Announcement

Collapse
No announcement yet.

OK, here's the deal- set true chuck

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

  • OK, here's the deal- set true chuck

    About the Pratt Bernerd set true chuck, I talked about in a previous post. After disassembly, clean up, finding a new set of dog point set screws, and reassembly. I was able to get it rotating with less .0005" run out at the chuck jaws, However, when checking at 3 inches from the jaws I find out I have .0035" run out. How do I take this out? I want to thank everyone who were so helpful on my last post.

    Getting this old lathe up and running is almost to much for this old man!

  • #2
    There is a tilt to "something". The jaws are not holding the part in-line with the spindle axis.

    That may or may not be the fault of the jaws.

    Step 1..... put an indicator on the face of the chuck, and see if the face has any runout, which would be caused by tilt.
    You will need to deal with the jaw slots. One way is to adjust so you have only a few thou deflection, so that the ball at the indicator tip can easily slide up onto the surface again.

    If the face is not running out, then the chuck is almost certainly mounted OK, and the next step is jaws.

    If the face IS running out, there is some sort of mounting problem, either in the backplate, or where the backplate touches the spindle. A mis-machining, a piece of swarf, a ding, dust, whatever.

    Step 2 .......If it really is the jaws, then grinding them while they are held in the position they take when clamping a part, is likely to fix the issue. Double check everything first before you "cut metal" on the jaws, though, the problem might be a burr on a jaw, or some other easy to fix problem.
    Last edited by J Tiers; 10-02-2020, 06:46 PM.
    1601 2137 5683 1002 1437

    Keep eye on ball.
    Hashim Khan

    If you look closely at a digital signal, you find out it is really analog......

    Comment


    • #3
      What system do you use for mounting the chuck to the spindle?

      Comment


      • #4
        So assuming that you have a piece of rod chucked up and dialed into .0005 at the chuck. Now you have .0035 at the far end of the rod. Absolutely first thing to confirm; Is the rod straight and round? If so then make a mark on the measured high spot of the rod, and the corresponding place on the chuck while chucked up. Now loosen the the chuck jaw and rotate the rod 180*. Tighten and recheck. Does the high spot follow the chuck or the rod? You can also loosen and tighten the chuck while putting some bias force on the end. This will show bell mouth if you have it. You can blue up the rod and check the pattern of the jaw contact on the rod. Grip the rod very lightly and push/pull sideways at the end of the rod. Is it moving about.

        Are the jaws in the chuck tight or are they moving about in the key way slot indicating a well worn chuck? Do you see problems with different sized rods?

        All things to check and use as data points for deciding what the failure mode is. It also may be none of the above.

        lg
        no neat sig line
        near Salem OR

        Comment


        • #5
          Do you want the home shop method of controlling this or the real world machine shop version?
          You most likely will not like either method very much.

          Comment


          • #6
            Absolutely what Larry said. Qualify your test rod first
            I cut it off twice; it's still too short
            Oregon, USA

            Comment


            • #7
              Originally posted by rogee07 View Post
              About the Pratt Bernerd set true chuck, I talked about in a previous post. After disassembly, clean up, finding a new set of dog point set screws, and reassembly. I was able to get it rotating with less .0005" run out at the chuck jaws, However, when checking at 3 inches from the jaws I find out I have .0035" run out. How do I take this out? I want to thank everyone who were so helpful on my last post.

              Getting this old lathe up and running is almost to much for this old man!
              I have some of those. I buy them used. Each time I get one I remove the mounting plate from the chuck and turn the plate square to my spindle. Sometimes not much more than a shave really.. Thats my process. JR
              My old yahoo group. Bridgeport Mill Group

              https://groups.yahoo.com/neo/groups/...port_mill/info

              Comment


              • #8
                I am curious, just what are they? And how do they differ.

                It seems to me that you start at the spindle and work your way out one part at a time.

                1. Is the spindle running true?

                2. Is the back plate running true?

                Etc.

                If this is a used chuck, it is probably wear on the jaws. but the only way to be sure is step by step.



                Originally posted by Bented View Post
                Do you want the home shop method of controlling this or the real world machine shop version?
                You most likely will not like either method very much.
                Paul A.
                SE Texas

                Make it fit.
                You can't win and there IS a penalty for trying!

