Gear Hobbing: Servo or Stepper motor - Page 2
Close
Login to Your Account
Page 2 of 9 FirstFirst 1234 ... LastLast
Results 21 to 40 of 166
  1. #21
    Join Date
    Jun 2002
    Country
    CANADA
    State/Province
    British Columbia
    Posts
    2,492
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    1847

    Default

    Hi again drcoelho:
    You wrote:
    " when the spindle settles down to a more or less fixed speed, I would then get the rotary spinning in a perfectly matched speed"
    and:
    "
    The quality of my results will be driven I suspect mostly by the quality of the gear hobbers I use,"

    As you know, with hobbing synchronicity of the work spindle and the hob is the crux of the whole problem, so your contention that the quality of the hob will be the driver of the quality of the result somehow misses the importance of this core relationship.

    I know nothing about the intricacies of how you go about this, but it seems to me, that the problem is fundamentally made more difficult by the substitution of "real" gearing between them with "virtual" gearing between them.

    Back in the day when mechanical gearing linked the two spindles, the quality of the drive motor was irrelevant to the quality of the output, because the two were slaved together in a way that could not be disrupted no matter what the drive motor did.
    All you needed was good quality gears.

    Now you have a new circumstance in which the quality of both motors is crucial in a whole new way, and the problem of linking a slave motor to a master motor with all of the potential sources of mismatch involved looms much larger.

    I think it's presumptuous to presume that a standard MANUAL milling machine motor, even a very good one will be optimized to produce a rock steady constant speed under variable load or variable input voltage, because in the case of normal MANUAL milling it's an irrelevant attribute.

    That means to me, that you will have only one of two undesirable options:
    a) Replace the main drive motor on the Deckel with one that IS optimized to be able to produce very constant speed.
    b) have one helluva good way to interrogate the hob spindle in real time (at a rate that is actually relevant to being able to accommodate fluctuations in hob motor speed and compensate with matching changes in workhead motor speed).

    Carbide Bob alludes to the notion that it's not a trivial problem, but you state the assumption that hob speed will be pretty steady, such that your predictive software will be able to make a real time compensation.

    I don't see how you can get there without a super good hob spindle drive motor and I suspect it would have to be a direct drive servo motor to eliminate the spindle motion errors associated with the gearing between motor and spindle in a conventional Deckel unless you can predict those errors, map them out, and compensate for them in your software with some kind of compensation table.

    Yes, I understand interrogating the hob spindle directly theoretically makes that problem go away, but as has been pointed out, the responsiveness of the system limits show fast you can spin the hob and still make a correction in time for it to matter so to make it work you still need very steady predictable hob speed so the compensation doesn't have to do so much.

    So, the 64 million dollar question:
    Is this electronic gearing actually going to be a worthwhile rabbit hole to dive down when mechanical gearing is bulletproof and may not be any more work to implement?

    I recognize the elegance of "doing it digital" but is it also "doing it the hard way"?

    After all, Barber Coleman and many others have solved this problem and solved it very well a long long time ago.
    If the goal is the best quality gears rather than the coolest gear cutting implementation, maybe this whole electronic gearing thing is worth a re-evaluation.

    Cheers

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


  2. Likes awake liked this post
  3. #22
    Join Date
    Dec 2003
    Country
    UNITED STATES
    State/Province
    Washington
    Posts
    4,627
    Post Thanks / Like
    Likes (Given)
    2129
    Likes (Received)
    3918

    Default

    Quote Originally Posted by drcoelho View Post
    Hm. I'm no expert with gear hobbing but I do know a thing or two about programming, and given sufficient money applied to the problem it seems to me that I can find a fast enough PLC with some predictive software programming (my code) that should get the spindle and servo synchronized. After all, CNC machines do this all the time. Not saying it is trivial, just saying I'm pretty sure I can write the code.....

    Also, I had already envisioned having the programming doing some averaging of spindle speed values such that when the spindle settles down to a more or less fixed speed, I would then get the rotary spinning in a perfectly matched speed. The idea to get away from the latencies associated with reading from spindle then feeding speed to the rotary...if the spindle is at constant speed, one could just tune the rotary servo to be the perfect matching speed.
    Ummmmmm . . . Having done high speed registration with a broad range of controllers including PLCs, I would guess that if you have not done this before that you will spend far more time and money trying than someone knowledgeable in the art can accomplish in an hours time using standard off-the-shelf hardware designed to do this that is far cheaper, requiring no “custom code”, and producing far more accurate results.

