Gear Hobbing: Servo or Stepper motor - Page 4
Close
Login to Your Account
Page 4 of 8 FirstFirst ... 23456 ... LastLast
Results 61 to 80 of 143
  1. #61
    Join Date
    Apr 2018
    Country
    UNITED KINGDOM
    Posts
    1,303
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    620

    Default

    Quote Originally Posted by CarbideBob View Post
    I wish you luck and it is doable.
    Sure, but the question is, why ? I could go buy a bunch of sand and iron ore to cast up my own small block chevy but when you can go to the dealer and get a crate motor for a cheap price, why the hell would you not do that ?

  2. Likes Oldwrench liked this post
  3. #62
    Join Date
    Feb 2017
    Country
    UNITED STATES
    State/Province
    California
    Posts
    324
    Post Thanks / Like
    Likes (Given)
    5
    Likes (Received)
    72

    Default

    Quote Originally Posted by EmanuelGoldstein View Post
    Sure, but the question is, why ? I could go buy a bunch of sand and iron ore to cast up my own small block chevy but when you can go to the dealer and get a crate motor for a cheap price, why the hell would you not do that ?
    Partly because it is an interesting project. Partly because I don't have the room in my shop for a full gear hobber. Partly because I thought it would be really cool to have a gear hobber accessory for my Deckel to go along with all the other Deckel accessories. And mostly because I want to be able to create gears on the fly without having to order and wait, and also so I can experiment with unusual and exotic gear forms that might otherwise be unavailable.

  4. #63
    Join Date
    Feb 2017
    Country
    UNITED STATES
    State/Province
    California
    Posts
    324
    Post Thanks / Like
    Likes (Given)
    5
    Likes (Received)
    72

    Default

    Interesting PM thread here Converting horizontal mill to gear hobber showing several examples of gear hobbing accessory on horizontal mill using mechanical linkage between the cutter spindle and the rotary. My solution will be using a much higher quality milling machine than the examples shown, and will use electronic linkage, but same basic idea.

  5. #64
    Join Date
    Jun 2008
    Location
    Dassel,MN,USA
    Posts
    67
    Post Thanks / Like
    Likes (Given)
    9
    Likes (Received)
    29

    Default

    FWIW, I got interested in hobbing gears several years ago.

    I used my 2J head knee mill with a Camsoft control and Galil motion card. Set head at helix angle with hob in spindle. Used a 4rth axis to rotate the blank. galil has electronic gearing so slaving the gear to the hob was a piece of cake.

    This setup worked well for light duty gears without super tight tolerances. Better gears than I was able to make on the same machine with involute cutters.

    My son runs a $600K 5 ax machine. it would make wonderful gears.

  6. #65
    Join Date
    Jun 2002
    Country
    CANADA
    State/Province
    British Columbia
    Posts
    2,426
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    1785

    Default

    Hi again drcoelho:
    You wrote:
    "so I can experiment with unusual and exotic gear forms that might otherwise be unavailable."

    Have you considered the cost and availability of the unusual and exotic hobs you will need to be able to hob these unusual and exotic gear forms?

    Is this going to be a significant part of your justification for the device or is it just an incidental benefit?
    I ask because the hobbing process will be likely be a royal pain in the ass if you want to mostly experiment, but will likely be more worthwhile if a typical project for you involves a whole bunch of gears for a gear train and you really don't care about being profitable when you make them.

    If the ultimate in versatility is your goal but productivity is not, getting your mitts on a small used wire EDM would be a logical way forward providing you had the space, the cash and were willing to dive down that whole other rabbit hole.
    When I have metal gears to cut, that's what I think of first, but I already have the toy and can justify the space and money it consumes because it's profitable for me in other ways.

    When I have cluster gears or plastic gears to cut, my GO TO machine is my Haas Minimill with a custom G code writing program done for me by a buddy years ago that allows me to make a single tooth rack profile shaper cutter, and use spindle orient to generate an involute gear using the same strategy as the now ancient and long obsolete Sunderland gear shaping process uses.

    The only mods to the mill are a toolholder modified to allow me to position my single tooth shaper cutter in the proper relationship to my gear blank and lock it in place.
    In 45 years of doing this for a living, I've used the setup maybe 10 times, but when I NEED it, it's been a bacon saver and it's totally frickin cool to watch it dancing around the blank.
    For the simple price of a custom ground (by me) rack tooth with rake and relief, I can get any DP, or pressure angle or anything else I want...a cutter is under an hour to grind on the tool and cutter grinder and it consumes a buck or so in HSS.

    So I encourage you to think about these things before you dive too deep into this project.
    If on the other hand, the coolness of the build is mostly what's attracting you, then my recommendation is to go for it and enjoy it!

