584,860 active members*
5,163 visitors online*
Register for free
Login
IndustryArena Forum > CNC Electronics > Servo Motors / Drives > Ancient CNC lathe retrofit / fanuc Servos
Results 1 to 10 of 10
  1. #1
    Join Date
    Aug 2013
    Posts
    22

    Ancient CNC lathe retrofit / fanuc Servos

    Hi!

    Some while ago I bought several old Gutel G100 precision CNC lathes and slaughtered all but one of them which left me nice compact machine + enough of spares to last a lifetime of hobby machinist. The only real gripe with this machine is ancient (Fanuc 5M / Sinumeric) brain that I would like to replace with Mach3/4 driven electronics.

    The X-Y servos are Fanuc A06B-0613-B031 56v/12a motors with pulse encoder. Is there a reasonable way to drive those or would I be better of replacing them with new motors + drivers combo?

    Thanks
    Attached Thumbnails Attached Thumbnails fanuc_motor.jpg  

  2. #2
    Join Date
    Aug 2013
    Posts
    22

    Re: Ancient CNC lathe retrofit / fanuc Servos

    I was told that I can indeed control those drives using M4-L075K driver from CS Lab.

    I will bite the bullet and order entire control system from them, will post an update when successful or not.

  3. #3
    Join Date
    Oct 2010
    Posts
    1189

    Re: Ancient CNC lathe retrofit / fanuc Servos

    My learning from an Bridgeport interact 412 is get new drive kits it saves you months of headaches


    Gesendet von iPad mit Tapatalk

  4. #4
    Join Date
    Mar 2005
    Posts
    17

    Re: Ancient CNC lathe retrofit / fanuc Servos

    fdm_guy,
    Those appear to be [reasonably commodity] Fanuc DC servos - Perhaps Al can confirm, but the pulsecoder on that stack should be plain incremental, not FanucSerial (wrong vintage by my chart) -You can look up the standard Fanuc Pulsecoder pinout for the mating Amphenol connector - it's 5v direct drop in for a standard quadrature encoder. (Yay! Way easier that way). I would have liked to see the "2000P" suffix, to be sure, but not every nameplate stated it, IIRC.

    If they're incremental (likely), then we'll continue the conversation. If they're serial, then, it's a fun project but more worth your while to dump them.

    I had them on my Tsugami lathe of similar vintage, and could have dropped any old V/T or intelligent DC servo drive on to control them - I'm an LCNC guy, so had the advantage of not being limited to just step/dir for control (yes, Mach can now do fully closed servo loop, however you didn't mention your level of extra hardware capability, so as many Mach-er's do, they are limited to digital parallel port pins only, which does dictate step/dir and no encoder return). If you have the desire for additional hardware that can do servo-loop speed Velocity or Torque PWM generation, you could host the control loop inside of Mach. But that's just copying the 1970's (and earlier even) technology result. NO fun in that.
    Cheap DSP chips let a lot of manufacturers put that PID loop inside the drive now and even beat what a PC can do with it. Thus, even though I could do (and have done) the PID loop inside of LCNC, I typically choose to go with a stepper-like config. Truth be told, I use high resolution encoder follower modes, but they're still finite steps. I then choose drives that understand that. (I get the benefit of high-speed acceleration with command failsafe, but get rid of the nasty null-drift of an analog command). I lose on-the-fly acceleration rate changes, however even with high-speed-machining work, I have yet to fall out of the fixed acceleration-rate windows I have created.

    [The geeky talk: any control can have the capability to change the acceleration within it's defined window - let's say from 0 to max 200 pulses per second per second, but that window is predefined and that's all it gets - in a stepper system it's the max the drive and motor will respond to without losing steps, or less. If the control is in charge of a V/T loop servo, it can change the acceleration rate window to practically anything it wants on the fly - so if it had lookahead and HSM, in that mode it could ramp the acceleration to a slower window in some sections for high detail and high resolution, then for other sections "open up the taps" to fly around the workspace. This is acceleration rate, not feedrate, so don't look at it as a restriction on rapids or cutting feeds. It makes really cool sounds, but I haven't missed it any. Note that some stepper-drive controls will go even simpler and lock the acceleration to a single fixed value. Even then, on a short-stroke machine like a lathe, you may not notice any difference.]