    This application would be a walk in the park using this integrated controller/drive and any motor you pick . . . Delta Tau Data Systems, Inc., Motion Control Solutions

    Alternately you could use a Siemens S7-1500T + S210 drive and motor, or a Compact Logix + Kinetix Servo Drive . . . these are true PLC options with built in registration capabilities, but you need to stick with their proprietary servo systems. You could use a Siemens PM240 drive with the appropriate interface card to Drive CLiQ from the 1500T PLC and likely commutation and control a third party motor.

    The Delta Tau solution will be the most flexible and likely easiest to learn, program and optimize.

  4. Likes David_M, drcoelho, Milland, awake, greggv and 3 others liked this post
  5. #23
    Join Date
    May 2011
    Country
    UNITED STATES
    State/Province
    New Jersey
    Posts
    1,842
    Post Thanks / Like
    Likes (Given)
    167
    Likes (Received)
    1003

    Default

    Quote Originally Posted by motion guru View Post
    Ummmmmm . . . Having done high speed registration with a broad range of controllers including PLCs, I would guess that if you have not done this before that you will spend far more time and money trying than someone knowledgeable in the art can accomplish in an hours time using standard off-the-shelf hardware designed to do this that is far cheaper, requiring no “custom code”, and producing far more accurate results.

    This application would be a walk in the park using this integrated controller/drive and any motor you pick . . . Delta Tau Data Systems, Inc., Motion Control Solutions

    Alternately you could use a Siemens S7-1500T + S210 drive and motor, or a Compact Logix + Kinetix Servo Drive . . . these are true PLC options with built in registration capabilities, but you need to stick with their proprietary servo systems. You could use a Siemens PM240 drive with the appropriate interface card to Drive CLiQ from the 1500T PLC and likely commutation and control a third party motor.

    The Delta Tau solution will be the most flexible and likely easiest to learn, program and optimize.
    Something tells me this guy probably has money to burn lol.

  6. #24
    Join Date
    Jun 2002
    Country
    CANADA
    State/Province
    British Columbia
    Posts
    2,492
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    1847

    Default

    Hi motion guru:
    A quick simple question from someone who knows almost nothing about this subject:

    In this scenario do BOTH motors need to be servo motors or can you realistically make a single servo motor follow the erratic behavior of an ordinary 3 phase motor driven spindle running at several hundred RPM if the synchronization needs to be as close as a mechanical coupling would produce?

    Cheers

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

  7. #25
    Join Date
    Dec 2003
    Country
    UNITED STATES
    State/Province
    Washington
    Posts
    4,627
    Post Thanks / Like
    Likes (Given)
    2129
    Likes (Received)
    3918

    Default

    Quote Originally Posted by implmex View Post
    Hi motion guru:
    A quick simple question from someone who knows almost nothing about this subject:

    In this scenario do BOTH motors need to be servo motors or can you realistically make a single servo motor follow the erratic behavior of an ordinary 3 phase motor driven spindle running at several hundred RPM if the synchronization needs to be as close as a mechanical coupling would produce?

    Cheers

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

    The latter . . . we often synch to an encoder over which we have no control. If the spindle encoder is not directly tied to the spindle (I.e. it is on the motor, and the spindle is geared to the motor) . . . it is a simple matter of adding a once per revolution prox switch of sufficient bandwidth to the spindle to capture the encoder position every rev (done with hardware latch of encoder counter / NOT in software) and then calculate the gear ratio on the fly and use the ratio in your path planning on the slave servo axis. And in this case the slave axis with the drive I linked can be a 3-phase induction motor and perform as well or better than a permanent magnet servo motor.

  8. #26
    Join Date
    Apr 2018
    Country
    UNITED KINGDOM
    Posts
    1,694
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    780

    Default

    Quote Originally Posted by drcoelho View Post
    I just want my gears to be very high quality for my own purposes,
    Then skip this idea and go buy a nice little used hobber. Your plan will cost you way more money and produce worse parts. In fact, if you want really accurate small gears, maybe even look for a little threaded-wheel grinder. I had a David Brown that would do 6" fine-pitch gears from solid, something like that would be ideal (for what you described).

