Announcement

Collapse
No announcement yet.

Trouble indexing carriage to cut rack (newbie question)

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

  • Trouble indexing carriage to cut rack (newbie question)

    Hi..
    I'm a hobbyist with little experience and could use some advise.

    Recently, using my home lathe, I attempted to make a small rack gear to repair a friend's
    microscope. I used a milling attachment to hold a small plate of brass, about 0.200" thick.
    The cutting tool was a short length of 1/4" HSS ground to the 40 deg angle and mounted
    into a 1/2" boring bar as an arbor. This boring bar was then mounted through the spindle
    with a 3 jaw chuck. The milling attachment mounted on the saddle was fed parallel to the
    boring bar and indexed by the pitch of the rack. To index the carriage, I mounted a dial indicator
    (Chinese 1") on the ways and fed the carriage 1.5mm for each rack tooth. I also would lock the
    carriage before each cut.

    Results *looked* nice but I found that my indexing was off (compounding error). Since my indicator
    is only 1", I would cut about 9-10 teeth and then need to reposition the dial indicator to cut another
    9-10 teeth. I needed to do this about 4 times in order to cut the total number of rack teeth.
    I was as careful as possible to insure accuracy but there was an accumulating error that ruined the work.
    I used a 1.5 mm pitch gauge to check the end result.

    So, I am wondering how I managed to create such an error. I realize that my cheap indicator
    is not particularly precise (+/- 0.001" or so). The indicator does have some non linearity
    to it. I checked and saw that is some cases, for each 0.100" of travel it may be off 0.002", while
    sometimes it would right on. Would it be to my advantage to 're-zero' the indicator after each
    tooth cut? Maybe my indicator is not accurate enough for this task? What are my other options
    for accurately indexing the carriage?

    Hopefully, this post hasn't been too long winded. Any help that you can offer would be appreciated.
    Thanks.
    Jim Flanagan

  • #2
    Welcome to the forum, more help will be along soon (or may have already been).

    Pretty green at this myself but first question re: your OP, when you repositioned the dial indicator, how do you know you got it in the same relative spot for each move? That could account for some error.

    Maybe way off on this idea but can you set up some sort of stop based on the previously cut tooth?
    So you would make a male fitting to fit in the space just cut, put it into that space but with calculations and a stop along the direction of travel this "fitting" would bump up against said stop which is where you would cut the next "space", continue to completed. "Indexing" as it were...

    Comment


    • #3
      It may be a bit far out but if your lathe can be geared to cut a 1.5MM thread pitch you might be able to cut it like a thread. You would probably need to angle the work so the cut was straight across, start with a shallow cut and feed up for the remaining cuts. Or you may be able to feed back to front with the vertical set for full depth. You would have to leave the feed engaged and reverse the lathe between cuts.

      Another possibility would be to gear for 1.5MM pitch, leave the feed disengaged, and turn the stud gear a full turn between cuts. This should work if the spindle to stud gear ratio is 1/1, otherwise figure out how many teeth to turn the gear for a 1.5MM feed.
      Don Young

      Comment


      • #4
        I would try a 6in digital caliper. Find a way to attach to your set-up, and cut without having to re- index. Bob.

        Comment


        • #5
          If the original broken rack is available, you might try using that as an index.

          Comment


          • #6
            Originally posted by rfdes View Post
            Results *looked* nice but I found that my indexing was off (compounding error). Since my indicator
            is only 1", I would cut about 9-10 teeth and then need to reposition the dial indicator to cut another
            9-10 teeth. I needed to do this about 4 times in order to cut the total number of rack teeth.
            I was as careful as possible to insure accuracy but there was an accumulating error that ruined the work.
            I used a 1.5 mm pitch gauge to check the end result.
            Jim Flanagan
            Could you be more specific about "ruined the work"? Thick teeth? Thin teeth?

            I can't imagine lead errors on the order of thousandths effecting the operation of a manually operated rack.

            I would be inclined to suspect the boring bar moving axially in the chuck. Any axial thrust on the bar, from say different clearance angles on respective sides of the bit, could cause the bar to shift in the chuck. This would be encouraged by the interrupted nature of the cut.

            Dave

            Comment


            • #7
              Agree with Don- cut it like a thread. There is going to be some helix angle to the rack teeth, but the wider the radius the cutter sweeps out, the less this angle will be. It might be a good idea to mount the cutter on a long bar, with the tailstock center supporting that end. The bar can be significantly large in diameter, giving rigidity as well as a means to get the cutter as far from the spindle axis as practical. This is with the rack held stationary on the carriage somehow.

              You can also mount the rack to the side of the bar and mount the cutter on the toolpost like you would normally do for threading. In this case, the challenge is to mount the rack securely.

