the Web. One carriage stop should be
made to carry a dial indicator. With a simple stop toward the headstock
and an indicator-bearing one toward the tailstock, boring to a fixed depth is
just a matter of
-
picking up the surface with the boring bar;
-
locking the right-hand stop close enough to the
carriage so that the indicator engages the carriage;
-
zeroing the indicator;
-
moving the cross slide so that the boring bar
clears the work;
-
advancing the carriage toward the headstock until
the indicator shows the desired depth;
-
bringing the left-hand stop next to the carriage
and locking it; and
-
repositioning the cross slide and boring,
advancing the carriage toward the headstock until it runs into the stop.
Stops for the cross slide and compound are not quite
as useful, but the effective rigidity of the lathe turns out to be strikingly
improved when simple locks for the cross slide and compound can be
engaged when these components should not move. The gib screws for the
cross slide and compound can be tightened to achieve locking, but the screws
provided with the lathe (Sieg #218 and #245) are shoddy, and their straight-slot
heads make adjustment clumsy. I replaced all the gib screws with hardened,
socket-headed setscrews (M4/0.7 × 16 mm for the rear of the
carriage, M5/0.8 × 30 mm for the cross slide, and M3/0.5 × 16
mm for
the compound). With the new screws in place, gib adjustment for locking
and unlocking is easy.
Before attacking the gib screws, I had
improvised locks for the
cross slide and compound, relying on the existing threaded holes for the follow
rest and for the compound rotation mechanism, respectively. These
improvisations are now superfluous.
The mechanism that advances the lathe's
tailstock quill is graduated in thousandths of an inch, and I have no reason to
doubt its accuracy, but its backlash makes the scale almost valueless when a
pecking technique is used for drilling. To drill to controlled depth, I
use a collar that fastens to the quill and projects downward to where it can run
into the carriage; a drawing for the collar is here. With
the collar in place, drilling to fixed depth involves
-
picking up the surface with the drill bit;
-
locking the tailstock;
-
moving the carriage to run up against the bottom
of the collar;
-
locking the right-hand stop close enough to the
carriage so that the indicator engages the carriage;
-
zeroing the indicator;
-
advancing the carriage toward the headstock until
the indicator shows the desired depth;
-
locking the carriage; and
-
drilling, advancing the quill until the collar
runs into the locked carriage.
Coping
with the 0XA quick-change tool post The
quick-change tool post most widely suggested for use with the SC4 lathe is the
0XA size; for example, see here.
This QCTP is not adequate. The tool-post mounting hole on the SC4 lathe is
M12/1.75, so the puny M6/1.0 central screw of the 0XA post must be fattened with
an adapter before the post can be attached. With only this feeble
attachment, it's difficult to keep the 0XA tool post
from rotating away from its set orientation on the compound slide. Before
I decided to get rid of the 0XA QCTP (see below), I developed a variety of
coping strategies:
1.
Sometimes one can do away with the tool post and/or compound slide altogether.
boring
without the tool post or compound slide
One occasionally
has reason to bore a tapered hole, but a boring bar should usually be parallel to
the spindle. I built a boring-bar holder that replaces the compound
slide. It is a one-trick pony, allowing no rotation about any axis except that
of the spindle, and no height adjustment (the tool-holding hole was drilled by
mounting a drill chuck on the spindle, so its height is guaranteed). A
drawing (shared with the drawing of the QCTP mounting block described below) is here; the
photograph shows this fixture in place on the cross slide.
a
rotary
broach that avoids the tool post and compound slide
Construction of a rotary broach,
based on Mike
Cox's design, is another application in which
the compound slide can only get in the way. I started with another
2-inch-square tower screwed to a base plate. With the base plate tightened
down to the cross slide at 90° and a drill chuck in the spindle, I drilled the
hole in which the cutters would spin, and then I reversed the tower to drill the
grease-escape hole. I didn't need the height-adjustment screw shown on
Cox's site, nor — because I machined the
2°-angled cutter faces on the mill, using a 2° angle block and a 5C collet in
square/hex collet blocks — did I
need the locking screw that is present in the original design. I
didn't need to machine the sides of the tower to any specific angles at all. To skew the cutter to 1°, I simply twist the base plate as far
counterclockwise as it will go before tightening down the nuts that hold it to
the cross slide. As it turns out (the mounting holes are 0.358" ID;
the T-bolts are M8/1.25), the limits of motion before the nuts are tightened are
almost exactly ±1°.
mounting
the 0XA QCTP without the compound slide
The compound slide can be rotated
360° on the cross slide, and the tool post can similarly rotate on the
compound slide. Sometimes one wants the convenience of the QCTP, but the
two potential rotations are troublesome. When parting, knurling, or using a form tool, the lathe must be
configured to eliminate all these degrees of freedom, so that the tool is
exactly perpendicular to the spindle. I
constructed a mounting block to support the 0XA QCTP in a fixed orientation.
The block is attached directly to the cross slide in place of the compound
slide. A drawing is here;
the photographs show the block with and without the QCTP in place.
2.
When the tool post and compound slide must both be used (for example, for
threading), measures can be taken to increase the rigidity of the selected
configuration.
tool-post
positioning pin
Unwanted rotation of the 0XA tool post is reduced by the Positioning Pin,
Sieg part #238. This small part is spring-mounted in a well in the
compound slide, from which it pushes a tooth up against the base of the tool
post. I discovered that the
Positioning Pin of my lathe had been lost, probably fired away by its spring
during manipulation or replacement of the tool post. I believed
(correctly) that it would be easy to fabricate another, but I was uneasy in
the absence of any sort of drawing or description (the pin is shown on Sieg's
assembly drawing of the lathe, but as little more than a smudge). I
appealed to various Web-accessible sources of information, and I am happy to
acknowledge the help of Luc Morin, who went
to the trouble of dissecting his own C4 lathe and producing this elegant
drawing. I made the replacement pin of mild steel. Along the
way, I was struck by the uncanny similarity of shape between the Positioning
Pin and this old xray anode that I happened to have around the house.
bracing
the QCTP against unwanted rotation
To resist forces that might tend to rotate the
tool post out of its orientation, I produced a simple brace that can be bolted
to the compound slide. I made mine of 0.375" mild steel.
Installation
of an AXA toolpost
My friend Alain Aubry showed me that an AXA QCTP can easily be fitted to
the SC4 lathe, with immediate improvement in the rigidity of the system.
There are two necessary adaptations before the AXA toolpost can be attached and
used:
-
The AXA toolpost (at
least the
one I purchased) is provided with an attachment shaft that is said to be
9/16" OD, but in fact it is an M14/1.5 threaded rod. To attach
the toolpost to the SC4 compound slide, I produced an M14/1.5 shaft that was
turned down and threaded M12/1.75 for a centimeter or so at one end.
-
The AXA toolholders
are beefier than their 0XA counterparts. When one of these toolholders
is resting on the top of the compound slide, some tools will have their
cutting edges above the lathe's center of rotation. Following Alain's
lead, I milled relief,
down to the level of the bottom of the T-slots, into the compound slide as
shown here:

The
AXA toolpost is more rigid than the old 0XA post, but it is liable to forced
counterclockwise rotation when, for example, a cutoff tool must be extended to
truncate a large-diameter workpiece. When using a cutoff tool, the
compound slide is noncontributory, just as it is in the boring-bar case
described above. Mounted on the cross slide in place of the compound
slide, a simple frame (seen here bare and then with the QCTP mounted on it)
provides solid protection against unwanted rotation.
Home
Page revised: 01/28/2019 21:39
|