Servo or Stepper for CNC? - Page 2


Page 2 of 4 FirstFirst 1234 LastLast
Results 21 to 40 of 62

Thread: Servo or Stepper for CNC?

  1. #21
    Registered walter's Avatar
    Join Date
    May 2003
    Location
    USA
    Posts
    792
    Downloads
    0
    Uploads
    0

    Default

    Please continue Mariss, your effort is much appreciated.



  2. #22
    Gold Member
    Join Date
    Mar 2003
    Location
    United States
    Posts
    2839
    Downloads
    0
    Uploads
    0

    Default

    Sorry for the lapse; I've been up to my neck in alligators while draining a swamp.:-) I've bookmarked the thread; give a few days and I'll pick it up again after I've turned a few of these nice alligator hides I have now into shoes or belts or something...

    Mariss



  3. #23
    Member
    Join Date
    Aug 2006
    Location
    UK
    Posts
    1602
    Downloads
    0
    Uploads
    0

    Default

    How does the use of gears/timing belts effect the overall accuracy of the system?

    Say I wanted 1Nm at 1000 rpm, I could do it with a 4:1 reduction on a 4,000 rpm, 0.25Nm servo, or by gearing up a 4Nm stepper running at 250rpm by 1:4.

    I'm guessing that the geared down servo will have more positional accuracy due the the gearing effectively quartering the step size, wheras the geared up stepper will effectively quadruple its step size, but won't both systems be a lot less accurate due to backlash in the gearing?

    I guess what I'm asking is does it actually make sense to use ballscrews if you're not going to drive them directly from the motor?



  4. #24
    Registered
    Join Date
    Dec 2005
    Location
    USA
    Posts
    3319
    Downloads
    0
    Uploads
    0

    Default

    In our situation (2:1 ratio belt driven preloaded ball screw, gibb based Bridgeport Eztrak mill with with high $$ true ball screw support bearings) we found the "error" contributed by the OEM "gilmer" style square tooted timing belt to be inconsequential.

    Granted, we spent a small fortune on the ball screw bearings and weeks tuning and adjusting the machine to remove/eliminate slop. During the tuning process, we also spent a ton of time playing with belt tension and the like.

    A belt in good condition had virtually NO effect on net accuracy once we got the slop out of the rest of the system. The BIGGESt contribution came when we had a pro tune the servos so as to get the follower error of the servo's "tuned out".

    We tested via cutting perfectly round circles via a canned circle milling program. We then checked the form for roundness on our camshaft measuring machine (measures to 0.0000025" resolution).

    When the machine was properly tuned, the circles were round within 0.0003" or better and the "flats" at the 4 direction changes were 0.0001" or better - the "flats' are nearly IMPOSSIBLE to eliminate due to gibb slop needed to allow the table to move without binding. We ultimately learned that the above "errors" were due to other hysterisis issues that simply weren't practical to resolve on a gibbs based "bridgeport mill".

    Does the average guy need such accuracy? No.

    But we CONCLUSIVELY showed that "belt induced errors" are much less in reality than what one might "expect" IF you have a well adjusted/tuned system made of parts with decent integrity.

    Worn/tired crap isn't worth spending the time on. Get good sprockets and premium belts, properly tension them and make sure they "wrap properly" and spend your time and money on other stuff that provides a better cost/benefit ratio - like a good servo/stepper tuning regime and the best damn ball screws ans ball screw bearings that you can hardly afford.



  5. #25
    Member
    Join Date
    Aug 2006
    Location
    UK
    Posts
    1602
    Downloads
    0
    Uploads
    0

    Default

    Thanks NC Cams - that's very useful

    I am trying to choose motors at the moment - I have calculated my cutting forces to be about 200N, and with the screws I am planning to use, this would need about 0.2-0.3Nm/30-40ozin of torque at about 500-1000 rpm.

    Looking at the motors I could easily lay my hands on in the UK, I seem to have the choice of 2-pole steppers, 3-pole/phase steppers and brushless DC servos.

    If I use 2-pole steppers, it looks like I'd have to use huge motors just to get any torque at 500-1000 rpm, and perhaps some sort of gearing up. This doesn't sound good for acceleration as big motors have lots of inertia.

    If I use brushless DC servos, it looks like I'd probably have to gear them down as well as the torque would be in the right range, but I'd be wasting power as I'd never hit the full speed of 4000+ rpm. I'd also have to either sell a kidney, or build my own drivers as brushless DC and cheap don't seem to go hand in hand.

    Finally, 3-phase steppers seem to provide flat torque across the 500-1000 rpm range, but no extra torque at low speeds. I could probably use them without gearing, but I am a bit worried that no one seems to use them!

    Hmm...



