Table Saw Tuneup

Basis: 1976 Craftsman 10" contractor saw; cast table, steel extensions, sheet steel cabinet and legs.

I've been able to get this saw to perform almost as well as a cabinet saw (though I know its trunnions will one day fail me.) Key improvements included:

  1. Addition of machined cast iron pulleys and PowerTwist link drive belt to eliminate drive vibration;
  2. Fabrication of wood zero-clearance throat inserts as needed;
  3. Center-punch-and-file sides of original miter bar to eliminate slop in miter slot;
  4. Extremely patient and frustrating alignment of blade parallel to table slot, to immeasurable deviation. As you may know, the act of tightening the trunnion bolts tends to skew the trunnion, and alignment is therefore a repeat-til-you-get-lucky proposition.
  5. Addition of various shop-built miter gauge and fence auxiliary fences (see home page link);
  6. Design and Replacement of incredibly poorly machined arbor (see below);
  7. Addition of case reinforcing bolts to (almost) eliminate surface movement and leg twisting (see below);
  8. Addition of cast iron extensions; (Sears has them if you're diligent in searching.)
  9. Addition of Incra 1000SE miter gauge;
  10. Regular (hard paste) waxing of all sliding surfaces, including top, miter bar, fence rails, etc.;
  11. Addition of protractor-based blade height micro-adjustment (see home page link);


  12. (Eventual) replacement of fence - still researching. I've actually learned how to maintain alignment for the original, and am not desperate for the upgrade, though it will undoubtedly speed up work.
  13. There's a replacement set "out there" for the trunnion bolts that includes a pair of opposed setscrews that negate trunnion movement when its bolts are tightened. Coming ASAP!

Replacement arbor
Because arbor diameter varied greatly over its length, the saw could not cut flat-bottom dados. There was no recourse but to replace the arbor. Two replacements ordered from Sears were as bad as the original, so I sketched the design below and had it fabbed at a local machine shop for $80.00

The old arbor was taken to the machine shop along with this sketch. Had to be sure to take the nut, too, as I was fearful that the tighter arbor specs would cause interference fit for threads. As it turned out, the old nut worked just fine, but barely (when the arbor is cold, it nearly spins on; when it's warm, the nut will still turn with mild finger effort, but will definitely not spin.) Another half thousandth and it would have required a little machine work. That's so close that I'd just recommend taking it along as a matter of course. Here's the sketch, which you're welcome to save and print : (The sketch is a small .gif suitable for download and printing.)

  1. Dimensioning looks a little weird, as I wanted to keep the image as physically small as possible. Look carefully, as many of the tolerances are in tenths, not thousandths. Call your machinist's attention to their importance.

  2. FYI machinist: the flange body is a light shrink fit on the shaft - the old one can be re-used if possible.

  3. If you follow this path, be sure to remove your throat plate, tilt the arbor to 450, and double-check the allowable overall length for the working end! Imagine that the nut is in place at the end, and measure to confirm that there will be clearance between outboard end and throat opening. The sketch shows that I measured an allowable projection of 1-5/8" from flange to end of arbor.

  4. If you go so far as to tear down to this level, then I'd suggest that it might be wise to replace the arbor bearings, too. Hint: there are several metric/English variations of that bearing with slightly differing actual dimensions. Also, there are U.S., Turkish, and heaven knows what else - go for the best you can get, and be sure that I.D, O.D., and thickness match the old ones.

  5. It's a good opportunity to reassemble with machined cast iron pulleys and a segmented Power Twist V-belt - minimal cost and really noticeable benefit!

  6. The very tight diameter dimension will make blade change a little slower - it takes a bit of patience to work the blade off without binding. Cheap blades may be undersized on hole diameter and require just a touch with a reamer to get 'em on at all. I've had no difficulties with blades from high-quality manufacturers - my sense of feel as a machine designer suggests that clearance is actually just about the planned half-thousandth at the arbor flange, and 2-3 thousandths along the threads.

Leg/Case Reinforcing Bolts

These two photos are taken at saw's feed end, left side.

The horizontal joint at the intersection between case and legs would open slightly at each corner, and the legs would twist when the top was side-loaded - obviously related to disconcerting table movement.

Looking up from floor into same corner.
The single bolt originally provided at each corner was so far inboard from the outside corner that the case frame (folded under in the horizontal plane) was able to bend at the corner, and the outside corner of the leg was free to wiggle and twist. Simple addition of a new 1/4" machine screw with nut and star lockwasher at each corner as shown tied all the sheet steel together and virtually eliminated table movement - the difference was truly remarkable. Note that at the location chosen, both horizontal legs of the case and both top folds of the leg overlap and all are captured by the bolt.
Suggestions for proceeding:
  • Be sure to wear wraparound eye protection if you drill and install from underneath, as I did.
  • So as to ensure maximum benefit even if the screw should work loose slightly, I used a drill gauge to determine the actual diameter of the screws, and chose a drill diameter just a few thousandths larger.