                Comment


                • #9
                  simple solution for "skewed" chuck: chuck up a piece of round (preferably big and ground) stock, stick in into a lathe and skim the back side.

                  Comment


                  • #10
                    Originally posted by Tim Clarke View Post
                    Absolutely what Larry said. Qualify your test rod first
                    "Assumption" is that if the rod looks like a corkscrew, that just may have been noticed or considered already........ but the chuck and mounting can be eliminated by checking the face runout. I have yet to find a chuck that has face runout inherent to it. Manufacturers do not want the chuck to look like it is way out of square.
                    1601 2137 5683 1002 1437

                    Keep eye on ball.
                    Hashim Khan

                    If you look closely at a digital signal, you find out it is really analog......

                    Comment


                    • #11
                      After making sure the rod is ok, then check the front face of the backplate for any runout and skim if necessary. When the chuck is fitted without the jaws, the front face can be checked for runout, assuming it has not been damaged by misuse. If any runout of the backplate was rectified, then retest. If the error persists then the jaws could benifit from a little grinding. The condition of the chuck will undoubtedly be a factor in its precision or lack of.

                      Comment


                      • #12
                        Originally posted by Paul Alciatore View Post
                        I am curious, just what are they? And how do they differ.
                        There are several approaches, one may solve the root problem or treat the symptoms.

                        Hobbyists are often very concerned with accuracy and repeatability which are two different things. Making an accurate part (many make just 1 or a small number of parts) is easily done if time is not a consideration as most claim. Someone such as yourself are more concerned about the machine being “correct” then actually making the part accurately.

                        The hobbyist method is, as you mentioned, to first determine the root cause of any inaccuracy, this includes much pulling of hair.
                        Then devise home brewed methods of correcting it such as grinding the spindle nose in place, turning the back plate in place, grinding the hard jaws in place and so on, this may take weeks or indeed months. This costs little money yet a good deal of time.

                        In an operating machine shop if a single part needs to be made you simply treat the symptoms and make the part, repeatability is unimportant when making one part. If producing 100 parts per hour you do not have the luxury of painstaking setup or even measuring each part. In this case repeatability is paramount, if the process will not repeat then the machine issues must be addressed using the same methods as listed above. This may include buying a new machine or hiring a company that does nothing but correct spindle or chuck errors in place, this is often done in 2 days or less but is a bit costly for a hobbyist (-:
                        Last edited by Bented; 10-03-2020, 06:39 PM.

                        Comment


                        • #13
                          Originally posted by Bented View Post
                          .........
                          The hobbyist method is, as you mentioned, to first determine the root cause of any inaccuracy, this includes much pulling of hair.
                          Then devise home brewed methods of correcting it such as grinding the spindle nose in place, turning the back plate in place, grinding the hard jaws in place and so on, this may take weeks or indeed months. .........
                          If it took me more than an afternoon to get to the bottom of it, and maybe do the correction, I'd go take up knitting.........

                          Yes I have seen a number of businesses which are satisfied to do 100 different one-offs the long, hard way, and won't fix the machine. But they WILL if they think they can do a big run. Bogus thinking, of course. The big run may be no more pieces than a couple dozen small runs involves.

                          If they just don't want small runs, they are at liberty to decline them.... but the boss keeps accepting them.
                          1601 2137 5683 1002 1437

                          Keep eye on ball.
                          Hashim Khan

                          If you look closely at a digital signal, you find out it is really analog......

                          Comment


                          • #14
                            Originally posted by Bented View Post
                            Hobbyists are often very concerned with accuracy and repeatability:
                            I have not noticed that.

                            Accuracy for sure. Repeat? Not production.. JR

                            My old yahoo group. Bridgeport Mill Group

                            https://groups.yahoo.com/neo/groups/...port_mill/info

                            Comment


                            • #15
                              I'm always looking for the easy way, which often ends up taking longer than doing something the right way. So let me ask this question. Would you be able to increase the parallelism of a set true chuck by inserting shims between the backing plate and the chuck body?

                              Comment

                              Working...
                              X