No announcement yet.

finding 'test bar'

  • Filter
  • Time
  • Show
Clear All
new posts

  • finding 'test bar'

    I'm looking for something that'll indicate spindle runout in my Rockwell 11 lathe.

    I've seen references to a lathe 'test bar', and even though I think that's something more elaborate than what I'm looking for, I searched for it on this forum.

    I searched every which way I could think of and always got 500 results, all of which included 'test' but never 'test bar'.

    What I need is something straight and round so I can figure out what's wrong with my lathe. Some time ago, I bought 3/4 x 6" dowel pin from McMaster Carr, but it was crooked. Might have been round, but wasn't straight.

    I tried using an Enco Jacobs chuck arbor held in a 5c collet and am seeing over a thou of runout. After finding out everything I could about installing spindle bearings properly I ended up with even more. Before I scrap the lathe, I thought I'd try to find something known round and straight.

    Frustrating all around.

  • #2
    A true test bar is very expensive, sometimes in the $500 range. Most of the old lathes are just plain worn out, and not a whole lot you can do. If you need something pretty close the only realistic solution would be use a 4 jaw chuck and dial in the workpiece before cutting. I would not scrap the machine, unless you buy a brand new import or find a Hardinge or Monarch it will be the same old story.


    • #3
      I've always thought that Jose Rodriguez's description of a lathe test bar was a very well written article.
      Rather than ad lib I'll direct you to a link with Jose's article.
      Home, down in the valley behind the Red Angus
      Bad Decisions Make Good Stories​

      Location: British Columbia


      • #4
        You can do a lot of tests without a test bar. What's wrong with your lathe?
        I seldom do anything within the scope of logical reason and calculated cost/benefit, etc- I'm following my passion-


        • #5
          I don't know what level of precision you are needing or looking for and most lathes leave the factory with less than that much runout but if you are getting that kind of results with unknown test components I would think your lathe is probably at least as good as an awful lot of lathes.
          Don Young


          • #6
            Brownell's has 2 different length test bars


            I hope this is what your looking for


            PS...if you get one, have it sent USPS, first two tries for me through UPS
            were failures.
            First one I got shipping tube, but no bar.
            Second one looked like it may been stepped on and was rejected
            Third VIA USPS finally was good
            Senior Member
            Last edited by alcova; 01-30-2011, 10:44 PM.


            • #7
              I once read on a machining forum that old style dot matrix computer printers contain a steel bar that is quite straight and uniform in diameter.


              • #8
                There are very accurate pins available for dies that are not too expensive and will serve as a test bar. I was given one by a die maker friend of mine a while back. It is 1" in diameter and about 10 " long, ground with centers. It is tapered about .0005" over it's length but knowing the taper it can be calculated when measurements are made. Just a thought.


                • #9
                  I stop out back of the thrift store and get the printers that are headed for the trash,all kinds of bar stock in them. And the stuff is straight.
                  Plus all sorts of gadgets in there that I find usefull/hordeable.


                  • #10
                    Warco sell them quite cheap. Good enough for most HSM'ers I expect:

                    Peter - novice home machinist, modern motorcycle enthusiast.

                    Denford Viceroy 280 Synchro (11 x 24)
                    Herbert 0V adapted to R8 by 'Sir John'.
                    Monarch 10EE 1942


                    • #11
                      The spindle runout is determined by the space between the spindle bearings and the spindle. I'm not sure why you need a perfectly round "test bar" to measure the runout. The procedure, at least for my South Bend lathe, is to use a dial indicactor and a bar that will fit the spindle bore to lift the spindle.

                      Using a chuck arbor held in a 5c collet and (I assume) rotating the spindle is doing nothing more than measuring the combined errors of the arbor, collet, and collet taper. Depending on the condition of the collet, etc., if you loosen the collet and rotate it 90* in the taper you may get a different "runout" reading. If you are rotating the spindle by hand, it is sitting in the bottom of the bearings and you are not measuring the spindle runout.


                      P.S. - Jose Rodriguez's description of creating a test bar referenced in a previous post is used to check the taper of your lathe, not for checking spindle runout.
                      Senior Member
                      Last edited by flathead4; 01-30-2011, 11:57 PM.
                      Tom - Spotsylvania, VA


                      • #12
                        well then in that case why bother with the inside of the spindle as the chuck is mounted to the outside of the spindle. the only time the inside would come into play is with a center and as for a center i prefer to touch it up every time i use it to ensure the center is right on the money, even with an adapter either for collet or center there is bound to be some run out even if it is in the millionths of an inch. run out is run out no mater what the amount.


                        • #13
                          Miller Machine makes test bars for several different lathes.
                          They have a web site with prices. They also make and have some lathe parts and gears.



                          • #14
                            There seems to be some misunderstanding what a test bar is for. A test bar is for checking bearing runout and also alignment of the headstock to the bed. You can use them to determine if you have bad bearings. Also they can be used when scraping in the head to the ways.

                            If you have bad bearings it does not matter what kind of chuck you put on you cant fix it. Thats why test bars are ground with a taper that fits the internal taper of the spindle, to get a direct reading of the spindle. Bad ball bearings can have non-periodic error and the south bend method is no good for modern machines with ball bearings. If the error is periodic it may mean you taper is damaged or your spindle is bent. How can you bend a spindle? A lot of them come from people getting the bright idea to lift the lathe with a bar through the spindle.

                            One does not use a test bar for a 3 jaw, a 3 jaw is not accurate and not intended to be. Bad ball bearings can have non-periodic error and the south bend method is no good for modern machines with ball bearings.


                            • #15
                              +1 for miller machine. Not $500 for a 12 inch MT2.
                              There is also a technique that will use anything that is sort of straight and similar diameter.
                              Here is a link
                              I did find the test bar easier to use. After leveling and changing out the bearings on my myford clone aligned the headstock and got rid of almost all of my taper.