"Projects" at georgesbasement.com

Oufitting a Goodell-Pratt No.29-1/2 polishing lathe for screwcutting

   The original Goodell-Pratt No. 29-1/2 polishing lathe was supplied with a graver tool rest and items for buffing wheels and holding small round objects.    Somewhere I acquired an extremely dirty slide rest that looked hardly worth the effort of cleaning it up. Once I exposed the original unblemished, unused surfaces, it became a perfect match.
   Here's the Montgomery & Co. (New York) catalog image, page 199, ca. 1900, that is of this exact slide rest.
No. 29-1/2 polishing lathe in Goodell-Pratt Catalog No. 14, ca. 1920
Starting condition for the Goodell-Pratt lathe & slide rest
Montgomery & Co. image of the slide rest

   The gears shown below are all from my junk drawers, albeit with center bushings made from scrap bronze tensile specimens.  The shafts on which they run are all 0.229+ inch diameter.  I proportioned the banjo after the one I made for the Sebastian treadle lathe.
   The brackets & banjo that hold the gears were made from angle iron and plate steel about 1/4 inch thick. The T-nuts were made from bar stock.  The heavy bracket that fits the extreme left end of the lathe was machined from a block of steel and hand fitted by filing, as was the body clamp shown at right.
  The two clamping brackets are a fairly close fit to the lathe's standard, but what saves the situation is the elasticity of the heavy black paint, a.k.a. japanning.  The attachment is quite secure and does not wobble or deflect in use.
Outftting the lathe for longitudinal feed
Gear train
Body clamp that holds gear train onto lathe

   Shown below is the left-hand end of the drive shaft with its Woodruff keyseat cut.  The shorter thread is for a retaining nut that minimizes end play.  The longer thread is for the thumb nut that retains the screw gear portion of the gear train, so that it is easily changed for different feeds/threads.
   The Woodruff cutter is the 5/16 inch diameter, 3/32 inch wide size that is quite delicate, so I held it in the Atlas milling machine in a No.2 MT collet and clamped the drive shaft with a my own special-made, one-time-only holding device.
   The drive shaft has to be disengaged in a manner that lets me continue to run the lathe without disengaging the gear train. The pawl device fulfills this requirement. The brass sleeve has a friction device (two wire segments) in an annular space, so it stays wherever it's put. The spring lifts the pawl.
   I cut the pawl with a jeweler's saw in one go, with just a little filing to get it to function properly. It's made from 1/8 inch thick brass plate.
Left-hand end of drive shaft
Fixture for making Woodruff keyseat

Cutting the Woodrull keyseat with the Atlas milling machine
Drive shaft with pawl disengaged
Drive shaft with pawl engaged to drive the longitudinal feed
Drive shaft pawl partially cut with jewelers saw

   Presently I'm stymied by the lack of a gear cutter for 35 to 54 teeth, but I do have the cutters that cover 55 to 134 and 26 to 34 teeth; below I'm cutting the 56-tooth gear, using one I made for the Sebastian treadle lathe as a template.
   Really large gears would require that I lower the knee of the Atlas miller too far, so I made a fixture that allows the template and gear being made to rotate on a common vertical axis.  It's more work to raise and lower the knee in order to cut each tooth, though.
   The gear train below, with a 40-tooth stud gear and 64-tooth screw gear, uses a 56-tooth gear as an idler that also causes the thread being cut to come out as a right-handed 32 tpi, as the right-handed feed screw of the compound is 20 tpi.
Gear cutting on the Atlas milling machine
Cutting the 127-tooth metric transposing gear
Gear train for 32 tpi


   Also of note is the tool post, which incorporates a third slide so that I can feed the 60-degree thread cutting tool at 29 degrees to the transverse axis so that the chip is cut mainly on one side of the tool.

