No announcement yet.

Ball turners, and how to use them

  • Filter
  • Time
  • Show
Clear All
new posts

  • Ball turners, and how to use them

    I've spent the last several weeks making a ball turning attachment. Well, that time includes making various other tools that I needed to make the ball turner, as you might expect.

    Like I needed an unobtainable Acme (trapezoidal) thread, so I cut a tap and...

    then I needed to cut the relief on the tap, but I had no tailstock for my rotary table...

    and so on...

    I've noticed several people have made their own. Maybe if I start this thred with a run down of how my ballturner came into being, others could share their comments about what's good, what's bad, and tips and tricks of ball turning in general.

    I wanted a good beefy ball turning tool to go on my spindly plain Lorch lathe. For those that don't know, a plain lathe has no lead screw (or rack). You move the cross slide along the bed by hand, lock it up, and do everything with the top slide. Resetting the top slide to be parallel to the bed after using it at an angle is a prime PITA.

    I've converted my lathe to threading by sending a feed via UJs to the left end of the top slide screw, like Lorch did with some models. It works well, but it means I'm limited to threads about 5 inches long IIRC.

    I have only 64mm centre height above my cross slide, which with my normal fat top slide gives me a paltry 11mm of tool height, and I can't use the cross slide as an anchor for rotation as there's no space below decks before the feedscrew gets in the way. There's a central shallow hole for centring the usual top slide, and I use that to centre my ball turner. The bottom of the pivot is a close fit in that hole, so the two bolts to attachthe mounting plate to the crossslide are not for positioning.

    I wanted to match my everyday tool clearance of 11mm with the same for the ball turner tool post. That gave me 53mm into which I had to fit a full swivelling dovetailed slide with a tool post mounted on top.

    I wanted to turn big balls, which meant making the slide and swivel as low as possible, and providing considerable advancement for the tool.

    I ended up with 5mm of swivel plate, 22mm of slide, a 5mm adaptor plate and a tool post of 20mm. Yes, that made 52mm which I'm still pissed off about (64mm minus 52mm being 12mm). It means I cut the tool opening in the tool post 1mm too low. I've got to add a 1mm shim, or shim up every one of my normal tools I want to try out in the ball turner. Grrrr #1 !

    The slide is 75mm wide, 22mm high, and has 45mm of travel.

    This should allow 64mm minus 22mm minus 5mm of material, say 35mm of material radius for safety, above the slide, which incidentally allows me to cheat with my main objective, which is to turn a boule for petanque.
    Offically, boules are 70.5mm upwards in diameter, although the one I really wanted to copy is 73mm. Grrrr #2 !

    If I made another set of Acme/Trap taps I'd make the taper of the first tap much less, and maybe make three taps. I thought I was going to snap the first tap it was so hard to turn. I'd also cut the relief a lot deeper. What possessed me to cut relief only to the depth of the minor diameter of the thread, I don't know.

    What I couldn't find a way to squeeze in was a cover for the feed screw. You copy what you know, and the GA for my slide is pretty much a half-scale version of the top slide of my Lorch. (I have no experience of Lorch and Boley top slides for watchmakers' lathes). Lorch make the circular base (top) of the nut (which I made out of brass, not bronze - tut tut) wider than the channel, so a thin steel 'roof' covers the channel, looping itself over the nut, between the nut and the roof of the slide. Well, that level of detailed design was beyond me first time round - maybe next time.

    However all may not be lost, for as you can see in the bottom photo of the ball turner in action, with a pile of chips in front not one chip has sneeked round the back into the channel.

    Maybe turning balls becomes a matter of routine once you've had some practice. It certainly needs some thinking about when you start out.

    When I started to use the tool I was quite confused about what was happening until I realised a number of things (when I refer to o'clock below I mean from the user's point of view - the tailstock is three o'clock):

    1) Setting the cross slide so the ballturner's pivot is directly under the lathe axis is crucial, and I must make measurements so I can reproduce this at will. I suppose I need a cross slide stop mechanism ?

