Announcement

Collapse
No announcement yet.

Threading question for the folks with metric machines

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

  • #16
    Metric and imperial threads both use common ratios, and threads per inch is just an inverse ratio to millimeters per turn, so gears can accomplish either easily. A 2 mm lead screw can produce exact thread pitches of 1.0 (1/2), 0.5 (1/4), 0.25 (1/8), 1.5 (3/2), 1.25 (6/5), 3.0 (3/2), etc. For threads per cm, the ratios are just reversed, but they may not be nice decimal numbers.
    http://pauleschoen.com/pix/PM08_P76_P54.png
    Paul , P S Technology, Inc. and MrTibbs
    USA Maryland 21030

    Comment


    • #17
      If the leadscrew is metric then the QCGB will give true metric threads. However some industrial lathes with imperial leadscrews have a more complex gearbox than the plain Norton to provide good approximation to metric without going as far as using a 127 gear. The Colchester is within 0.0002% and the Harrison within 0.0027%. Larger industrial lathes probably have true conversions so it would be important for some applications to know what exactly the lathe is providing however it is likely that other errors in general machining are greater than the error given above for the Colchester.

      Comment


      • #18
        Originally posted by The Artful Bodger View Post

        Do you really think metric is different? Countries that are different to metric include Liberia, Myanmar and of course the USA, most every other country uses the metric system.

        Seems everyone else is out of step?
        Where ever did you get that?

        I like metric for everything except threads. More specifically, I prefer "SI", which is not quite the same as "metric".

        Metric threads are the one place where you get to select the standard you like... and the decimal ratios are a bit odd for thread cutting, as witness the need for extra gears on some threads, and 3 or 4 (or 5?) different gears on the thread dial.

        By the way, you are incorrect. The US officially adopted the metric system 150 years ago, somewhere around 1870. So the US is not different, it is just slower to change over entirely, because we never made imperial measurements illegal, as some countries did. Making alternates illegal tends to produce an incentive to adopt metric, especially if the Guillotine is an option.

        Originally posted by PStechPaul View Post
        Metric and imperial threads both use common ratios, and threads per inch is just an inverse ratio to millimeters per turn, so gears can accomplish either easily. A 2 mm lead screw can produce exact thread pitches of 1.0 (1/2), 0.5 (1/4), 0.25 (1/8), 1.5 (3/2), 1.25 (6/5), 3.0 (3/2), etc. For threads per cm, the ratios are just reversed, but they may not be nice decimal numbers.
        it appeared when I was checking it out that there is not a specific available ratio using two gears, one of which remains the same, as in a Norton box.

        It seems that a selection of gears is needed for the correct ratios.
        Last edited by J Tiers; 02-29-2020, 07:08 PM.
        1601

        Keep eye on ball.
        Hashim Khan

        Comment


        • #19
          It appears that, using two gears, you need a selection of about 7 or 8 different gears as gear 1, along with a selection as gear 2. I know that the picture shows gear 2 as in some cases of fine threads being toothcounts that are totally impractical, that makes little difference.

          The point remains that the metric gearbox has to have a selection of gears for both "gear 1" and "gear 2", with no simple set of changes to gear 2 to cover a range.

          The imperial norton box has a simple set of gears as "gear 2" that repeats several times in different ranges with a 2:1 division on the input for each range, and no need to change "gear 1" over the range of pitch steps.

          There IS a 10:1 "step" that s[lits the range. And the sequence "sort of" repeats, but there is a difference, because some of the gears do not divide by 10 evenly in toothcount, so those 10x larger pitches (2mm and 5 mm) need a different set of gears, a simple division by 10 does not work the way 2:1 would for an imperial Norton gearbox. And, some exist in larger pitches, but not in smaller. There is apparently no 0.55 mm pitch,but there is a 5.5mm pitch.

          The metric fixation on decimal seems clumsy.

          Here is the chart I came up with. Again, ignore the large toothcounts, and just notice that any range needs different "gear 2" AND "gear 1" choices for what would be a contiguous range on an Imperial Norton type box.

          Making a gearbox for metric? Yea, YOU can DO it, but it takes some finagling........



          Metric reference, exhaustively complete:

          http://mdmetric.com/tech/M-thead%20600.htm
          Last edited by J Tiers; 02-29-2020, 08:03 PM.
          1601

          Keep eye on ball.
          Hashim Khan

          Comment


          • #20
            Here are some of the standard metric pitches a SB with the transposing gear set can make. With the 127/100 set, they are exact pitches.