  6. #26
    Registered
    Join Date
    Jan 2007
    Location
    MI. USA
    Posts
    210
    Downloads
    0
    Uploads
    0

    Talking

    Servos or Steppers? --- Hmmm --- Coke or Pepsi?

    A couple of notes from my experiences;

    Steppers are cheaper.

    Steppers are so much easier to get up and running Four to six wires to hookup, apply power and go.
    Correctly sized and tuned a servo will run circles around a stepper.

    You can mismatch motor size with steppers and the thing will still run.
    Mismatch servo motor size to loads and tuning will be a nightmare.

    Stepper systems come in basically 2 flavors full/half step and microstep.
    There is about a zillion different combinations in servo systems ( DC brush, brushless, AC, torque mode, vel mode, encoder feedback, resolver feedback …… )

    Stepper motors go where you tell them to.
    Servo motors are error driven devices. No position ( or velocity) error no motion. It’s impossible servo system to maintain better than +/-5 counts over a range of speeds. Figure on +/- 30 encoder counts for an average well tuned system. Many commercial CNC machine tools run +/-100 counts or more.

    If all this accel/decel, steps per second, and inertia matching stuff is confusing just wait until you get to figure out how this Proportional , Integral, Derivative(PID), stuff works.

    Some myths:
    Stepper motors lose steps: If running within its torque/speed capabilities a stepper will never miss a step.
    Servo motors overshoot: This is all in the tuning. You can tune a servo to overshoot or not. Sometimes a little overshoot is desirable to compensate for mechanical lag (screw torsional wrapup, slide stickslip, bearing drag, etc.).

    I started out with steppers and looking back I’m glad I did. I learned about basic motion control without having to worry about having to get the motors to move. Now I use only servos but its been a long (and mostly fun) learning experience. If you have some basic motion control experience I would recommend servos. The performance benefits are great and its always good to learn new things. I once read that tuning servos was like riding a bike, hard to explain but once you master it you never forget. They forgot to mention that you’re going to crash a lot while learning. Limit switches are a must on servo systems, eventually the thing is go to run away on you and take off at the highest speed it can. Mount them well inboard during testing and remember servos coast to a stop without power so allow plenty of room for overtravel. In addition to the motors and controller your going to need some kind of tuning software so you can see the response of the motors as you tune. DC brush torque mode servos are the easiest to tune. DC brushless velocity mode are the worst. AC servos fall in between. Velocity mode systems (tach feedback) have 2 loops to tune, one in the drive and one in the controller.

    Anyone using the packaged Sanyo-Denki motor drives from AutomationDirect and MotionOnline. About $1000.00 per axis, AC servo systems which can be driven with stepper signals (giving you and encoder feedback ultra-high speed stepper) or servo signals. Automatic tuning built into the drive and cables pre-made. About as close to plug and play as you get in the servo world. 100 watts in the size of a cigarette pack and 400 watts in about the size of a NEMA-23 (all metric sized though).

    You can always spot the pioneers -- They're the ones with the arrows in their backs.


  7. #27
    Member
    Join Date
    Aug 2006
    Location
    UK
    Posts
    1602
    Downloads
    0
    Uploads
    0

    Default

    Thanks CarbideBob - I've currently got a stepper system, which works well, but is too slow for what I've got planned next. I would happily use steppers again if I could get 1Nm/140oz in at 1000-2000 rpm, but this seems just too fast unless I take massive steppers and gear them up, thus losing resolution...



  8. #28
    Member
    Join Date
    Oct 2005
    Location
    USA
    Posts
    14
    Downloads
    0
    Uploads
    0

    Default

    Mariss:

    Another vote for you to continue sharing your servo sizing wisdom.



  9. #29
    Registered
    Join Date
    Dec 2005
    Location
    USA
    Posts
    3319
    Downloads
    0
    Uploads
    0

    Default

    A small caveat to consider along with CARBIDEBOB"s reply about steppers:

    I know of cases where steppers are/were used to make a CNC OD grinder that doubled as as profile shape grinder.

    Yes, they were geared down via belts and/or ball screws,

    I do NOT know if they used proper ball screw support bearings (see my prior "Extrak" posts on this topic (real critical for this application)

    I do NOT know the resolution that was designed into the system (more critical details - the guy who designed it claimed to be an expert).

    BUT, I can definitely say that "steps/facets" were/are a problem in the finished part. The builders were NOT able to tune or gear the problem out of the device.

    Seems that the "steps" of the stepper would superimpose themselves onto the position of the grinder wheel and visible "facets" (micro steps/flats)were unavoidably created/cut in the parts by the stepped motion of the system. The thing was especially noticeable on 50-55mm dia bearing journals that we looked at.