              It's probably easier to have the cutter rotating, since you already have a milling attachment to mount the rack in.

              I made my own tailstock-supported bar to hold a cutter for boring with the workpiece held to the cross slide. I drilled a cross hole in the center of the bar to mount the cutter into. The hole weakens the bar a little, but if the diameter I'm boring is relatively small, there isn't much option. If the hole is considerably larger than the diameter of the bar, then the cutter could be mounted on its own piece of steel, which is secured to the OD of the bar. Just make sure the cutting edge is positioned to give a neutral to positive rake. This could be a worthwhile project in itself.

              This has brought up a couple more ideas- if your boring bar can be machined on one end to fit the spindle taper, then a collar machined behind that, you end up with one end of the bar snug in the taper and the chuck jaws bearing down on the collar. That's going to be more rigid than relying on the chuck jaws only. I just did a similar thing for a temporary arbor- I didn't machine the taper, but I did prepare one end of the bar to fit the through hole in the chuck snugly. It made a noticeable difference to the rigidity of the arbor. In your case it would let you extend the cutter out from the chuck a bit more, which might help.
              I seldom do anything within the scope of logical reason and calculated cost/benefit, etc- I'm following my passion-

              Comment


              • #8
                What makes you think the proper pitch is 1.5mm?

                Comment


                • #9
                  Hi..
                  I used a metric thread gauge to measure the existing, brass, rack. The rack is only about 2" long and 0.2" wide.
                  There are approx 30 teeth. Do you believe that maybe I could be mistaken with my measurement?

                  thanks for all the responses that I have received.

                  -jim

                  Comment


                  • #10
                    Can you mesh the teeth of the new rack into the teeth of the old rack and see how they compare/line up?

                    How does the new rack not function, exactly??

                    How about a picture of the old rack and the new rack side by side?

                    We'll help you get it sorted out.

                    doug

                    Comment


                    • #11
                      Your setup and procedure seem fairly sound. Why do you think it's a cumulative error? What are the indications that the gear is not correct?

                      Comment


                      • #12
                        I can readily see an error when I place the original rack gear side by side with my homemade
                        version. After about the 4-5 tooth the tooth spacing starts to widen until I'm about a complete
                        tooth off after, say, the 10 teeth or so. This weekend I plan on modifying my setup so that I don't
                        have to re baseline my 1" indicator. I should mention that I lockdown the carriage each time
                        I reindex for the next tooth. I'll attempt to add a photo tonight.

                        Comment


                        • #13
                          A metric 1.5 mm pitch rack is 16.933 teeth per inch. If the old rack was really 16 tpi, then after 10 teeth, the distance (15mm in metric) would be 10/16 *25.4 = 15.875mm. If the old rack was really 18 tpi, the dictance for 10 teeth would be 10/18 *25.4 = 14.11mm.

                          So, in the first case you would be .875 of a tooth out, and in the second case you would be .899 of a tooth out. Whichever is shorter will give the game away.

                          So, you need to either measure the old rack more accurately, or decide on the basis of the history of the microscope, where it was made etc, whether it is likely to conform to imperial or metric standards.

                          What do other things measure ? Not things like the barrel diameter, or lens threads - they may be optical standards. But measurements of things that will be proprieotry to the manufacturers.

                          ANother point is that you should not be using your rack if your lathe has a leadscrew. If you find the pitch is the same side of the English Channel as your lathe, fit a handwheel to the righthand end, with some a division plate made to match the rack. If not, as mentined above, fit a handwheel or division plates to part of the change wheel train so that what is left incorporates the matric/imperial conversion, and index with that.
                          Richard - SW London, UK, EU.

                          Comment


                          • #14
                            Originally posted by rfdes View Post
                            Hi..
                            The rack is only about 2" long and 0.2" wide.
                            There are approx 30 teeth. Do you believe that maybe I could be mistaken with my measurement?

                            -jim
                            30 teeth over 2 inches = 15 teeth per inch. 1/15 = .0667 inches per tooth. That's a pitch of 1.69mm. That's not *that* far off from 1.5mm in terms of guesses and swags. If there are 34 teeth, not 30, you'd be ok.

                            No way to ensure the pitch without measuring carefully.

                            Comment


                            • #15
                              When ever I had to cut a rack I usualy would set my indicator (thou or tenths to zero againt the mill table. Cut a tooth back off and to move to the next tooth I would use a gage block that equaled the DP or Module of the rack being cut and place that against the mill table and bring the table and block up to the indicator to zero without moving the indicator. Tiresome and repetitive but it worked. We did this primarily on a Horizontal mill. Isn't there rack stock available from Small Drive Products or Thompson?
                              Forty plus years and I still have ten toes, ten fingers and both eyes. I must be doing something right.

                              Comment

                              Working...
                              X