            Click image for larger version

Name:	metricgearchart.jpg
Views:	75
Size:	147.8 KB
ID:	1858882

            Comment


            • #21
              Interesting thread Jerry
              A bit different perspective but I believe it follows your original question about gear train accuracy.

              In studying gear boxes, both Imperial and Metric, one has to remember we have a common numbering system for both which is called Base 10
              Both threads are based on a Linear measurement, where Metric measures Pitch from apex to apex of a single thread and Imperial measures the number of apex's per inch
              What if you measured Imperial like you measure Metric ?
              You quickly realize that there is a problem with gear ratios- lets start with a 1 thread per inch and follow it { The Numerator (top) is one inch and the Denominator is # of threads}
              1 / 1 = 1
              1 / 2 = .5
              1 / 3 = .3333333333
              1 / 4 =.25
              1 / 5 = .2
              1 / 6 = .166666666
              1 / 7 = .142857142.
              1 / 8 = .125
              1 / 9 = ,1111111111
              1 / 10 = .1
              So what you see is this number system 1,2,4,5,8,10 as almost prime numbers in the sense that their functions are finite ( they end with 5 decimal places or less)
              If you carry it further , any combination of two of THESE numbers also work , like 2 x 8 = 16 so 1/16 = .0625
              or 8 x 10 = 80 so 1 / 80 = .0125 and so on.
              All other ratios become endless .

              Unfortunately the Imperial thread system has odd balls ( like 27 TPI or 11 1/2 TPI ) and that makes gear box design difficult
              But guess what ?
              Metrics have the same problem. Substitute Metric for the above table and you can see the exact same pattern ( Base 10 remember ?)
              so a variation of 1,2,4,5,8,10 works perfectly gear wise , but when you have .7 or .75 or 1.5 , you have problems

              Now before you get your Nickers in a twist, realize that the challenge is not insurmountable and designs in gear ratios come into effect, but there is no perfect solution and when you
              put both systems in the same lathe , regardless of the 127 conversion factor, errors can occur as Jerry commented on
              Just my thoughts
              Rich
              Green Bay, WI

              Comment


              • #22
                Why is that. I have explained it in some detail several times. And it is one area where the good old, English measure stuff is actually better than the metric. Here are the quick strokes:

                English measure threads are defined in Threads PER INCH. That "per inch" means that there is a reciprocal relationship between the actual number, which is almost always chosen as a nice, round number, and the actual, linear pitch. So a 20 TPI thread is actually 0.050" linear pitch between adjacent threads.

                But metric threads are always defined in the simple, linear pitch. NO RECIPROCAL relationship. A 1mm thread has 1mm between adjacent threads. And again, they choose nice round numbers but because the mm is a lot smaller unit, those numbers are mostly round tenths or twentieths of a mm (0.5mm, 0.75mm, etc.)

                These two different ways of defining threads produce mathematical sequences (and therefore gear ratio sequences) that are fundamentally different. Therefore a QC gear box that has the appropriate gearing for English threads will not have the same sequence needed for metric threads and vice versa. This basic, mathematical difference is why you not only need a 127 tooth gear to convert from one sequence to the other, but you also need assorted additional gears to convert the actual spacing or ratios of the sequences.

                PS: If a lathe has a metric lead screw, there is no reason why all the threads can't be "exact". Now, does every manufacturer do that? Who knows? As far as I can tell, all English lead screw lathes and their QC gear boxes do give "exact" threads.

                I put the work exact in quotes above because there are a host of other factors involved in the precision of a thread.

                Oh, why did I say that the English threads were actually better. Well, it is a lot easier to design an English thread QC gearbox. And you have a lot more threads in the English measure sequence than in the metric.



                Originally posted by dian View Post
                "Classic Norton gearbox Style doesn’t match well with the metric pitches."

                why is that?
                Paul A.
                SE Texas

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

                Comment


                • #23
                  Actually, I sort of screwed up. In the chart, for the 0.25 pitch, 15 and 240 teeth works just as well. And the 0.5mm would be just as well with the 15 and 120 tooth combo.

                  That will actually allow the two ranges to use the same set of "gear 1" sizes, except that there IS a 5.5 mm pitch, but 0.55 mm is not used. It could be included just to make the two series the same, nobody would care if there was an unused setting.

                  The SB chart has several unused non-existent thread pitches included in it (0.55, 0.65, and 1.1mm), and misses some oddballs (1.12mm, 1.4mm, 1.6mm, 1.8mm and a bunch of watch threads under 0.2mm). Again, no issue if they are there, no need to use them.

