"Projects" at georgesbasement.com
  

Ugrading the Atlas milling machine's indexing centers
Part I - Adapting template gears with bores smaller than 3/4 inch to the Atlas milling machine's indexing centers.
Adapter for South Bend Change Gears   Just as I did for my Sebastian treadle lathe, rather than making any permanent, irreversible changes to the Atlas milling machine's plain indexing centers, I made an adapter to enable use of the South Bend lathe's change gears (which mostly have a 9/16ths inch bore) as templates.
  

   The adapter consists of a plug closely fit to the bore of the indexing center's spindle and incorporating a tapered collar that matches the center bearing of the spindle, and which is clamped in place by the indexing center's nut (shown sitting on the Atlas MM's table).
Adapter for 100 tooth South Bend Change gear   The adapter also has a sleeve to match the bore of the gear being used as a template to the diameter of the adapter plug.
  

   In the image at left, the adapter is holding a 100 tooth change gear (needed for making metric threads on the South Bend lathe) onto the spindle of the Atlas milling machine's indexing centers. There is no key used - the friction of the 5/16-18 clamping screw is sufficient.  

   Note the stops that I added to the longitudinal table travel. held in place by the former gearbox screws.
Adapter for South Bend change gears   Another necessary part of the above adapter is an angle plate that holds the indexing pawl, which simply clamps to the table of the Atlas milling machine.  

   I made a new pin for the indexing pawl which has a 14-1/2 degree cone to match the pressure angle of the template gear.  The Atlas indexing centers come with a pin whose nose radius is a bit too large for the smaller gear pitches.
   
   The angle plate has three mounting holes for the pawl carrier so that it can be appropriately placed for any template within the range of the milling machine.
Making a 100 tooth gear with the Atlas MM indexing centers

   As seen in the image at left, even template gears which the indexing centers cannot swing over the table of theAtlas milling machine can be hung over the end of the table.
  

   The 100 toth gear being made here is 32 diametrical pitch, intended for my Goodell-Pratt 29-1/2 "treadle" lathe's screw threading adapter.
Solid arbor for gear cutting on the Atlas milling machine
   The 36-tooth gear shown in process at left is being made on an arbor which I made that would be stiffer than the tapered arbors used previously, which do not hold the gear blank securely perpendicular to the arbor axis and which deflect excessively in the smaller sizes, such as 3/8 inch.
   
   The arbor has a threaded extension sized to fit the gear blank's bore and a loose collar to facilitate making more than just one gear at a time and to place the threads of the arbor outside the bore of the gear blank.
Solid arbor for cutting gears in the Atlas indexing centers  
   The arbor is driven by a lathe dog and is carried on the two centers of the indexing head. 
 

   That's a 72-tooth template gear (from the set that came with the indexing centers) in the background. 
  

   This arbor is far stiffer than a simple tapered arbor and helps to prevent most of the chattering that is created by interactions between the arbor and the knee of the milling machine.
Part II - Freeing the Atlas milling machine's indexing centers from the need for templates.
First assembly of 80-tooth worm gear   While rummaging through my parts drawers in search of 32 DP gears, I came across this 80-tooth worm gear.  That's just twice the number of teeth in a standard universal index's 40-tooth worm gear, so I decided to put my found gear to use.
  

   The 5/16-18 tpi cap screw and plain washer did not hold the 80-tooth worm gear securely, so I used a 5/16-18 tpi stud and flanged nut from the workpiece clamping set for the Atlas miling machine instead.
Cutting the 35-tooth gear  The indexing plates came from a CRAFTS of NJ tailgating session, I think ... or from a Cabin Fever auction ... I made the crank from scratch, using the experience I gained from making a banjo for my Derbyshire lathe thread cutting adapter.

  Shown at left is the first assembly of the gear to the adapter which I made that permits the gear to drive the spindle of the Atlas milling machine's indexing centers without any permanent modifications.  The vertical cylinder is press fit to its base and carries a 3/4-5 tpi Acme-threaded worm in bronze bearings.
35-tooth gear and 36-tooth gear side by side
   The 36-tooth 32 DP gear at far left was made using a 72-tooth Atlas gear as a template, whereas the 35-tooth gear to its right was cut with the 49-hole index in the newly upgraded Atlas indexing centers using one and 7/49 turns of the crank executed twice per tooth.

   The gears mesh perfectly !
Still to come: 
 

I. A guard to protect the delicate 80-tooth cast bronze worm gear, being made by bending a piece of 3/4 inch square steel tube to a ca. 2-1/2 inch radius and then milling a 3/8 inch wide slot around its inner perimeter. The bending task is complete; I didn't break anything or get too sore.
  

II. An accessory to protect my fingertips while lifting the indexing pin of the crank. The present 1/8 inch cross pin's tips are too small for the rather strong spring that forces the pin into the circle of holes in the index plate. 
 

III. I already made a clamp that prevents inadvertent movement of the indexing sector, which can be disastrous when it occurs during cranking of the worm while gear-cutting.
  

IV. Now I need a bunch of 5 inch diameter indexing plates ...
Part III - The promised extras.
Making the gear guard   The gear guard started out as a 16 inch long piece of 3/4 inch square welded tubing, which I was able to bend cold with the aid of a cheater bar and a piece of heavy steel tubing that was just the right diameter around which to form the more-than-180 degree curve of the guard.  A bar clamp allowed me to complete the bending task, as it let me bend the tubing with my own hands without great effort.  Of course, I filled the tube with fine sand and hammered in oak plugs to keep it inside. One of those plugs is visible above.  I slotted the bent tube with the Atlas milling machine.
The finished gear guard  The slotting went uneventfully, as the tube's sides did not try to clamp the sides of the 3/8 inch wide milling cutter.  To center the slot, I milled all the way around the guard and then reversed the guard in the vise.