    Home-made hobber, pffffft.

    @Marcus : I asked this same question at an AGMA show once, way back when. Controls were $40,000 then and it didn't make sense to me when there were already good mechanical machines and no one who can afford a $250,000 hobber uses hobbers for quick changeover uses, like a cnc mill would be.

    Most of the gear machine companies had no good answer but the Liebherr guys knew right away. It's not that electronic motor control is any better but by going electronic you lose the windup in connecting shafts, the backlash in the gear trains, and you remove a bunch of holes in the casting so that, done right, an electronic gear hobber is more rigid and accurate.

    But (no offense intended) I am pretty sure our OP is not Liebherr.

  9. Likes Oldwrench liked this post
  10. #27
    Join Date
    Dec 2003
    Country
    UNITED STATES
    State/Province
    Washington
    Posts
    4,627
    Post Thanks / Like
    Likes (Given)
    2129
    Likes (Received)
    3918

    Default

    And once you have a high quality gear and need to inspect it (and the gear weighs 15 tons and is of herringbone design) . . . A direct drive 360 pole motor with optical encoder 1m in diameter and the right controller and drive allow you to measure absolute gear tooth accuracy to +/- 1 arc second (while actively compensating for deviations due to lube film thickness in the hydro-dynamic bearings supporting the gear)

  11. Likes johansen, Zahnrad Kopf liked this post
  12. #28
    Join Date
    Jun 2002
    Country
    CANADA
    State/Province
    British Columbia
    Posts
    2,492
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    1847

    Default

    Hi motion guru and EmanuelGoldstein:
    That is super interesting information; I was completely unaware that you could use an encoder to interrogate irregular motion so closely at those kinds of processing speeds, and I was also unaware of the errors in mechanical drives and where they come from.

    I assumed you'd need a prettty sophisticated setup to process irregular frequency encoder signals at that speed and keep a slave motor in synch to within fractions of a degree, even when the slave is not a servo motor.
    To me that's pretty amazing and it sounds like this is all much less of a problem than I believed.
    Thank you both for enlightening me...I learned something new and cool today.

    Cheers

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

  13. Likes motion guru, Milland, awake liked this post
  14. #29
    Join Date
    Dec 2003
    Country
    UNITED STATES
    State/Province
    Washington
    Posts
    4,627
    Post Thanks / Like
    Likes (Given)
    2129
    Likes (Received)
    3918

    Default

    We recently retrofitted a mechanical “flying saw” on a plywood line. It had previously been linked to the plywood lug chain drive system and was dragged back and forth using a roller chain with a link attachment.

    We retrofitted with a motorized carriage with 10 feet of travel and put an encoder on the lug chain. The lug chain drive was extremely jerky and we processed the master encoder signal through a digital filter that clipped the peaks and added the clipped encoder counts to the signal valleys so as so not lose encoder counts.

    The system worked extremely well and allowed faster production rates, adjustable cut lengths, and far less wear and tear on the lug chain drive and saw carriage.

    We have done similar flying saws for I-Joist lines with saws weighing up to 2500 lbs and line speeds to 500fpm cutting 40ft lengths . . . Carriage travel in these cases is 15 to 20 feet. That’s a cut cycle every 4.8 seconds hauling a 2500 lb saw from zero speed to 500fpm in under a second, synching up with the joist, cutting it, decelling back to a stop, and then racing home to do it all over again . . . All the while maintaining +/- 0.125 inch cut accuracy. Crashes are exciting when something goes wrong.

  15. Likes memphisjed liked this post
  16. #30
    Join Date
    Feb 2017
    Country
    UNITED STATES
    State/Province
    California
    Posts
    380
    Post Thanks / Like
    Likes (Given)
    6
    Likes (Received)
    105

    Default

    Quote Originally Posted by motion guru View Post
    Ummmmmm . . . Having done high speed registration with a broad range of controllers including PLCs, I would guess that if you have not done this before that you will spend far more time and money trying than someone knowledgeable in the art can accomplish in an hours time using standard off-the-shelf hardware designed to do this that is far cheaper, requiring no “custom code”, and producing far more accurate results.