    The same problem was eliminated with the incorporation of servo's and life rolled merrily along.

    One addendum to servo tuning: the faster you try to move with servo's, the more/greater the potential follower error, especially with a "slow" PC. In our case, due to the accuracy requirement for the intended prime use, we had the technician NOT tune the servo's to run at light speed jogs.

    Also, high pole count DC motors are notoriously smoother running than AC motors due to them NOT having the 50-60hz current ripple affecting them like that which occurs with AC motors. We learned and verified this when I worked in the bearing industry with motor testing and re-verified it when we started doing cam and OD grinding most recently with our cam grinder. This is why you DON'T run bow string tight V belts when AC motors drive grinding spindles.

    Moreover, when we finish cut something that has to be REAL precise, testing on our machine showed that darn near perfection could be achieved if you ran REAL SLOW feed rates on the spring pass final cut (like maybe 1.5"/minute at 100SFM with HSS cutter on mild steel).

    The finish isn't "ground" smooth but when you run a tracer over it, with a follower like that which will be used on the finished part, you're hard pressed to tell the difference between the milled part and a CNC ground one.



  10. #30
    Gold Member
    Join Date
    Mar 2003
    Location
    United States
    Posts
    2839
    Downloads
    0
    Uploads
    0

    Default

    jamma, everyone,

    Thanks. My plate is kind of full right now because we are in the process of moving to bigger digs. Our curse is we cannot decommission our current shop until the new one is up and running. We can't shut off production for even one day without having all Hades to pay; we have product shipping obligations that cannot be interrupted.

    The result is drives are being turned out at both locations right now and they are 10 miles apart. That also means the last 3 weeks two complete surface-mount production lines are running at each location 10 miles apart. My 'day' job is here, my 'night' job is 10 miles away straightening out the daily hairballs entailed in commissioning a new facility and there are plenty. That is 12-hrs a day, 7 days a week job until the move is completed.

    It will take another 4 weeks for that. I will then sleep for 48 hours and be a human again. The topic is on my list of things to do, please bear with me until I can get to it. I've got to get this nightmare out of the way first.:-)

    Mariss



  11. #31
    Gold Member
    Join Date
    Mar 2003
    Location
    United States
    Posts
    2839
    Downloads
    0
    Uploads
    0

    Default

    NC Cams,

    Good points; let's look at them:

    1) A 10-microstep drive on a 1.8-degree motor takes 2,000 steps for one revolution. At low speeds, each step is a 0.18-degree increment of motion (move, stop, move again).

    2) A servomotor equipped with a 500-line encoder also takes 2,000 steps for one revolution. At low speeds the motion is also incremental (0.18-degrees) because the encoder is incremental. It also has a move, stop, move again motion.

    There is no difference between the two.

    Servos can have a continuous following error if the feedback is PD and a transient following error if the feedback is PID.

    Steppers always have a no-load following error of -1.8 degrees at moderate to high speeds because of inductive current phase lag (-90 electrical degrees). Add to this a zero to -1.8 degrees of load dependent following error and a stepper can be up to -3.6 degrees out of position at higher speeds. This is equivalent to -20 encoder count following error for a 500-line encoder.

    Step motors aren't really step motors. They are high pole-count (50-poles for a 1.8 degree motor) AC synchronous motors.

    Mariss



  12. #32
    Registered
    Join Date
    Mar 2007
    Location
    usa
    Posts
    84
    Downloads
    0
    Uploads
    0

    Default

    I read some where that larger stepper motors lose resolution. I'm not sure what this means, but the post suggest 600oz/in as the largest recommended stepper size. I'm thinking about getting some 900oz steppers from Kelinginc, but I'm not sure about this lost resolution issue. Can someone clarify? thanks.



  13. #33
    Registered Lt.Scott's Avatar
    Join Date
    Oct 2005
    Location
    USA
    Posts
    3
    Downloads
    0
    Uploads
    0

    Default

    Wow, Mariss has made my choice for me. I already have a Xylotex 3 axis board and steppers, but my son-in-law working at Rubbermaid in the maint. shop said he can get all kinds of "removed from service" servo motors. (I will also see about steppers) But with Mariss' comparisons early on in the thread, I will stick with my steppers because:: 1) Already have them 2) Don't need all that torque 3) Servos sound too complicated for me
    I am building a smallish sized router table based on jgro drawings.

    John

    We need not all think alike, but we should all THINK!


