Machining a Watch Stem on a Lathe.

Discussion in 'Horological Tools' started by Jerry Kieffer, Aug 3, 2013.

  1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.
  2. NAWCC Login Function is Not Working. If you know your username and password for the Message Board, you can log in using the small login link at the upper right.
  1. Jerry Kieffer

    Jerry Kieffer Registered User

    May 31, 2005
    #1 Jerry Kieffer, Aug 3, 2013
    Last edited: Aug 3, 2013
    I recently received a request for a procedure to machine square flats a watch stem on a machine Lathe and thought the answer may be of interest to some.

    The square flats on watch stems have been traditionally filed on a watchmakers Lathe with the use of a file rest and indexing of the spindle.

    However, when using machine tools, one would normally use the appropriate size Lathe and Milling machine that are compatible with each other. Stock would then be placed in a collet and the collet would be mounted in a indexer on the mill for the simple machining of the flats. The collet and stock would then be transferred back to the Lathe to machine the rest of the stem.

    In this case, I understood the question to be how one would machine the square flats with just the Lathe. While the above method is preferred, it is quite simple to do on just the Lathe if you observe a couple of cautions.

    My personal method on just a Lathe is as follows with a couple of quickie photo`s.

    (1) I mount square carbon steel stock designed to be machined in a self centering four jaw chuck. The stock is then quickly machined in steps per the first attached photo with the smallest step sized and length for machining the square flats only. The steps support the machining process.

    (2) The square stock is then mounted in the tool post set 90 degrees from the spindle center line per the second photo. A small Endmill with maximum diameter of about 1.5 times larger than the distance across the flats is then mounted in the spindle also per the photo. Of course the stock is shimmed to center on the endmill.

    (3) from this point, flats can be machined on the end of the stock utilizing handwheel controlled slides and handwheel settings for sizing the flats. After each flat is machined, the square stock can be rotated 90 degrees in the tool post until the procedure is completed.
    The third photo shows relative size of a endmill in relation to stock size when highly accurate micro machining is required. The stock shown is .040" and the endmill is .046".

    This is an especially effective method if for some reason one needs to machine a tapered square surface.

    One caution.
    Make sure that each time the square stock is mounted in the tool post, that it is securely mounted against the rear vertical surface of the post or accuracy of the machined square will suffer.

    Jerry Kieffer

    Attached Files: Download all post attachments

    Max Phillips likes this.
  2. motormaker

    motormaker Registered User

    Apr 5, 2010
    That is a nice solution to the problem. Thanks for sharing.
  3. Dushan Grujich

    Dushan Grujich Registered User

    Jun 20, 2003
    G'Day Gerry!

    Majority of wrist watch winding stems are between 40/100 mm - 70/100 mm (0.016" - 0.030") across flats and often are 3 mm - 5 mm long (0.120" - 0.200"). What about flexing of the work-piece when a square part of stem is machined without a support at the other end, using the above described method?

    I am also curious about flexing of the milling cutter as it appears to be fairly long and is of about the same diameter, or little larger, than the square being milled. Of course, I assume that You are milling "drill rod" quality steel that is cold drawn thus it is rather hard itself, without annealing.


  4. Jerry Kieffer

    Jerry Kieffer Registered User

    May 31, 2005
    #4 Jerry Kieffer, Aug 5, 2013
    Last edited: Aug 5, 2013
    Actually a very good question.

    Micro machining is something that must be experienced to be understood, but I will try to explain.

    First, you are correct in that a center may be required depending on the procedure. For example on a watchmakers lathe with a file rest, you are applying file pressure across the full length and width of the flat at a very slow rate causing a high degree of stress on a small part. Under these conditions, support on both ends is required for accuracy.

    However, when Micro machining Micro parts with machine tools, I find that the use of support is far less efficient and accurate than support on both ends. I find that starting out with a large piece of stock and leaving support steps on the work piece to support each machining step to be far superior in all respects. Again however, this requires a Lathe and Mill that are compatible and except each others accessories. In this case I would normally mount the stock in the lathe chuck, transfer the chuck to a indexer to machine the steps, then back to the lathe for final machining. The stock would never have been removed from the chuck until the completed part was parted off.
    With a machine Lathe, large amounts of metal can be removed quickly and easily making the process even more efficient. In this case, the initial steps were machined in less than 60 seconds.

    In regard to flexing.
    Each step of the work piece is left heavy enough to support each machining procedure without flexing.
    In this case, I measured across the corners of the flats of the stem shown in the attached photo. It measured .040" so the stock was machined slightly over .040" at about .042" and only as long as required to machine the flats. Across the flats it measured .029"
    Again in this case, only about 10 percent of the of the stock is removed when each flat is machined. The other 80 or 90 percent has several times more strength than is required to support the very minimal machining stress.
    The reason machining stress is so minimal besides quality tooling, is the way it is utilized. Yet again in this case, the cutting action of the endmill is across the flat only not over the whole length. In addition, the small .046" endmill has four cutting flutes and is turning at 2500 RPM. The cutting stress under these conditions is so minimal it is undetectable.
    The attached photo should give a little better idea of the proportions of what is being done. Note that the finish left by the Endmill is at least as good as the sample stem.

    I use Carbide tooling for this type work allowing machining of whatever can be machined/worked with hand tools.

    The example shown in the photo`s is A-2 (Air hard) steel, however as mentioned it is not an issue to machine much harder high quality steel if there is an advantage.

    Jerry Kieffer

    Attached Files:

  5. flynwill

    flynwill Registered User

    Feb 1, 2007

    You may be happy to know that a former student of yours (me) used somewhat similar technique when I needed to make a stem for the first time. It was for an Elgin 760 that I was re-casing in a modern case. I did invest in two bits of tooling in addition to the lathe are the vertical milling table, and the index block set. That enabled me to put together the setup shown in the photo and machine the flats with an end mill much as you describe.

    I think I started about 15-20 different pieces before I finally fabricated a part that was good enough. The second picture is one of the middle efforts partially completed alongside the original part.

  6. Jerry Kieffer

    Jerry Kieffer Registered User

    May 31, 2005
    Very interesting and creative setup that importantly allows the work piece to remain in the collet until machining is complete. I am happy to see that you are making use of what appears to be that stub endmill you were forced to purchase. You guys were really a great group.

    Again, I am happy it worked out for you.

    Jerry Kieffer
  7. davestanda

    davestanda Registered User

    May 23, 2011
    this is the set up i tried,like flywheel's but i tried making the indexing tool from sqaure steel stock and a tailstock collet holder from a WW lathe...i used set screws to hold the collet holder....That didn't work because the WW collet holder has ridges ,there wasn't enough surface i milled three sides of the work piece and then the fourth was easy to index afterwards...
Similar Threads - Machining Watch Stem Forum Date
Hand operated watch cleaing machine Horological Tools Dec 3, 2016
Elma Accutron MiniTest ( Mini-Test ) watch testing machine - help. Horological Tools Mar 19, 2016
Anyone use a parts cleaning machine called - SUPERIOR WATCH CLEANER - PITTSBURGH PA Horological Tools Mar 3, 2015
Watchmaster G-D-2 Watch Timing Machine Manual Horological Tools Nov 23, 2014
L&r watch cleaning machine ultasonic vari-matic fault Horological Tools Aug 3, 2014

Share This Page