No announcement yet.

Threads again

  • Filter
  • Time
  • Show
Clear All
new posts

  • Threads again

    Gentlemen. Just this last hour I've cut a thread at 40 tpi in order to test the accuracy of the lathe in this mode. I cleaned everything up, lubed the gear bosses, etc- picked the bits and pieces out of the teeth, et al. Thinking was to give myself the best chance at nothing getting in the way of machining a 'perfect' thread.

    So I made the test piece and cleaned it up. Looks really nice- that printer rod machines beautifully. Then I set it up with the mike jig and tested it. I can't find any drunkeness, and if there's any error over length it seems to be within the flexibility of my jig. In other words, less than a half thou, so I'm calling it a half thou out over an inch. Because it's that close, I plan to re-fixture it to take some of the springiness out of the jig and test again. However, I'm happy with that result, and it looks like I'll be able to make myself a mike head with a longer range. So that's fine.

    Now I'm wondering about a couple of things- first off regarding the accuracy of the leadscrew. This isn't a toolroom lathe by any means, so I don't expect that an extra expense was incurred by the manufacturer to equip it with a high accuracy part. I wonder about the 'good' usa made lathes like the south bend, etc, and any other sort of equivalents in other countries (leaving out the current crop of cheap offshore machines)- what tolerance grades might exist for leadscrews on those machines? I'm just wondering if it was or is common for those manufacturers to set a high standard for such a part.

    I'm pretty sure I know the answer when it comes to todays offshore stuff-

    The other thing I'm wondering about is -who ever uses the high tpi range of thread cutting? I mean , on average, not the guy who makes his own no-see-um bolts for the dime sized 24 cylinder engine-

    I play with small things a fair amount, and I've yet to need to cut a thread past 20 tpi. What have you guys done that needed finer threads than that?
    I seldom do anything within the scope of logical reason and calculated cost/benefit, etc- I'm following my passion-

  • #2
    Originally posted by darryl
    I play with small things a fair amount, and I've yet to need to cut a thread past 20 tpi. What have you guys done that needed finer threads than that?
    When I made my chuck spider, I cut the threads on the adjustment for the spinner at M12 x 40 tpi
    I probably didn't need them that fine, but I thought it would give me plenty of fine adjustment.



    • #3
      Rifles and pistols commonly use 24 28 and 36 TPI, just for muzzle threads.

      Scope and sight screws run 40 48 TPI.

      Oddballs are the norm with guns...


      • #4
        With the average machine of any size and some use, anything much past maybe 48 or 56 tpi is probably a joke.

        The slop in the leadscrew/halfnuts/bearings system probably means the thing CANNOT repeat well enough to land in the thread no matter WHAT you do with perfect handling of the halfnuts and thread dial.

        I cut up to 48 tpi, and it works OK

        With finer threads, any fixed error is a larger percentage of the pitch, so "camming" of the leadscrew, pitch errors etc are much more significant at finer threads. By 100 tpi, you are probably fooling yourself if you think you can cut it

        Keep eye on ball.
        Hashim Khan


        • #5
          What have I made with finer threads? Well, almost everything I have threaded has been 20 TPI or finer. I guess I just make smaller things than some of the folks here. Camera lenses, micrometer screws (like you are attempting), various fixtures, etc. All seem to want finer threads.

          I actually chuckle to myself at comments I have seen about anything finer than 16 or 20 TPI being "special" or "too small" or whatever. For me they are routine.

          But to each his own.

          As for the accuracy of your lead screw, it may depend. First, the overall, long distance accuracy may be off by several thousanths per foot. Temperature can play a part in this, but it is a differential between the expansion of the lead screw and the work, not just one of them alone. But if your lathe is older, wear can produce short term differences. This will be specially noticable in the area near the headstock where most operations take place.

          You could get a handle on this by rigging a test setup with a large digital caliper mounted at the tailstock end and bearing on a good flat mounted on the carriage. Add a means of manually turning the lead screw exactly one turn at a time and note the readings on the caliper. They should show any errors - within the accuracy of the caliper, of course.

          To do it better you would need an even better set up with Jo blocks. A standard set of blocks would probably get you up to about 8 inches. For longer distances you would need more 4" or longer blocks. Again temperature would play a role and you would need to allow time for it to stabilize every time you handled the blocks.

