Announcement

Collapse
No announcement yet.

Set Up For Grindind Saw Arbor Blade Flange

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Set Up For Grindind Saw Arbor Blade Flange

    I put one of my new bearings on my table saw arbor the other day. after pressing the flange back on the shaft I chucked the shaft up in my lathe to check the run out of the flange. It's off by about .0015 thou. Doesn't sound like much, especially for a table saw arbor but I have to wonder what the error would be on the cutting edge of a 10" blade, or even a 7" blade. I know the error is greater the further from the center of the shaft you go. Basically the effect of this run out is like what you would have with a wobble dado blade only on a smaller scale.

    I'm tempted to set it up in my T&C grinder and face it but am afraid I may make it worse, or not cut the flange face perfectly square to the arbor.
    The only place I can chuck up on would be the center of the arbor which is 13/16" CF round and I can feel the unevenness in it as I turn it between my fingers.
    The threaded end is out of the question for this set up and so would be the pulley end of the shaft.
    For being square all I would have to rely on would be the accuracy of my grinders work head being true with the table travel.

    I was going to mount the flange on a shaft or expanding mandrel but it should really be ground in place.

    JL.................



    Last edited by JoeLee; 04-25-2019, 01:06 PM.

  • #2
    How about taking a light face cut in your lathe? Grab a short section of the arbor's tail in a 4-jaw and dial it in, or grab in a good collet setup, and grab the bearing seat or even the bearing itself in a steady. Push the flange against the bearing inner race or use a couple of purpose-made sleeves and the arbor nut (or disassemble the inner race from the old bearing), and then take a light face cut. You could hardly get any less runout.

    Comment


    • #3
      It's not likely that you will be able to improve on 1 1/2 tenths.

      The error at the saw tips is proportional, so if your run out was measured at 1 inch radius, the run out at the tips of a 10 inch dia. saw blade
      (5 inch radius) would be less than .001". They don't make wood butcher blades that accurate.

      If you want to be fussy, when you mount a blade on your saw, measure the run out, and then reposition the blade on the arbor in several orientations to find the position of least run out.

      Run out in the few thousandths of an inch never hurt any saw cut in wood that I have ever seen. It gives the blade a little breathing room.

      Comment


      • #4
        I meant to say 1 1/2 thou, not tenths. Typo on my part.
        I would never worry about a couple tenths.

        JL.....

        Comment


        • #5
          I like rklopp's advice, agreed this is a lathe job, grinding not necessary.

          You could evaluate the 13/16" section compared to the bearing journals on v-blocks in a surface plate to see if it's good enough to chuck on. Lacking that equipment, 4-jaw chuck or collet the bearing journal on the non-threaded end and evaluate the 13/16" section before you break out the steady rest. The shaft was likely turned between centers though and the 13/16" section is not ideal for holding.

          Comment


          • #6
            I had an old Walker Turner drill press that the chuck was running out. What I did is chuck up the shaft on one bearing diameter (bearing removed) and got that to perfect, On the other bearing point, I used my steady rest on that bearing surface. Then I recut the Jacob taper just enough to clean it up. Installed in the drill press and it was within .0005 which is plenty good enough for a drill press - it original was out about .015.

            Comment


            • #7
              I think you are talking about variation in the axial direction, along the axle right. That will as you say work like a wobble saw and if measured at 1 in radius become 5x as much at 5 in radius on a 10 in blade. But the other side of the blade along the diameter is also wobbling by that much so total variation is 10x total. So 10thou. I think that is less than the variation in set of the teeth.

              Comment


              • #8
                As above... and if you are really persnickety you'd need to check the "run-out" at full speed - the blade is flexible and maybe be quite different at 5000 rpm

                Comment


                • #9
                  Originally posted by Baz View Post
                  I think you are talking about variation in the axial direction, along the axle right. That will as you say work like a wobble saw and if measured at 1 in radius become 5x as much at 5 in radius on a 10 in blade. But the other side of the blade along the diameter is also wobbling by that much so total variation is 10x total. So 10thou. I think that is less than the variation in set of the teeth.
                  Yes, that is correct, that's exactly what I meant. How much is the error amplified 5" out from the center.

                  JL..............

                  Comment


                  • #10
                    Well, here is what I did.

                    I set it up in my T&C grinder. Aligned the work head of it's reference line. I put the high spot on the arbor between the jaws to eliminate any mounting error.
                    Took about .002 to clean up and called it done. I checked it with my tenths dial and have 0 dial deflection.




                    Then I put it back in the lathe, put a ground parallel across it and checked it for squareness to the shaft, I got 0 dial deflection.
                    So it all worked out good.




                    The only thing I don't like about grinding / facing on the T&C grinder is it's easy to concave or convex with out knowing until you check. That's why alignment of the work head is so critical. If not set up properly that surface I just ground could easily end up tapering inward or outward.

                    JL..............
                    Last edited by JoeLee; 04-25-2019, 04:29 PM.

                    Comment


                    • #11
                      It's simple scaling or trig. If it's out by .0015 total over the diameter of a 2" flange ( max high to max low as in indicated just at the outer edge of the flat face) then it'll be 10/2 times as bad at the edge of the blade. So 5 times as much, not ten times since we're looking at the full diameter of the flange.

                      Now that means .0075 total or thereabouts. And that's not bad at all. With a good blade that will give you a nice clean cut.

                      As it happens I've got a fairly nice rip blade on my saw at the moment that gives me surfaces straight off the saw that are smooth and suitable for gluing or need only a slight finish sanding or a very thin shaving or two from a hand plane. I just checked with a dial gauge I use for a variety of tests on the table saw and my total runout is .005 to .006. So it's right in the ball park with your saw.

                      Mind you if you can tune the runout up to half of what you have now and if the spindle does not have any float that causes play then I'd say you've got a prime candidate for very nice cuts with a good blade.
                      Chilliwack BC, Canada

                      Comment


                      • #12
                        I re-cut the face of the arbor on my Unisaw in place on the saw with a small tool slide and carbide tool bit. Took very fine cut and it worked fine. True as you are going to get it running in it's own bearings.
                        Last edited by toolznthings; 04-25-2019, 10:47 PM.
                        Toolznthings

                        Comment


                        • #13
                          Originally posted by TOOLZNTHINGS View Post
                          I re-cut the face of the arbor on my Unisaw in place on the saw with a small tool slide and carbide tool bit. Took very fine cut and it worked fine. True as you are going to get it running in it's own bearings.
                          I agree but it would be a major job to try and set this up running in it's own bearings.
                          The important thing here is that I have eliminated the run out and the surface is square to the arbor.

                          JL.............

                          Comment


                          • #14
                            Originally posted by JoeLee View Post
                            I agree but it would be a major job to try and set this up running in it's own bearings.
                            The important thing here is that I have eliminated the run out and the surface is square to the arbor.

                            JL.............
                            Glad that worked ! Note : I actually did it on the saw itself using the saws motor. Not a separate setup elsewhere.
                            Toolznthings

                            Comment


                            • #15
                              Originally posted by TOOLZNTHINGS View Post
                              Glad that worked ! Note : I actually did it on the saw itself using the saws motor. Not a separate setup elsewhere.
                              That seems like an almost impossible task especially given the room you have reaching down through the table saw's blade access hole. The set up would be too time consuming.
                              My arbor now shows 0 runout as compared to .0015 before. Doubtful that I'll pickup any error when it's mounted in the casting. I'll check it anyway. The accuracy should remain even when mounted. The only place error may be picked up is in the bearings. We'll see.

                              JL...............

                              Comment

                              Working...
                              X