    I'm going to get into trouble here for a minute, but bear with: just because the nameplate rating gives a "max" number doesn't mean you absolutely have to run at that number. (fire extinguishers and Firetrace(TM) at the ready...)

    I'm not a marketer for any brand in particular, but if you wanted to explore, you may find that an offering from Gecko could drop right in between Mach and that DC motor, and you're running. IIRC, Mariss now has his 320's in the sub-$150 range. That drive (and others like Leadshine's DCS series, or Jon's (Pico) PPMC system, etc) pretty much turn a DC servo into a stepper motor, as far as Mach is concerned. For instance, if you were to PM Mariss, he would gladly give a good and informed opinion on your circumstance. Sometimes someone saying "not a good fit" is just as important as "of course, we can do that". (However I'm pretty confident a 320x is a dream match for your motor.)

    That said, I'm not a DC servo guy. I tend to dump them at first sight for retrofits, unless there is a really big reason to keep them. I'm an AC servo guy, and have no problem with making bolt-pattern match plates to pull a DC servo off a mount and put an AC servo in its place. I also loop back feedback directly to my control of choice; most AC servo drives return either a duplicate of the encoder signal or a modification of it as a specially buffered output - rather handy than dealing with tapping off the real encoder and possibly injecting noise accidentally. My number one reason to dump DC servos is not having to deal with sourcing brushes to replace and taking the time to properly burnish them in once it's time for a replacement. Burnishing is easily done with the motor not attached to the load. Less fun when it's on a short stroke screw (read: lathe) and the PID loop no longer responds cleanly due to the motor having a different response profile (new brushes). Fair enough, it's not as bad as it sounds, lots of equip still uses DC mot's very reliably and will surely do so for a long time.

    If you wish to keep the DC servos, I'd recommend one of those drive choices (or their equivalents) above. The encoder connects to the drive, but that's where it stops. The drive takes care of the servo loop (adjusted either by trim pots or a serial-controlled interface of some sort) - and once tuned you leave the servo loop alone. Mach is set up to drive steppers (a pin each for enable, step and direction for each axis, plus whatever estop/limit and ancillary triggers you need) and you're off to the races. A big ol transformer/bridge/capacitor becomes your unregulated linear power supply and the lathe is likely moving in an afternoon. (Unloaded tuning and all good safety practices recommended, of course.)

    This is really just to say that you don't have to dump all the original parts - retrofitting to a different motor style would take more effort, but depending on your application, perhaps not enough return on that investment. Mach may or may not be suited to a production application for you, but if it's hobby to mid-entry, the combination can be a successful marriage. If that's the case, no need to put a supercharger on a yugo, unless you're looking for coolness factor.

    With an intelligent DC drive on your existing motor, you can also try just one axis to see if it is something you want to pursue at very little cost or investment - moving over to another motor or drive platform takes a bigger commitment. Running the Z (axis parallel with the spindle) back and forth would let you get the hang of what remains in those motors; maybe they no longer perform optimally on that chassis (hoping not the case), but if you need to unload them, I'd recommend unloading with the couplers as a set; if they're the same Fan5T vintage, the motor shafts were taper-stud and likely coupled with a zero disk coupler - that combo is typically something folks would pay for rather than just the motor, and in my case, it provided funding to make AC drives that much easier to acquire.

    Cheers, best of luck!

    Ted.


    Quote Originally Posted by fdm_guy View Post
    Hi!

    Some while ago I bought several old Gutel G100 precision CNC lathes and slaughtered all but one of them which left me nice compact machine + enough of spares to last a lifetime of hobby machinist. The only real gripe with this machine is ancient (Fanuc 5M / Sinumeric) brain that I would like to replace with Mach3/4 driven electronics.

    The X-Y servos are Fanuc A06B-0613-B031 56v/12a motors with pulse encoder. Is there a reasonable way to drive those or would I be better of replacing them with new motors + drivers combo?