          I doubt that many people have a real need for this kind of accuracy assurance. But if you are trying to make micrometer screws, you may be one of them. The difficulties you are seeing here are probably why most micrometers are made with only one inch or 25mm of threads. That and the pure pain of cranking a long, fine thread through so many turns. A five or ten inch micrometer screw would almost have to be motorized.
          Paul A.
          SE Texas

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


          • #6
            In the 70's I made a 120 thread tap with my Craftsman 12" lathe just fine. Still have it. It has 2 flutes.


            • #7
              Leadscrew accuracy has always been the mark of a good lathe.
              Southbend's were no different. They were good quality and precision ground, not cut!
              Slop and loose nuts have little effect on the results as all play is taken up during the thread feed period.
              But, as J Tiers pointed out, "Camming" is a definite detractor of precision, as it causes surging, and overfeed.
              Camming comes from several sources, such as damaged change gears,or broken teeth,
              bent leadscrews and even roughened way surfaces .
              The next detractor is worn leadscrews, which means, as a lathe gets older, the accuracy falls off as a natural function.
              Cutting fine threads ,like 60 TPI is really not a problem.
              You just have to use very sharp tools and a magnifying glass

              Here is a photo of a 32 TPI square thread cut on my 10 inch Boxford ( British Southbend)

              Major Diam is .172 and thread creast is .016"
              Last edited by Rich Carlstedt; 01-29-2010, 11:44 PM.
              Green Bay, WI


              • #8
                The SB9 goes up to 640 tpi. I tried it and it seemed to work ok but I don't have enough magnification to verify the thread form so I had to settle for 160 tpi.

                These threads were cut in 4 passes just to see if the lead screw following was accurate enough. They could have been cut in a single pass but that was no challenge. The material is hardened stainless steel and the threading tool is a triangle positive rake tungalloy carbide insert with the tip ground to almost zero radius. The threads were cut at 850 rpm.

                Free software for calculating bolt circles and similar: Click Here


                • #9
                  Evan, you keep that up and the SPAM will be after you.
                  Socity of prevention of abuse to microns.
                  Play Brutal Nature, Black Moons free to play highly realistic voxel sandbox game.


                  • #10
                    I did a couple of 1" x 50 TPI internal and external. I suppose that would be a Special.
                    The shortest distance between two points is a circle of infinite diameter.

                    Bluewater Model Engineering Society at

                    Southwestern Ontario. Canada


                    • #11
                      Interesting comments and pictures.

                      Looks like I just did 154per inch, calculated. I used hss ground to a sharp point, no radius at all, diamond honed. The thread doesn't look too bad, but of course I cheated by doing only a single pass. Maybe later I'll try multiple passes. I just used the feed lever instead of the threading lever.

                      I expected it to look bad, but it doesn't. I'll have to look at it under the microscope. Just for fun. I don't get out much

                      Actually, I have been getting out a bit lately. Guy I work with heads a band, so I've been out drinking and dancing with the girls. Fun but expensive.

                      Oh yeah, the threads. Evan, how does your lathe do 640 tpi- you must be fiddling with the gears

                      What I'd like to try is turning 40 tpi over a distance of about 3 1/2 inches on a rod that's about .3 or so diameter. I don't have to thread right to one end, since I plan to use half nuts for this project. I'm thinking to make up a holding jig for the tailstock, bored for a close fit of the donor rod. I'll insert say an inch or so of the rod, with lube, and the jig will keep the rod centered on axis and help prevent bowing a little better than a center would. There's going to be about 5 inches of rod between the chuck and the tailstock bushing jig, and I am going to get some bowing anyway, I'm sure. I suppose I could use a travelling steady- duh- I just this moment realized I actually have one.

                      I'll turn two zones, a lead-in and a lead-out zone for the threaded section, probably about 30 thou less diameter than the rod. I've got some Anchorlube G-771 metalworking compound that seems to work well, I'll use that.
                      Last edited by darryl; 01-30-2010, 02:36 AM.
                      I seldom do anything within the scope of logical reason and calculated cost/benefit, etc- I'm following my passion-


                      • #12
                        I built a PM Research boiler last year and then some 1/16th pipe size ftgs and valves. When it gets down to 2-56tpi for the stem nuts I used a tap and singlepointed the stem. Trying to cut 2-56 internal seemed way beyond reason to me. (10 in. Allas) I can't imagine grinding an internal bit that size.

                        I fired it up afew days ago and ran the engine.

                        Now I need to a make small Propane burner rather than the torch I used.


                        • #13
                          Oh yeah, the threads. Evan, how does your lathe do 640 tpi- you must be fiddling with the gears
                          No fiddling. It is listed on the brass plate for all the threading setups on the end cover over the gears.
                          Free software for calculating bolt circles and similar: Click Here