No announcement yet.

metric theading,,,,,methods...tried them?

  • Filter
  • Time
  • Show
Clear All
new posts

  • metric theading,,,,,methods...tried them?

    I have Atlas and a Clausing lathes. Have any of you done the gear purchase and adjustments necessary for either one to be able to cut metric? I've got the Atlas manual that tells which gear/s to purchase,,but the changing of the gear holder sounds like a real pain in the butt. Especially if you have to switch back to English for anything. I cut an occasional barrel thread to 1.25M for CZ rifles. No one has talked me into being 'ok' with 20 tpi which is very close,,but if it ain't right,,it ain't right,,,,,right?
    ps> been using a friend's huge Colchester [sp?]lathe for metric threading. It takes two of us at both ends of lathe to move levers [backing up cutter, etc]
    alan in ga

    [This message has been edited by Alaninga (edited 09-18-2004).]
    alan in ga

  • #2
    If you want a quick way to thread, check out Hardinge's method for threading..maybe you could build something like they use and stay away from gear changes..

    We had a 50 inch lathe (OD!) without a gear box, and dropped a chain on the spindle to a side shaft (1 to 1 ). the side shaft turned a threaded rod with the threads we wanted.
    The threaded shaft propelled the tool slide.. bingo . had to reverse the lathe and pull the tool out each time, but we were able to cut threads on a threadless lathe !


    • #3
      I cut a metric thread on a big monarch once by using a feed that worked out almost exactly the same as the metric thread. I engauged the cone clutch and left it engaged and used reverse to back it out. You could also use the feed lead reverse.


      • #4
        Rich Carlstedt I know what you are doing and many big lathes do it that way, no threading dial, you cannot miss the tread doing it that way.

        some have worms for threading and you cannot miss on them either.


        • #5
          I bought the 37T/47T change gears from Logan for my lathe. They provide 1% accuracy as opposed to 100% accuracy if the 100T/127T gears are used. The Logan gears I chose provide a ratio of 1.2702702, which when doubled gives 2.5405404. It worked out very well. The only change was that I had to reverse the lathe without disengaging the half nuts because the standard thread dial doens't do metric. Previously I tried to use the gear combinations recommended by Atlas to approximate metric threads. These are standard gears provided with their lathes, 44T/52T. Together they "combine to give a ratio of 44/52 or .846154, which is almost an exact function of 2.54, the English to Metric ratio" as stated in the Atlas Manual of Lathe Operation. They were supposed to produce threads with a small error (1 part in 3000), but it didn't work for me. Perhaps I was using them improperly. I do not claim to be a math wizard, but when I calculate the 44/52 ration I get 1.1818181 which when doubled gives 2.3636362. That doesn't seem to be as close to 2.54 as claimed by Atlas. I have no idea how they calculate that to be 1 part in 3000. I can tell you when tried, it didn't produce threads that were even close for me.


          • #6
            A VFD and 3 phase motor makes thread cutting a breeze (with gear changes as required, of course). I've got a Harrison M300 and after removing the thread counter for cleaning, have never installed it !

            With the VFD and forward / reverse lever or switch it becomes whirr --- whirr --- whirr

            No wear and tear on the lead screw or half nut. Just food for thought if you ever need to replace the motor/drive.



            • #7

              44/52 is 0.8461538

              Wheras you did 52/44 which is 1.18181 [ easily done ]

              Somewhere in that first gear train of 0.8461538 you need a X 3 and not a doubling up to get 0.8461538 x 3 = 2.5384614

              Sir John , Earl of Bligeport & Sudspumpwater. MBE [ Motor Bike Engineer ] Nottingham England.


              • #8
                A 63 tooth gear is all you need for a small lathe ... if you have a standard 80 tooth that is common with most gear sets. 80/63 is 1.2698... We had a 100/127 but it needed a special gear end cover as they were so large in diameter. The 80/63 combo fit inside the standard cover. I think this is what Myford uses as I have seen them for Myfords for sale on Ebay????



                • #9
                  I'm reading "Screwcutting In The Lathe" by Martin Cleeve: I'm in the chapter on metric threads now. He has a good approach to the math and even gives ways to check the calculations.

                  He has a table of 46 ratios that approximate the 1.27/2.45 exact figure, but the 44/52 is absent. Apparently this is not a closed subject and new numbers can still be found.

                  He also brings up an interesting point concerning the accuracy of lead screws. The arguement is that if your lead screw is off by a few thousanths, and who's isn't, why should the gearing ratio be any better. The problem with this is that errors can add together to produce a bigger error than either one alone would produce. A few thousanths here, a few thousanths there and pretty soon you have a few hundreths. But I guess it works out in most cases.

                  He also explains the gearing for cutting cutting worms to match any DP, threading dial theory and operation, etc.

                  This book is very good and I strongly recommend it.

                  Paul A.
                  Paul A.

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


                  • #10
                    Myfords use two 21T gears for metric although you can use a lot of different permutations.
                    It all depends on the operator and what he thinks he will need as regards precision.
                    The 127 gear is the lowest number that is theoretically right for metric threading but all the other ratio's put forward are probably well within any errors the machine can hold to.

                    After all has anyone ever had a lead screw off a Myford, South Bend, Atlas, Clausing etc tested for ACTUAL accuracy ??

                    These were just production parts when built, chances are the lead screw errors are greater that the errors caused by the gear train.