                  But the SB needs a different stud gear for basically every one of the common pitches, although 48, 42, 40, 36, 32, 28, and 26 seem to recur scattered through the range up to 2.5mm (roughly 10 tpi). But you need a different one of them for every pitch over a range that a Norton Imperial QCGB handles easily with only one lever movement.
                  1601

                  Keep eye on ball.
                  Hashim Khan

                  Comment


                  • #24
                    My Harbor Freight 9x20 lathe is set up primarily for imperial threads, as follows:

                    Click image for larger version  Name:	HF_Lathe_Thread_Chart.jpg Views:	0 Size:	52.0 KB ID:	1858912
                    The metric threads require a lot more changing of gears, because the gears in the QCGB are set up for imperial. It should be possible to use different gears designed specifically for metric. Here are some metric sizes that can be obtained with a 9 position QCGB.

                    Code:
                    6.00 3/2  1.5000 24 16
                    5.50 11/8 1.3750 22 16
                    5.00 5/4  1.2000 20 16
                    4.50 9/8  1.1250 18 16
                    4.00 1/1  1.0000 16 16
                    3.50 7/8  0.8750 14 16
                    3.00 3/4  0.7500 12 16
                    2.50 5/8  0.6250 10 16
                    2.00 1/2  0.5000  8 16
                    
                    2.50 5/8  0.6250 20 32
                    2.00 1/2  0.5000 16 32
                    1.75 7/16 0.4375 14 32
                    
                    1.50 3/8  0.3750 12 32
                    1.25 5/16 0.3125 10 32
                    1.00 1/4  0.2500  8 32
                    0.75 3/16 0.1875  6 32
                    
                    1.60 2/5  0.4000 16 40
                    1.50 3/8  0.3750 15 40
                    1.00 1/4  0.2500 10 40
                    0.80 1/5  0.2000  8 40
                    
                    0.70 7/40 0.1750  7 40
                    0.60 3/20 0.1500  6 40
                    0.50 1/8  0.1250  5 40


                    There is a lower limit to the number of teeth on a gear, so probably the 6 and 5 tooth will not work. These gears were adapted from JTiers' chart above.
                    Last edited by PStechPaul; 03-01-2020, 12:19 AM.
                    http://pauleschoen.com/pix/PM08_P76_P54.png
                    Paul , P S Technology, Inc. and MrTibbs
                    USA Maryland 21030

                    Comment


                    • #25
                      My imperial Boxford with an 8TPI leadscrew and QCGB cuts MOST common metric threads EXACTLY with the a change from 80 to a 127/100 gear and ALL common metric threads with a change of one more gear EXACTLY
                      Pretty much like the SB chart above..

                      Comment


                      • #26
                        I'm spoiled, I have a Holbrook Model C - the different approaches (Imperial turns per length, metric length per turn) are handled really well - swapping between the two swaps drive and driven between the cone of gears and the plain gear in the Norton box, although it's necessary to swap in a 127-tooth gear (on the other shaft) to replace the 45-tooth for Imperial threads. Simple. Effective. Versatile. Takes 2 minutes and moving a lever.

                        Click image for larger version

Name:	C13 QC Gearbox Plate.jpg
Views:	72
Size:	178.0 KB
ID:	1858930

                        There are some never-heard-of-'em oddball metric pitches, but all the standard ones are there, and who knows, might need a 1.625mm pitch one day, if only to make spare parts unobtainable

                        Dave H. (the other one)
                        Rules are for the obedience of fools, and the guidance of wise men.

                        Holbrook Model C Number 13 lathe, Testa 2U universal mill, bikes and tools

                        Comment


                        • #27
                          I added a 125/127 gear to the Smart & Brown model A with imperial leadscrew and with a few more gears got close approximations of the common metric pitches. This is plenty good enough for screws and nuts, but would not be any good for something like a leadscrew. They are within 1% of the exact pitches.

                          Comment


                          • #28
                            Originally posted by old mart View Post
                            I added a 125/127 gear to the Smart & Brown model A with imperial leadscrew and with a few more gears got close approximations of the common metric pitches. This is plenty good enough for screws and nuts, but would not be any good for something like a leadscrew. They are within 1% of the exact pitches.
                            If you had added a 100/127 or a 120/127 gear combo, your threads would have been exact. I've never heard of a 125/127 gear combo doing anything for imperial to metric conversion.

                            However, plugging in the 125/127 gear set shows that you CAN get exact threads for some of the metric threads. It's just that to get them ALL exact would require a lot more gears. For the 100/127 gear set, you only need 7 extra gears to generate the exact threads on the most common metric threads. I don't know what the standard metric threads are, but I can create exact threads for all the leaves in my metric thread gauge set and then some.

                            Comment

                            Working...
                            X