    This application would be a walk in the park using this integrated controller/drive and any motor you pick . . . Delta Tau Data Systems, Inc., Motion Control Solutions

    Alternately you could use a Siemens S7-1500T + S210 drive and motor, or a Compact Logix + Kinetix Servo Drive . . . these are true PLC options with built in registration capabilities, but you need to stick with their proprietary servo systems. You could use a Siemens PM240 drive with the appropriate interface card to Drive CLiQ from the 1500T PLC and likely commutation and control a third party motor.

    The Delta Tau solution will be the most flexible and likely easiest to learn, program and optimize.
    VERY interesting suggestions here, the Delta Tau Data Systems looks very interesting for my application, very high resolution sampling is possible on the spindle encoder which pretty much solves any concerns related to latency, and they are driving the servo motor directly...pretty cool! Programming this guy looks straight forward.

    I had already started researching the PLC/servo motor package approach as well, but thx for the additional references.

    VERY HELPFUL, thanks!

  17. #31
    Join Date
    Feb 2017
    Country
    UNITED STATES
    State/Province
    California
    Posts
    380
    Post Thanks / Like
    Likes (Given)
    6
    Likes (Received)
    105

    Default

    Quote Originally Posted by motion guru View Post
    Hi Marcus,

    The latter . . . we often synch to an encoder over which we have no control. If the spindle encoder is not directly tied to the spindle (I.e. it is on the motor, and the spindle is geared to the motor) . . . it is a simple matter of adding a once per revolution prox switch of sufficient bandwidth to the spindle to capture the encoder position every rev (done with hardware latch of encoder counter / NOT in software) and then calculate the gear ratio on the fly and use the ratio in your path planning on the slave servo axis. And in this case the slave axis with the drive I linked can be a 3-phase induction motor and perform as well or better than a permanent magnet servo motor.
    I'll be putting the encoder directly on the horizontal arbor of the hobbing cutter, thus will have direct readings of the actual cutter speeds.

  18. #32
    Join Date
    Feb 2017
    Country
    UNITED STATES
    State/Province
    California
    Posts
    380
    Post Thanks / Like
    Likes (Given)
    6
    Likes (Received)
    105

    Default

    Quote Originally Posted by motion guru View Post
    And once you have a high quality gear and need to inspect it (and the gear weighs 15 tons and is of herringbone design) . . . A direct drive 360 pole motor with optical encoder 1m in diameter and the right controller and drive allow you to measure absolute gear tooth accuracy to +/- 1 arc second (while actively compensating for deviations due to lube film thickness in the hydro-dynamic bearings supporting the gear)
    motion guru: VERY cool work you are doing

  19. #33
    Join Date
    Jul 2006
    Location
    Hillsboro, New Hampshire
    Posts
    7,003
    Post Thanks / Like
    Likes (Given)
    1649
    Likes (Received)
    4823

    Default

    I LOVE threads like this. So much interesting information comes out of them, especially when really experienced people like MG chime in. Thanks, all, following this with great interest.

    drcoelho, whatever you wind up doing, it would be great to maintain a thread of the actual build and programming, I'm sure it would be followed by many members.

  20. #34
    Join Date
    Aug 2006
    Location
    New Zealand
    Posts
    1,238
    Post Thanks / Like
    Likes (Given)
    70
    Likes (Received)
    246

    Default

    Hi
    The answer to the question of servos or steppers is dependent on the application. Steppers are usually preferred for slow speeds and stop/hold. The are not good for high rotational speeds. Servos can be used for any speed but perform better than steppers at higher rpm.


    Dan Gelbert built a lathe grinder accurate to microns using steppers (you can tell by the noise they make). Accuracy is not determined by the choice of stepper/servo.


    Andy Pugh did a retrofitted gear cutter with LinuxCNC. I think an electronic solution is a far simpler project than a wholly mechanical setup, but that might be because my background includes control engineering.


    Closed loop control is a very good thing that can be applied to steppers or servos. You will need encoders.

    Servo systems need to be designed and tuned to achieve stability.

    Backlash is highly non-linear and destabilizing. You should aim to minimize backlash.


    Dazz

  21. #35
    Join Date
    Feb 2017
    Country
    UNITED STATES
    State/Province
    California
    Posts
    380
    Post Thanks / Like
    Likes (Given)
    6
    Likes (Received)
    105

    Default

    Quote Originally Posted by Milland View Post
    I LOVE threads like this. So much interesting information comes out of them, especially when really experienced people like MG chime in. Thanks, all, following this with great interest.

    drcoelho, whatever you wind up doing, it would be great to maintain a thread of the actual build and programming, I'm sure it would be followed by many members.
    This project is most definately happening, and I will post my results to the forum.