                    Next thing is, what is going on these threads ?
                    A nut that's mass produced by the million per day by stamping? or it's mate that's cut on the same machine, with the same errors cancelling out.

                    Perhaps it's a threaded component made with a tap or die, but to what specs and by what company?

                    Does the Great Wall Silk Flower and Tractor Company, [ One Hung Low province, ] use a 127 gear or do they use the nearest ratio they have to hand.

                    I have to smile when this comes up because it's always over the 25.4 metric transposition ratio but what happens when you cut say 32 tpi on an imperial lathe?
                    NO ONE ever questions this but the same errors except the ratio come into play as regards lead screw, slop in the machine, tight spots on the carriage etc.

                    They must exist, in fact they do exist or why did the top end makers do corrected leadscrews.
                    Some went to the troble of having the zero mark floating and being moved by a hand scraped follower slide to match any errors in a GIVEN lead screw.

                    Here's a classic example of tolerance and machining.
                    Take a 3/8" UNC nut and bolt, 3/8" diameter x 16 tpi x 60 degrees thread angle.

                    Now take a 3/8" British Standard Whitworth nut and bolt, 3/8" x 16 tpi x 55 degrees thread angle.

                    Try interchanging them and you will find that whatever the combination they still fit even though they are different and the laugh is the feel the same in the fit.

                    John S.

                    Just read Paul's post above which went on whilst I was typing and he said roughly the same thing.
                    I will second the Martin Cleeve book, if there are two books in that series that are really worth reading it's the Martin Cleeve one on Screwcutting and the Ivan Law one on Gearcutting.

                    [This message has been edited by John Stevenson (edited 09-18-2004).]

                    Sir John , Earl of Bligeport & Sudspumpwater. MBE [ Motor Bike Engineer ] Nottingham England.


                    • #11
                      If you're like me and take spare moments from close friends in a fine restaurant to work gear ratios with a calculator on a napkin then you are truely a geek.

                      That 80/63 metric translation gear ratio results in 0.00013 error in in inch, plenty close enough for all but the most crictcal measuring instrument axis drive threads. I like if becuase you need only one gear.


                      • #12
                        For what it's worth, here's a list of all the gear pairs between 15 and 100
                        teeth that will produce the 1.27 (= 2.54/2) ratio within 0.5%.

                        Desired ratio = 1.27
                        (number in parens is error in per cent)

                        19:15 => 1.266667 (-0.262467%)
                        28:22 => 1.272727 (0.214746%)
                        33:26 => 1.269231 (-0.060569%)
                        37:29 => 1.275862 (0.461580%)
                        38:30 => 1.266667 (-0.262467%)
                        42:33 => 1.272727 (0.214746%)
                        43:34 => 1.264706 (-0.416860%)
                        47:37 => 1.270270 (0.021281%) **
                        51:40 => 1.275000 (0.393701%)
                        52:41 => 1.268293 (-0.134434%)
                        56:44 => 1.272727 (0.214746%)
                        57:45 => 1.266667 (-0.262467%)
                        61:48 => 1.270833 (0.065617%)
                        62:49 => 1.265306 (-0.369597%)
                        65:51 => 1.274510 (0.355103%)
                        66:52 => 1.269231 (-0.060569%)
                        67:53 => 1.264151 (-0.460556%)
                        70:55 => 1.272727 (0.214746%)
                        71:56 => 1.267857 (-0.168729%)
                        74:58 => 1.275862 (0.461580%)
                        75:59 => 1.271186 (0.093421%)
                        76:60 => 1.266667 (-0.262467%)
                        79:62 => 1.274194 (0.330201%)
                        80:63 => 1.269841 (-0.012498%) **
                        81:64 => 1.265625 (-0.344488%)
                        84:66 => 1.272727 (0.214746%)
                        85:67 => 1.268657 (-0.105770%)
                        86:68 => 1.264706 (-0.416860%)
                        88:69 => 1.275362 (0.422230%)
                        89:70 => 1.271429 (0.112486%)
                        90:71 => 1.267606 (-0.188533%)
                        91:72 => 1.263889 (-0.481190%)
                        93:73 => 1.273973 (0.312803%)
                        94:74 => 1.270270 (0.021281%)
                        95:75 => 1.266667 (-0.262467%)
                        97:76 => 1.276316 (0.497306%)
                        98:77 => 1.272727 (0.214746%)
                        99:78 => 1.269231 (-0.060569%)
                        100:79 => 1.265823 (-0.328915%)

                        As can be seen, 80:63 is the best with 47:37 being second best with smaller
                        physical envelope.

                        Regards, Marv

                        Home Shop Freeware - Tools for People Who Build Things
                        Regards, Marv

                        Home Shop Freeware - Tools for People Who Build Things


                        • #13
                          Marv wins.


                          • #14
                            All that does is prove that Marv has a bigger napkin

                            Sir John , Earl of Bligeport & Sudspumpwater. MBE [ Motor Bike Engineer ] Nottingham England.


                            • #15
                              I made a transposing set for my Atlas which will fit in the gear cover. The DP is 32 so the gears are 1/2 the diameter of the stock change gears. I used 127 tooth on the screw and a 40 tooth engages it. That makes one turn of 40 tooth gear moves the carriage one millimeter. The lead screw is 8tpi. If a different screw is used simply multiply it's tpi by 5 for the spur gear to use.
                              If .5mm thread is desired set change gears to drive spur gear 1/2 turn per 1 turn of spindle. And so forth.