    Cheers

    Marcus
    Implant Mechanix • Design & Innovation > HOME
    Vancouver Wire EDM -- Wire EDM Machining

    Last edited by implmex; 06-12-2019 at 11:56 AM.

  7. Likes motion guru liked this post
  8. #66
    Join Date
    Aug 2006
    Location
    Kitchener, on canada
    Posts
    856
    Post Thanks / Like
    Likes (Given)
    15
    Likes (Received)
    192

    Default

    Hi Marcus

    Next time you set that gear hobbing system up on your Haas please take a video to share with us.

  9. #67
    Join Date
    Jun 2002
    Country
    CANADA
    State/Province
    British Columbia
    Posts
    2,426
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    1785

    Default

    Hi SBAER:
    I will do my damndest, but first I have to figure out how to make the video and post the video.

    BTW, this is not a hobbing process, it is a gear shaping process.
    If you are at all familiar with a Fellows shaper, it shares some attributes of those machines.
    It is different however, in that it uses a rack shaped cutter rather than a gear shaped cutter.
    So it cannot make internal gears.
    My adaptation uses a single tooth rack cutter, so the cutter is easy to make accurately.
    Moving the cutter properly relative to the blank allows it to generate a proper involute flank, even though the cutter has straight sided flanks.
    The underlying principle, is that a gear of any tooth count will roll perfectly when meshed with a straight sided rack so long as the PA and DP match.

    Once I get a chance, and figure out how to shoot the video I will make it happen.

    Cheers

    Marcus
    Implant Mechanix • Design & Innovation > HOME
    Vancouver Wire EDM -- Wire EDM Machining

  10. #68
    Join Date
    Aug 2007
    Location
    Holmen, Wi
    Posts
    197
    Post Thanks / Like
    Likes (Given)
    1
    Likes (Received)
    15

    Default

    Again - linuxcnc can do this... He electronically gears the rotory axis to the spindle...






  11. #69
    Join Date
    Feb 2017
    Country
    UNITED STATES
    State/Province
    California
    Posts
    324
    Post Thanks / Like
    Likes (Given)
    5
    Likes (Received)
    72

    Default

    Yes, the above pictures are exactly what I'm planning but with these caveats: higher quality/more accurate rotary and motor, use of PLC for tighter tolerances. The orientation of the setup in the first example above is exactly the setup I'm planning, gear hob on the horizontal spindle, rotary with support on the table. My Deckel has universal table, so I can set up the helix angle easily, and has powered x,y,z so feeds can be setup for smooth operation.

  12. #70
    Join Date
    Aug 2007
    Location
    Holmen, Wi
    Posts
    197
    Post Thanks / Like
    Likes (Given)
    1
    Likes (Received)
    15

    Default

    Andy (the video creator..) says

    You might want to tell him that linuxCNC is working in double-precision floating-point throughout and that is almost certainly far more accurate than a PLC.
    He also said...

    It’s a fair comment about the roatary. I changed that to a harmonic drive (no videos of hobbing with that yet)


    Quote Originally Posted by drcoelho View Post
    Yes, the above pictures are exactly what I'm planning but with these caveats: higher quality/more accurate rotary and motor, use of PLC for tighter tolerances. The orientation of the setup in the first example above is exactly the setup I'm planning, gear hob on the horizontal spindle, rotary with support on the table. My Deckel has universal table, so I can set up the helix angle easily, and has powered x,y,z so feeds can be setup for smooth operation.

  13. #71
    Join Date
    Jan 2007
    Location
    Flushing/Flint, Michigan
    Posts
    7,201
    Post Thanks / Like
    Likes (Given)
    343
    Likes (Received)
    6074

    Default

    Many ways to "get er done".
    At some point you have to ask what level of accuracy to you want to make and how do you check it.
    Down in the guts CNC machine tools work in counts and that is a integer world, "double-precision floating-point throughout" is not so much help and if fact can mess you up.
    I do like linuxcnc but its more about sample times, resolution and system time constants.
    It is also not the easiest system to get up and running.
    Many had such high hopes in the early days of the Open CNC project that would become this. It did not quite pan out.
    Bob

  14. #72
    Join Date
    Dec 2003
    Country
    UNITED STATES
    State/Province
    Washington
    Posts
    4,573
    Post Thanks / Like
    Likes (Given)
    2070
    Likes (Received)
    3812

    Default

    Also, for perspective, the Delta Tau unit is using a DSP563xx processor with dual 24 bit data bus handling floating point math operations natively using a modified Harvard Data bus structure. This is rather old school compared to the newer PowerPC version, but more than adequate for this application.