This lathe's compound rest has only two axes, and even some three-axis tool rests have no means for setting the single-point cutting tools feed axis at 29 degrees to the transverse axis as most machinists know to do in order to get clean threads.  The two tool posts shown here provide that third axis in a way that avoids the need for angular graduations on the third slide.
   Perhaps you are wondering how I deal with the task of setting the point of the cutting tool exactly at the center height of the lathe ?  That's taken care of by the Montgomery & Co.'s tool slide, which as such an adjustment built into the base.  Just remember to set that before aligning the longitudinal axis !
Tool post for external threading   The main clamping screw of the external threading tool post clamps the outer, channel shaped part of the tool post. The socket head cap screw clamps the actual tool and holds it down against a shim that's keyed to the thick washer.

   The feed screw at right pushes the tool towards the work at an angle of 29 degrees to the transverse axis of the coupound.

   The Vee of the single point tool is rotated 29 degrees from the feed axis so that proper setting of the tool with respect to the work also sets the feed axis correctly.
Tool post for internal threading  For the internal threading tool post shown at left, the third axis of the compound is created by a slide between the bottom plate of the tool post and the upper part that actually holds the tool.  The bottom plate is clamped with the main clamp, and the pair of socket head cap screws clamp the upper part to the lower part. Feed is effectuated by the socket head cap screw whose head appears at the lower right hand corner of the tool post.  The keys between the upper and lower portions of the tool post run in slots milled into both parts at 29 degrees to the transverse axis.

   As this is billed as a treadle lathe, I built a suitable table and tried to invent a foot motion that could coast without waving the treadle back & forth.  This annular drive fills that bill, as in the image below the center of the annulus is lined up exactly with the axis of the flywheel's axle.  The flywheel can rotate freely at this position of the treadle (shown as a mock-up only). Very hard to adjust ...
   The annular drive can operate in two modes: (1) With the treadle swinging in a wide arc so that the crank pin (running on Torrington needle bearings) travels in a full circle; or (2) With the treadle restricted to a swing truncated at the middle (freewheeling) position, which causes the crank pin to run in a 180 degree arc, reversing the swing of the treadle arm twice per revolution.
Annular drive
   I bored the two holes in the crank as exactly as I could and then made an eccentric sleeve to allow for adjustment on erection.

   I also had to allow for adjustment of the position of the annulus so its arc of swing would pass exactly through the axis of the flywheel axle.  That was by trial & error ... tedious indeed.

   Operation was very jerky at first and I only progressed to an annoying level of jerkiness before I gave it up.
Annular drive   The annular drive would have worked great, but for this fatal flaw: If allowed to swing in the full arc, it would randomly try to reverse the flywheel's direction of motion upon passing the center position if there was the slightest hesitation. That brought the flywheel to an instantaneous halt.

  So it had to be restricted to the half swing treadle motion, in which the treadle would swing with next to no resistance for 90 percent of the arc, so I had to press the treadle hard during only part of its motion.

   In the end I just made a simple crank & foot pedal treadle mechanism which works OK, as the lathe doesn't want to coast very long anyway, due to the friction in the plain-bearing countershaft. The shaft hangers are antiques in their own right.  At low speed the spindle turns at eight times the flywheel's speed; at the high speed, the ratio is about twenty to one.  I also made a numbered dial for the crossfeed; I will eventually do the same for the longitudinal feed.  The drive belt that's presently turning the spindle has a metal connector that's too large - it hangs up in the narrow clearance underneath the smaller spindle pulley. Alignment adjustments for the belts are easy but fussy; the leather belts came from my clothes closet.

   I made no irreversible changes to the Goodell-Pratt lathe;
I did make a new spindle with a 3/4-10 threaded nose to hold the four-inch faceplate that I made with thirty 1/4-20 tapped holes and six Tee slots. I replaced the handles on the feed wheels of the slide rest; I made no changes to the flywheel, although I had to get creative to clamp it onto its 3/4 inch axle. I did find it necessary to raise the tailstock by 0.015 inch (with brass shims) in order to improve its alignment with the headstock center.  I'm looking for a small chuck that I can fit to the nose of the spindle. The 30 by 30 by 30 inch bench knocks down to three main parts that are easily transported.  The lathe can be driven by hand from the little wheel at the right-hand end of the countershaft, which might make threading a little less stressful.
Portrait of finished Goodell-Pratt No. 29-1/2 treadle lathe