    Thanks

  5. #5
    Join Date
    Dec 2003
    Posts
    24216

    Re: Ancient CNC lathe retrofit / fanuc Servos

    According to my Fanuc motor spec sheets those motors have 2000p/rev (B031) Absolute encoders, not incremental.
    The actual encoder # would confirm it.
    Otherwise nice motors, if the com is still OK and maybe just need new brushes.
    Al.
    CNC, Mechatronics Integration and Custom Machine Design

    “Logic will get you from A to B. Imagination will take you everywhere.”
    Albert E.

  6. #6
    Join Date
    Oct 2006
    Posts
    16

    Re: Ancient CNC lathe retrofit / fanuc Servos

    The 5 control would have used incremental encoders as they are most likely yellow caps. The issue I have had with using these old motors with new drives is that there may not be a analog tach which means you would have to adapt them to the motors. I would replace them if the budget allows

  7. #7
    Join Date
    Dec 2003
    Posts
    24216

    Re: Ancient CNC lathe retrofit / fanuc Servos

    I either remove the tach brushes or disable the tach as they are not needed on modern drives as the preferred way with later drives is to use torque mode rather than the old velocity command modes.
    As I say the Fanuc motor sheet shows B031 as absolute 2000p.
    Al.
    CNC, Mechatronics Integration and Custom Machine Design

    “Logic will get you from A to B. Imagination will take you everywhere.”
    Albert E.

  8. #8
    Join Date
    Oct 2006
    Posts
    16

    Re: Ancient CNC lathe retrofit / fanuc Servos

    This is true but I have never seen a Fanuc 5 series control using absolute encoders or serial encoders as the mother board uses the standard am26ls line receivers for quadrature input. There may be something that they used and called the that but but I suspect they are 2000 line pulse coders . Would like to see the sheet you have as you learn something new every day And sinumeric did have some odd options for the Fanuc brand.

  9. #9
    Join Date
    Dec 2003
    Posts
    24216

    Re: Ancient CNC lathe retrofit / fanuc Servos

    This is one of them.
    Al.
    CNC, Mechatronics Integration and Custom Machine Design

    “Logic will get you from A to B. Imagination will take you everywhere.”
    Albert E.

  10. #10
    Join Date
    Aug 2013
    Posts
    22

    Re: Ancient CNC lathe retrofit / fanuc Servos

    Hi everyone!

    Thank you very much for input, it is greatly appreciated and helpful! I will keep you all updated on my progress and write a blog / transformation log as well, so it could help someone in future. I understand that replacing the motors with a nice and modern driver / motor pair is probably the most sane thing to do, it will cost me less in $ (I have 6 of those old servos I could sell off) and guaranteed less in time.

    But, this is a hobby retrofit, I do not care about cost return, there will be no $ return and there was already lots of costs / time spent, for example my machine was filthier then any girl I ever met... She's sparkling clean now... In some way it would be cool to only replace the 'brain' and keep drives in the machine.

    I opened one of the servos and the encoder type nr. is: A860-0300-T001 2000P. Seems to be quite well known piece, there is pin out posted here by Al_The_Man: http://www.cnczone.com/forums/fanuc/177163-cnc.html

    What do you think? Am I opening a box of pain here by trying to get them working with modern controllers or is it doable?

    Best Regards

    Fdm Guy

    P.s. I am a bit more scared by main spindle driver here: http://www.cnczone.com/forums/spindl...ml#post1986644

Similar Threads

  1. Ancient CNC lathe retrofit / fanuc Servos
    By fdm_guy in forum EMCO CNC Machines
    Replies: 0
    Last Post: 01-03-2017, 08:26 AM
  2. Handyman cnc retrofit with dmm servos
    By ira7481 in forum Bridgeport / Hardinge Mills
    Replies: 1
    Last Post: 03-28-2014, 10:19 PM
  3. Fanuc 5T Lathe retrofit
    By NC Cams in forum Fanuc
    Replies: 3
    Last Post: 03-19-2010, 09:10 PM
  4. Ancient Knee mill retrofit
    By kong in forum Knee Vertical Mills
    Replies: 38
    Last Post: 11-10-2007, 09:06 PM
  5. plasma table retrofit: smart servos or closed loop steppers?
    By Josh Leber in forum Waterjet General Topics
    Replies: 12
    Last Post: 03-25-2006, 04:25 AM

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
  •