  22. Likes Karl_T, Milland, BugRobotics liked this post
  23. #36
    Join Date
    Jan 2007
    Location
    Flushing/Flint, Michigan
    Posts
    7,505
    Post Thanks / Like
    Likes (Given)
    371
    Likes (Received)
    6280

    Default

    Quote Originally Posted by plastikdreams View Post
    Something tells me this guy probably has money to burn lol.
    If you put any real value on your time and effort the Delta Tau is maybe the lowest cost approach.
    There is a reason precision gear hobbing machines stayed mechanical for quite a while after nc/cnc and plcs where being put on machine tools.
    You don't just close the loop and sync. You have to do it exactly every time.
    The controls work in the digital or time sliced domain and you don't know what happens or have control between those time slices.
    Now the controls are fast enough to outpace the response of the motor/amp or mechanical ability to move. Now one has to worry about that end.
    The part wants to be cut analogue or continuous over and over.

    Today's PLCs will run circles around a 084 or PLC3 but this is still a hard task for many systems.
    It is a system of all the combined components. Blow one and the results are not what you want.
    One sees this in such automation and then there is the "it mostly works" followed by huge amounts of time in debug in that last 10%.

    Yet you build to what seems capable, test and go forward. Get your butt kicked and try again. That's how machines are built.
    Bob

  24. Likes Oldwrench liked this post
  25. #37
    Join Date
    Feb 2017
    Country
    UNITED STATES
    State/Province
    California
    Posts
    380
    Post Thanks / Like
    Likes (Given)
    6
    Likes (Received)
    105

    Default

    This to me is an interesting project, and matches my skill set. I'm seeing encoders that can do 1 million per revolution resolution so it seems to me that latency is pretty minimal as long as PLC can keep up. I also envision having a settling in period where I set the manual spindle speed, let the PLC sample it and sync the rotary axis and then start the cut once things are calibrated. The questions to be answered are whether the PLC can keep up with variations in spindle speed during the cuts, and whether the rotary server can be adjusted in real-time during cuts. While one would think that mechanical linkages would have no latency, in fact given back-lash and gearing gaps and just the transmission time along the mechanical drive train, I suspect that old style mechanical systems also have latency in terms of small changes in spindle speed, and as previously stated, my goal is to match old style gear hobber quality, not hit todays CNC standards, so maybe that is an achievable goal.

    And yes, iteration is the key to a project like this. Try it, measure it, improve it, etc....

  26. #38
    Join Date
    Feb 2017
    Country
    UNITED STATES
    State/Province
    California
    Posts
    380
    Post Thanks / Like
    Likes (Given)
    6
    Likes (Received)
    105

    Default

    Quote Originally Posted by CarbideBob View Post
    If you put any real value on your time and effort the Delta Tau is maybe the lowest cost approach.
    Bob
    If i went with the Delta Tau power controller, what motor would you recommend? I'm inclined to go with the Sankyo RCD rotary table...they specifically reference servo motors from FANUC, Mitsubishi, Sanyo, OKUMA in their product catalog. I've talked to Sankyo and they would need to factory set the unit with lower gearing for lower RPMs that I'll be using, but they confirm that this rotary table will work in continuous mode use for my application.

    Also, any feedback regarding what gearing hob RPMs are typical would be helpful....maybe someone that has used the old style gear hobbers could provide some insight into what range of RPMs are typical for a range of spur gear cuts?

  27. #39
    Join Date
    Nov 2004
    Location
    Webster Groves, MO
    Posts
    7,113
    Post Thanks / Like
    Likes (Given)
    1780
    Likes (Received)
    3214

    Default

    One of my friends was a digital guy while I am more analog. He liked to point out that an analog loop, no matter how tight, must have an error to function.

    Bill

  28. #40
    Join Date
    Jan 2010
    Location
    Canada
    Posts
    322
    Post Thanks / Like
    Likes (Given)
    61
    Likes (Received)
    164

    Default

    I think I would go with the stepper and an encoder driven by the spindle.

    Like the electronic lead screw on a lathe.

    A bit something like in this video, forget the lathe size, please.

    YouTube


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
  •