No announcement yet.

Metric quick change gearbox mystery

  • Filter
  • Time
  • Show
Clear All
new posts

  • Metric quick change gearbox mystery

    My Harrison lathe has a qc gearbox which will cut metric or inch threads. The leadscrew is 4 TPI. Looking at the qc gearbox parts assembly, I do not see any multiple gear ratios which would give the 127 factor needed for "perfect" metric off the inch leadscrew. Putting a digital scale on the carriage and measuring over 20, 30 or 40 threads, there is ZERO error.

    Does anyone know what sort of gear combinations the Brits used in these things to get this (perfect??) conversion. I can see where a multiple ratio scheme might work where one ratio is on the low side but the next is on the high side ... but perfect cancellation ??

    Paging Marv Klotz ?


  • #2
    Out of curiosity, what is changed to switch?

    Whatever lever is moved or flipper is flipped to change over obviously has to be the key to the mystery. Start there.

    Keep eye on ball.
    Hashim Khan


    • #3
      Take a look at this thread.



      • #4
        On the Harrison we have at work (urrently set up for some 12mm ISO Acme nuts I'm making) to cut metric threads you have to install the appropriate gear train (establishes the 127 tooth ratio I think through four compoinded gears) and once you engage the lead screw you cannot disengage it untill you are done. The nuts are in Ampco 18 by the way
        Forty plus years and I still have ten toes, ten fingers and both eyes. I must be doing something right.


        • #5
          This link may have some relevant info...


          [This message has been edited by Bruce Griffing (edited 12-01-2004).]


          • #6
            Here's the M300 gearbox from Tony's UK site:


            The pitch (or feed) is selected by front panel table and 4 knobs such as "C 1 T W"

            Until or unless I get brave and take it off for a cleaning, I'll never know for sure but it appears after more calculations that some of the gear ratios, such as the one for 1mm, may have an error down around 0.002%. That would explain why my 40 turns at 1mm pitch gave 40.00mm of digital readout movement.

            The error may be lurking in the third decimal place ... just fine for my needs

            In the image, the 2nd shaft from the top (with coupler) is the leadscrew, the one below it is the feed shaft and below that are the gear "selectors" which I believe are moved by cams off the 4 knobs.

            There are all sorts of gears in there including a bunch of 22T:

            Top shaft (idler??): 19, 50, 38, 23, 19, 35

            Lead shaft: 19, 20, 22, 19, 20, 24, 23, 27, 24, 28, 26, 38, 35

            Feed shaft: 22, 22, 22, 33, 22, 36

            Large gears from spindle to qc (not shown): 40:95 and 44:88

            Why do gearboxes and headaches always appear in the same threads Thank God they all work !


            [This message has been edited by nheng (edited 12-01-2004).]


            • #7
              Running my gearfind programs for a ratio of 1.27, I find the following

              47:37 => 1.270270 (0.021281%)
              80:63 => 1.269841 (-0.012498%)

              for a single gear pair and

              24:22 - 78:67 => 1.270014 (0.001068%)
              26:23 - 91:81 => 1.269995 (-0.000423%)
              53:41 - 56:57 => 1.270004 (0.000337%)
              89:63 - 89:99 => 1.270002 (0.000126%)
              89:77 - 89:81 => 1.270002 (0.000126%)

              for two pairs.

              Note that they're *all* approximations. Since 127 is prime, you can't
              duplicate the 1.27 ratio *exactly* except for the trivial case of
              127:100 and integer multiples of that combination, none of which get you gears
              with less than 127 teeth, which is the presumed goal.

              Regards, Marv

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

              Home Shop Freeware - Tools for People Who Build Things

              Location: LA, CA, USA


              • #8
                Marv, Thanks for the reminder about your program. I forgot that gearfind.exe would use arbitrary gears and additional gear pairs and that is what I was looking for. Is suspect that this machine may use up to 4 pairs (6 including the 2 "fixed" pairs).


                • #9
                  Here is a diagram from Martin Cleeve's "Screwcutting in the Lathe" that shows how you can change the effective lead of the leadscrew with gearing in the quick-change gearbox:


                  Perhaps your lathe employs this concept.

                  [This message has been edited by dkinzer (edited 12-01-2004).]
                  Don Kinzer
                  Portland, OR


                  • #10
                    I think that's similar to what is happening in order to support up to 3 gear pairs.

                    Thanks to Marv's GEARATIO program, it looks like the metric pitches have errors down below 0.003% for a handful I ran.



                    • #11
                      Have you not noticed that the maximat cuts metric, inch, and module just fine using compound gearing.

                      The only trick is reading the tiny black and silver gearing chart inside that dark oily cover...after I screwed up 40 hours of work on a spindle I tore that little bastard off and scanned it into the computer - now I can read the damn thing! Unfortunately I can't blame the chart any more either for messin' up. It's tough being brain-dead.


                      • #12
                        Above I mentioned the nuts I was making this week for a repair job to be done some time in the future. Also note my high buck "scale"


                        With the screws still in the machine I had to have something to fit the nuts to. Nuts were cut in two steps, one pass to depth with a rougher and a finish tool following up behind. With the compound set at 90D to the cross slide and feeding straight out. Sometimes you got to know when to break the rules The thing that worried me was shearing the safety pin in the gear train for the Lead Screw
                        Forty plus years and I still have ten toes, ten fingers and both eyes. I must be doing something right.


                        • #13
                          Dave, I have a Maximat 7 threading chart which I sat and typed into Excel in a large font and did some cleanup. I'll email you a copy. BTW, I think the last time I calculated the Maximat gear ratios, the error for metric was pretty high (but not killer). I found a few, more accurate ratios but don't recall if I ever checked to see if they would fit on the machine ... that is always a good thing

                          Spin Doc, Very nice work !

                          [This message has been edited by nheng (edited 12-02-2004).]


                          • #14
                            Thrud said It's tough being brain-dead.

                            NO it's not Dave I find it quite an advantage Alistair
                            Please excuse my typing as I have a form of parkinsons disease


                            • #15
                              Thanks Den. One thing I did do is to use the steady rest after indicating the work in so that was one thing I did not have to worry about. That was also part of the reason for setting the compound to 90D. The Steady on the Harrison we have (V350?) is a nice unit but the saddle can't get real close to it
                              Forty plus years and I still have ten toes, ten fingers and both eyes. I must be doing something right.