    2) Where the tool tip is in relation to the pivot does not really matter - the result will still be a ball. But the further to the left the tool tip is, the fuller a ball you can turn. To put this another way, if the tool is pointing down the centre line of the tool post and the slide at the twelve o'clock position, there'll be more things getting in the way of each then other when the tool tip is offset to the left. A little here helps a lot. Moving the tool tip to the left 5mm makes the difference between getting to seven o'clock and getting to eight o'clock. And that's the difference between a round lump on the end of a rod, and a sphere that looks as if it's about to part itself off.

    3) The tool tip needs some thinking about. Offset about 15 degrees to the left I got a reasonable ball using a tool shaped like an acme thread cutter, ie a narrowing parting tool, but it wasn't much good between three and four o'clock. If I point the tool to the left, then when I'm cutting near the lathe centre line, at three o'clock, the tool is pointing back at me, so it needs relief on its right flank. Once it's round at seven o'clock, the tool is cutting like a parting tool, so all it needs is a more normal front rake.

    4) All bets about tool positioning will be off as soon as I try to turn a ball halfway down a rod.

    I'm sure there's lots more to learn, and I've only been using collet mounted work. When the chuck comes into play there'll be a lot more pressure on lathe bed real estate. I may have to design another tool post that
    'peers round the corner' as it were. Maybe I'll get away with a wide tool with the cutting point sticking out on the right like a reversed boring bar.
    Last edited by rohart; 06-27-2010, 09:11 AM. Reason: Getting pictures working
    Richard - SW London, UK, EU.

  • #2
    Impressive build.

    You seem to be concerned about chips getting on the screw. Why not mill a 1.5875mm wide lip on the dovetails so a piece of sheet steel can lay over the screw? Secure it with a weak loc-tite.
    Last edited by Tony Ennis; 06-27-2010, 09:44 AM.


    • #3
      Tony - don't forget this is a lefthand feedscrew, with the nut attached to the underside of the sliding part. The nut is near the front of the slider so there's nowhere for the cover to go as the slider retracts.

      Lorch handle it by having a gap above the nut that's wider than the channel, so the nut itself is wider than the channel and the attachment screws together. Then the cover is the full length of the lower base of the slide, and is looped over the nut.

      It's just a matter of designing it right in the first place, but as it was my first slide, and I was squeezing a whole lot in to not much vertical space, without really understanding how my existing compromises were going to affect the main task - ball turning - squeezing this bit of detail in was beyond me at the time.
      Richard - SW London, UK, EU.


      • #4
        Originally posted by rohart
        ...Setting the cross slide so the ballturner's pivot is directly under the lathe axis is crucial, and I must make measurements so I can reproduce this at will. I suppose I need a cross slide stop mechanism?
        Have a look at for an idea. The W20 arbor with 50% milled away goes in the spindle and you wind the cross slide in or out so the flat bit on the attachment touches it.

        Could also be done with a vertical pin concentric with the radius turner's axis.

        I think the same sort of thing is done with the 'setting pins' on a Quorn.

        Could you use a piece of EPDM screwed to the slide to cover the gap? It would flop over the dial, however when fully retracted.


        • #5



          • #6
            Ball turners, QCTPs and tool sharpening jigs all command the "Better mousetrap" archives, the number of designs are prolific. If you have access to M/E 3rd May 1991 page 519 Richard, it shows a centre pin setting as DJC has suggested. If not, and you want a copy, PM me your E-mail address.

            Regards Ian.
            You might not like what I say,but that doesn't mean I'm wrong.


            • #7

              I like the idea of the flat intersecting the centre line, so a half cutaway arbour can be used to set it up. But that's an attachment designed to be installed on whichever actual lathe the user has. If the attachment is dedicated to one lathe and one cross slide, then a cross slide stop should be an easier solution.