   After that, it was just a matter of judiciously sawing tabs into the ends of the guard to fit the shape of the housing of the gear drive. Lots of filing and even some work with a jewelers saw to avoid struggles with clamping the guard in the bench vise.

   Lastly, I drilled the four attachment holes in the guard, clamped the guard onto the housing, and prick punched the sites to drill and tap the housing for No.10-32 tpi cap screws.  I opened up the clearance holes to 17/64 and did have to file one attachment hole oval to clear its screw.  It's very rigid and guards the delicate worm gear nicely.
Finished crank handle  


  The little cross pin that I used to lift the index plunger is too small in diameter for my fingers, so I made a tubular lifter and cut tabs from the tube as shown at left.  Now it's easy to lift and twist the plunger in order to latch it into place while cranking the worm gear to the next index position with no risk of catching the sector and moving it.  
 

   I also added a clamp to one of the sector arms as seen at the rear of the right- hand arm in the image at left.
The finished indexer with guard in place   The indexer is adjusted for zero backlash when the housing of the worm gear is clamped to the table of the milling machine. Then the end play is taken out of the worm gear by loosening the crank's clamp screw and screwing the crank down onto the upper thrust bearing.  Tightening the clamp screw of the crank takes up a little more end play, so this is a trial-and-error process to eliminate all the backlash in the indexing mechanism.  
 

   The gear guard not only protects the worm gear from inadvertent impacts, it also shields the gear from chips, as they cannot land in the teeth of the worm gear.  I do not have to remove the guard while mounting/demounting the accessory.
Part IV - Making indexing plates
Right angle adapter plate
   In order to drill the properly spaced holes in any new index plates that I might make, I needed an adapter to present the disk at right angles to the Atlas MM's spindle.  I found a two inch wide piece of half inch thick CRS which I draw filed flat to within 0.001 inch. 
 

   Then I drilled and countersunk two holes for 3/6-16 tpi cap screws to clamp this plate to the table, installed two 5/16-18 tpi studs and cut keyways in the plate for the indexing center's two 3/8 inch wide keys, and then fitted the worm gear accessory described above.
Dedicated adapter for index plates
   The Craftsman four inch chuck, while nicely made and fairly rigid, doesn't hold the disk securely for drilling, so I made a dedicated adapter threaded 1-10 tpi to fit the spindle nose of the indexing centers by chasing the threads in my South Bend lathe and then finishing the face and shoulder while the adapter was screwed onto another 1-1/2-8 tpi to 1-10 tpi adapter that I made years ago.
   The worm-gear accessory is just one way of indexing the holes in the planned disks.  I can just use another gear with a 3/4 inch bore right on the back end of the indexing center's spindle, or a different size bore (such as the South Bend lathe's 9/16 inch bore change gears) mounted on a suitable adapter made for the 80-tooth worm gear accessory.  In the image immediately above right I have mounted the angle-iron adapter made for that latter purpose onto the half-inch-by-two-inch CRS plate.  For example, I can use the 127-tooth, 48 DP gear for the Derbyshire instrument lathe as the template for drilling the 127-hole circle in a blank indexing plate.  The holes in the 5.0 inch diameter indexing disk would have to be smaller than 4.75Pi/127 = 0.115 inch in diameter in order not to part the disk [simple solution added in November 2016: don't drill clear through the disk !].
MachiningTheDiskEdge
Radially Lined Template
   Here's my proposed solution: Use a plotting program (here: PSI-Plot, running under Wine on a laptop running linux) to lay out 127 dots in a circle slightly smaller than the OD of the cast iron disk, with radial lines passing through each dot. The template is attached to the disk with Scotch double-sided tape and aligned within one-dot-width to the circumference of the disk; the double-sided tape doesn't "grab" the paper, so it can be pushed around with the dividers that I used to check the aligment and then pressed into place, where the sector arms and their clamps don't interfere with the paper.
Note: Right-click and select "view image" to see these pictures full size, whereupon the radial lines will look OK.
Setup for drilling the 127 holes
Back side of disk
Centering the dots
With the setup shown above, left, the sighting tube that I had just attached to the crank arm was about chin-high, so I had to use an angled dental mirror and (occasionally) an LED flashlight to see the dots as shown approximately in the photograph at right. This was easier than it looks, and I completed drilling the 127 holes in about two hours.
127th hole is above the marked 1st hole
127 holes on a 4.64 inch circle
Plotted 127 holes overlain on finished disk
   In the third image I have re-plotted the 127 hole locations on a 4.64 inch circle, converted the white background to transparency, and fitted the plotted circle's small white dots to the image of the disk. I had to adjust the width and height of the disk image to fit the calculated circle of holes because the Epson plotter may not be as accurate as my PSI-Plot plotting program.

   There are actually 127 holes in the disk !

My next step will be to countersink these 1/16th inch holes to set the sector arms with my original indexing setup, and then to drill another 127 holes in the second disk. Each pair of the 127 holes lies on 0.115 inch centers. View the overlain disk full size to view the hole alignments.