  14. #34
    Registered lgalla's Avatar
    Join Date
    Jul 2006
    Location
    canada
    Posts
    1256
    Downloads
    0
    Uploads
    0

    Default

    Geckos+Mack3+servos do not =closed loop operation.Mack3 is for steppers,geckos accept step+direction signals.Not the same thing as a full blown servo system.Which is better when using Mach3,stepper or servos?
    larry

    L GALILEO THE EPOXY SURFACE PLATE IS FLAT


  15. #35
    Registered
    Join Date
    Oct 2005
    Location
    United States
    Posts
    78
    Downloads
    0
    Uploads
    0

    Default

    Unless you are talking about a dedicated system, running its own hardware and software, you are simply not going to see your "closed loop sytem". Modern servo controllers and systems run their feedback loops at the hardware level- no loop feedback through software at all. To run global feedback in such a complex system would invite disaster, and is simply not feasible for a hobbyist.

    The only way to get a true closed loop system is to run a system based upon linear encoders. That way, the controller knows at any given moment exactly where the machine is. Mach 3 knows where the machine was commanded to go, and the servo drives themselves know whether or not the shaft of the motor responded. If the error builds up in the motor/ drive feedback loop beyond acceptable limits, the drive will fault, and shut down the system. Mach 3 will know this, and stop sending motion commands. As far as I am concerned, this IS closed loop operation, at an acceptable level.

    Why muddy the water by talking about global feedback? Global loop feedback is an impressively difficult thing to achieve, and nets no benefit for the average user. It is necessary on industrial machines that run unattended for hours or even days, but that is not what we are building here.

    Mach 3 + servos IS a closed loop system, in that the drives have local feedback (hence the term 'servo'), and the drives report fault conditions.

    Closed loop is at the HARDWARE layer, which allows us to build our machines with random parts!



  16. #36
    Member Halfnutz's Avatar
    Join Date
    Feb 2005
    Location
    USA
    Posts
    750
    Downloads
    0
    Uploads
    0

    Default

    Well put Corvus,

    I have Yaskawa Servo Drives, they have a fault output signal, they are all three (X,Y,Z) tied to an enable input to Mach. If Mach sends a command to X Axis to move 1500 steps and the for some reason the drive is unable to accomplish the command, measured in real time by by the encoder, the enable input to Mach is driven low and Mach shuts down in an error state. Closed loop at the hardware level.

    Very easy to implement, and 100% effective. The Servo Drive has parameters that allow the user to set all the error parameters and control how "tight" the system is.

    As far as I know this is how most modern closed loop systems are accomplished, irregardless of cost or complexity. It's just the best way for a multi-axis system to be implemented.

    Halfnutz


  17. #37
    Registered
    Join Date
    Mar 2009
    Location
    pakistan
    Posts
    2
    Downloads
    0
    Uploads
    0

    Default need help

    can any bro help me with information on steooer motor em152



  18. #38
    Registered
    Join Date
    Sep 2006
    Location
    uk
    Posts
    10
    Downloads
    0
    Uploads
    0

    Default

    hi all,

    by any chance any of you have or know of any websites that have comparison charts of various commercial available cnc controllers?

    could you also please explain to me, when choosing a cnc controller, what are the characteristics most machine builders look for?

    Thank you
    khalid



  19. #39
    Registered
    Join Date
    Feb 2007
    Location
    China
    Posts
    16
    Downloads
    0
    Uploads
    0

    Default www.cnc4cnc.cn

    Quote Originally Posted by pmac View Post
    hi all,

    by any chance any of you have or know of any websites that have comparison charts of various commercial available cnc controllers?

    could you also please explain to me, when choosing a cnc controller, what are the characteristics most machine builders look for?

    Thank you
    khalid
    Hi,

    Please visit website of www.cnc4cnc.cn to see if it's helpful for your project.

    Good luck!
    Kevin



  20. #40
    Registered
    Join Date
    Jul 2009
    Location
    united states
    Posts
    11
    Downloads
    0
    Uploads
    0

    Default

    hmm not that I am all that great at this stuff but from 90% of the post on this the only answer I can come up with is if I want speed I have to go servo but if I want accuracy and procession + consistency stepper is the only real way to go! steppers are cheaper, easy to set up, don't have to worry about over travail so much, or lost steps like a servo, well for me it sounds like I would down grade to stepper motors then use a servo set up any day of the week, the only use I see for servo set-up is for mass production do to speed loss when it comes to steppers but at that point I would rather build more stepper set-up machines first to make up the difference I would hate to spend $500 on a big chunk of aluminum have a servo over shoot and make it to where I have to fill it back in with a welder count me out I will stick with the steppers thank you very much!!!!

    thanks for the post learned a lot from it



Page 2 of 4 FirstFirst 1234 LastLast

Tags for this Thread

Posting Permissions

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


About CNCzone.com

    We are the largest and most active discussion forum for manufacturing industry. The site is 100% free to join and use, so join today!

Follow us on


Our Brands

Servo or Stepper for CNC?

Servo or Stepper for CNC?