  15. #73
    Join Date
    Feb 2017
    Country
    UNITED STATES
    State/Province
    California
    Posts
    324
    Post Thanks / Like
    Likes (Given)
    5
    Likes (Received)
    72

    Default

    Accuracy of calculation is not the challenge here IMHO, it is having a real-time operating system. LinuxCNC does have real-time OS options, and I am a Linux expert so Linux install would be a piece of cake for me, but quite honestly this is such a simple application and a PLC can operate with no OS interrupts and therefore you never have glitches related to higher level OS operations that I feel the PLC is better suited. I am very intrigued by the ideas motion guru has suggested for ways to really tighten up the specs on this accessory to get really accurate rotary motion and just don't want all the unknowns of dealing with Linux complexity overhead.

  16. #74
    Join Date
    Aug 2014
    Location
    bainbridge island
    Posts
    1,063
    Post Thanks / Like
    Likes (Given)
    200
    Likes (Received)
    254

    Default

    In many operations once the gear cutter touches the work the load is constant all the way to the end of the cut and theoretically the torque exerted by the cutter on the gear blank is either zero or its in the same direction all the time. A Servo or stepper driving a 40:1 worm gear (with very thick oil) rotary axis would not matter as long as the rotary axis is dynamically stiff enough to prevent chatter. ( likely the reason that linuxcnc guy went from a belt to a harmonic drive)

    But if the load is changing as the gear is cut,(for example a bevel or spiral bevel or a hypoid gear) then the stiffness of the rotary axis will matter and a servo with closed loop encoder could be justified, but you can do that with a stepper too.

  17. #75
    Join Date
    Feb 2017
    Country
    UNITED STATES
    State/Province
    California
    Posts
    324
    Post Thanks / Like
    Likes (Given)
    5
    Likes (Received)
    72

    Default

    Quote Originally Posted by johansen View Post
    In many operations once the gear cutter touches the work the load is constant all the way to the end of the cut and theoretically the torque exerted by the cutter on the gear blank is either zero or its in the same direction all the time. A Servo or stepper driving a 40:1 worm gear (with very thick oil) rotary axis would not matter as long as the rotary axis is dynamically stiff enough to prevent chatter. ( likely the reason that linuxcnc guy went from a belt to a harmonic drive)

    But if the load is changing as the gear is cut,(for example a bevel or spiral bevel or a hypoid gear) then the stiffness of the rotary axis will matter and a servo with closed loop encoder could be justified, but you can do that with a stepper too.
    I'm not too concerned with torque, I am however hoping to get the very smoothest most accurate association of the cutting RPM to the work holding RPM. For this reason, I believe a servo motor is superior to a stepper, and I believe a dedicated PLC with no interrupts is superior to Linux (which has OS interrupts which can cause very tiny errors in synchronization). I'll be doing spur gears only.

    I'm pretty much committed to going with Sankyo RCD rotary, which has zero backlash, with a 60:1 reduction, this is a good match for my milling machine horizontal RPM (Deckel). I'll also be setting the z-axis (cut depth) one time prior to starting the cut, and then using power x to move the work into the gear hob with no interrupts, the idea again to have a very smooth cut without interruption.

    I'm also thinking that I might run the cut using both a roughing gear hob first pass, following by a finishing gear hob, again to improve quality....feedback on this idea is welcomed.

  18. #76
    Join Date
    Jul 2006
    Location
    Hillsboro, New Hampshire
    Posts
    6,333
    Post Thanks / Like
    Likes (Given)
    1519
    Likes (Received)
    4432

    Default

    Quote Originally Posted by drcoelho View Post
    I'm also thinking that I might run the cut using both a roughing gear hob first pass, following by a finishing gear hob, again to improve quality....feedback on this idea is welcomed.
    If you're using a spirally ground hob (like those shown in the videos), just transition in Z by some multiple of the tooth pitch to get to a fresh section of the hob. Adjust Y slightly deeper for cut depth, it's much simpler than trying to sync a second cutter.

    And for goodness sake, get the TIR of the hob better than the guys in the vids, the "grunch" [pause] "grunch" [pause] was driving me batty...

  19. #77
    Join Date
    Apr 2018
    Country
    UNITED KINGDOM
    Posts
    1,303
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    620

    Default

    Quote Originally Posted by implmex View Post
    If you are at all familiar with a Fellows shaper, it shares some attributes of those machines.
    It is different however, in that it uses a rack shaped cutter rather than a gear shaped cutter.
    Marcus : you will be happy to know you have reinvented the wheel

    You know about Sunderlands and Maags, which are still used for certain jobs but there was also a fine pitch machine from Illinois Tool Works that worked the same way - except it had a 3" capacity. They were cool, I had one which eventually went to a fine pitch shop in Los Angeles. If you are searching, it's an ITW "Gerite" gear generator. 4610 was (I think) the model number.