              I can't see how the Schaublin attachment gets its toll anywhere near the ball. The tool post would seem to want to hold the tool so it's tangential to the ball. Maybe I'm just not getting it. I must look up Tony G's site to see if there was a Lorch add on.
              Richard - SW London, UK, EU.


              • #8
                Very impressive. You did yourself proud.


                • #9
                  Nice Job Indeed

                  Small strip s/s wit hholes drilled throough to hold a precut piece of rubber (aka bridgeport way cover) May work out


                  • #10
                    Looks great. Certainly appears to be beefy enough, and well designed.
                    I REALLY like it!

                    But one thing's unclear to me.
                    Ok, you have the round plate, which is bolted firmly to the cross slide, via the two countersunk screw holes shown. Right?
                    Then there's the round pivot, which is attached to the round plate, and around which the bottom half of the slide mechanism rotates. Right?
                    What then, holds that bottom half of the slide down onto the round plate and pivot?
                    I can see the round piece with two holes that appear to be there to engage pins of a spanner. Is it screwed directly down onto the pivot? Or is the pivot provided somehow with a round (undercut) dovetail that engages somehow the rotating bottom slide?

                    The reason that question is especially burning in my mind, is that a few years back someone (forgotten the name now) had about a 5 part article in HSM magazine describing the build of a similar ball turner I really liked, designed for a South Bend. But I could never come up with or visualize a suitable modification to make it fit my lathe. Perhaps your method would.
                    Lynn (Huntsville, AL)


                    • #11
                      Thanks for the great post. Looks like a really neat project. Simple as they are, I liked some of your construction technique as they are ones I've never thought of using.
                      - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
                      Thank you to our families of soldiers, many of whom have given so much more then the rest of us for the Freedom we enjoy.

                      It is true, there is nothing free about freedom, don't be so quick to give it away.


                      • #12

                        When I undo two screws and take the rotating base of the top slide off the cross slide of my Lorch, I am looking at a 40mm hole in the cross slide. The hole helps to locate the rotating top slide, which has a raised boss that fits (not very well) into the hole.

                        I can't use the hole for anything, as the cross slide feed screw is too close to the surface.

                        But I use the hole as somewhere to put the head of the pivot bolt for the ball turner.

                        In the top photo on this thread, on the far left is a wide headed boltish piece. This fits through the hole in the round plate, and two screws through the round plate into the head of the bolt prevent the bolt from turning. The bottom part of the dovetail slide fits closely over the shank of this bolt, and the circular peghole nut with washer holds the slide down. That assembly needed a fair degree of precision, but it worked out OK.

                        Does that clear it up ? Next time I do anything like this, I'll try harder for the 'exploded diagram' approach.
                        Richard - SW London, UK, EU.


                        • #13
                          djc and Circlip:

                          After several months (of not doing any ball turning) I've finally understood why the 50% centre wouldn't help much in setting my too to the centre line. On my tool there's no surface available to attach a marker to that indicates where the pivot is.

                          I suppose I could set up some kind of gantry-like assembly that would touch the 50% at the centre line, but that would seem a bit Heath Robinson/Rube Goldberg.

                          I've measured how far the front of the cross slide is out from it's base when I turn a sphere, and that's good enough for set up.

                          I can also measure the almost sphere when I'm almost there, both fore and aft (at 90 deg to the spindle axis) and at 45 degrees. It's quite easy to work out how far to move the cross slide in or out to get it spot on before the last few cuts.
                          Richard - SW London, UK, EU.


                          • #14
                            If you want to reduce the angle part toward the spindle you need to
                            offset the cutter to the left of the axis. See the following pictures.





                            • #15

                              Nice device!

                              To get more clearance, you could replace the entire toolpost with a vertical pillar, and fit a small carbide insert to the top of it with a single screw. Pick your favourite shape - round, triangular, diamond to suit.

                              All of the gear, no idea...