    Also Fellows built some "straight line" generators that used racks to cut fine pitch teeth. They were similar to the #3 fine pitch shapers but instead of a round cutter, they used a rack. I am not sure if they could do helicals tho. The ITW did. I never had one of those but Len Epstein went through several.

    You are absolutely correct that for the stated purpose, the OP's "solution" is dumb. And using roughing and finishing hobs on fine pitch ? Give me a break. How does he plan to make these unusual hobs ? That is an insurmountable problem with this whole excursion into fantasyland.

    (These small shapers take up a space about 3' square. If one were to cnc-ify a fine pitch hobber, at least it would have the right structure and also be around 2' by 2' or less. Space is not a good reason for this.)

    This whole thread is silly and suitable for a hobbyist site. I have no idea why the subject of gears causes people with no experience to come out of the closet with ridiculous ideas, but it sure seems to.

    Oh well.

  20. Likes Oldwrench, Zahnrad Kopf liked this post
  21. #78
    Join Date
    Feb 2017
    Country
    UNITED STATES
    State/Province
    California
    Posts
    324
    Post Thanks / Like
    Likes (Given)
    5
    Likes (Received)
    72

    Default

    I'm gonna buy commercial gear hobs, FYI. This is most definately not a hobby project, just checkout the quality of the components going into this project. I suspect that I'll get better gear accuracy than I could get from an old machine off the used market, but that is subject to verification.

    I looked at the Sunderland approach quite seriously, but the mechanism seems quite complex and finding cutters is a challenge.

    My understanding is that Fellows machines mostly do gear copies, e.g. you need to have a master gear to start with...didn't want to have that limitation.

    Trust me, I DO NOT have room for even a small gear hobbing machine in my shop....just my fate.....

    Finally, I just am not willing to accept that using the quality components I'm planning and have the stiffness of a FPS 300 M milling machine (Deckel clone) that I can't equal the quality of the many small machines you mention.

    EmanuelGoldstein: Rather than being insulting as you are, how about you provide specific technical reasons why the approach I'm taking cannot match or exceed the accuracy of the old school machines you mention?

  22. #79
    Join Date
    Feb 2017
    Country
    UNITED STATES
    State/Province
    California
    Posts
    324
    Post Thanks / Like
    Likes (Given)
    5
    Likes (Received)
    72

    Default

    Quote Originally Posted by Milland View Post
    If you're using a spirally ground hob (like those shown in the videos), just transition in Z by some multiple of the tooth pitch to get to a fresh section of the hob. Adjust Y slightly deeper for cut depth, it's much simpler than trying to sync a second cutter.

    And for goodness sake, get the TIR of the hob better than the guys in the vids, the "grunch" [pause] "grunch" [pause] was driving me batty...
    So, can you clarify for me the preferred technique:
    - option A: set the z for full depth of cut, and (in my case) run X to feed the work piece across the gear hob until full width of gear is cut, is there any benefit from running through a second pass with full depth?
    - option B: set the z for partial depth of cut, then run the X to feed the work piece across the gear, but in this case there are multiple passes of the x, e.g. wind the x and and run through a second or multiple times?
    - option C: my idea of using a different finish gear hob....guess I gotta figure out how to synchronize this to make this work....the trick it seems to me is have a very accurate x axis start point....my Deckel mill will in fact have this capability....

    Ya, gonna spend some time figuring how to get everything trammed up very nicely, most definately some trial and error in this department, might even build some fixturing to make things easier/faster, we'll see...

  23. #80
    Join Date
    Jul 2006
    Location
    Hillsboro, New Hampshire
    Posts
    6,333
    Post Thanks / Like
    Likes (Given)
    1519
    Likes (Received)
    4432

    Default

    Option A may improved tooth surface finish, but doesn't really give you a rough and finish pass.

    I really think "option C" is just not workable - as well as having the expense of a second hob, you have to sync to the previous cut path. Unless you have a precision rotation indexing method for the hobs on the arbor, and both hobs are ground to exactly the same specification except for the roughing hob being slightly smaller in diameter, it's like trying to pick up a thread on a lathe with a new tool, but worse.

    Option B is basically what I was suggesting, except (if we go by the convention that spindle axis is Z, feeding axis is X, and the axis that controls depth of cut is Y), then I'm suggesting "roughing" on one portion of the hob (which is much wider than the final tooth form on the gear), then shifting in Z by an exact multiple of the spiral pitch of the teeth on the hob so that the actual relationship between hob teeth and cut gear teeth on the workpiece is maintained.

    Then you adjust Y (moving the hob closer to the gear blank), and repeat the pass in X. This gives the finish pass.

    I hope this is understandable. I'm near my bedtime so may not be as cogent as I try to be...


Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •