Considering LinuxCNC, what limits rapid speeds? - Page 2
Close
Login to Your Account
Page 2 of 2 FirstFirst 12
Results 21 to 35 of 35
  1. #21
    Join Date
    Jul 2006
    Location
    Hillsboro, New Hampshire
    Posts
    3,920
    Post Thanks / Like
    Likes (Given)
    798
    Likes (Received)
    2558

    Default

    Quote Originally Posted by huleo View Post
    We have a couple older machines that are worth recontrolling.
    I'd like to stop right here and find out what are these wonderful machines? If they're specialty designs it might be worthwhile, but if they're smaller plain 'ol VMC's then it's almost a certainty that you'd be better off buying a new machine with the capabilities you want over retrofitting an old Fadal or Tree.

    So what are they? It'll help the discussion...

  2. #22
    Join Date
    Jan 2007
    Location
    Flushing/Flint, Michigan
    Posts
    6,079
    Post Thanks / Like
    Likes (Given)
    238
    Likes (Received)
    4892

    Default

    The controller in no way limits machine acceleration or rapids.
    You can run 2000 ipm rapids and acc/decs near 1 G on a Sinclair ZX81 running interpreted basic.
    It is screws, their mount style, motors, motor to screw mechanics, system reflected inertia and the servo drives that determine this side not the control.
    The control and it's sample time control accuracy during profiling or feed moves.
    This limits the highest speed you can do a G01, G02 etc at.
    During rapid moves the error term is normally ignored or opened way up.

    Upgrading a control will not get you higher rapids as the old one is already asking for 100% out of the drive system and basically slamming it aginst the wall.
    Upgrading motors and amps may buy you some acc/dec but likely the rapid will still be limited by the mechanical side.
    Screws can only turn so fast before you hit critical speeds, belted drives have distinct limits.
    It is a system only as strong or fast as the weakest link.

    People ask me to put new high speed AC drives and high speed controller on a older machine expecting a huge speed up. I have to tell them that it will not be faster without replacing much more.
    I'm sure Motion gets the same type of requests.....

    Just trash the drive system, install linear motors and 1/10 micron feedback. How hard can it be?
    Bob

  3. #23
    Join Date
    Dec 2007
    Location
    Southeastern US
    Posts
    6,079
    Post Thanks / Like
    Likes (Given)
    652
    Likes (Received)
    2403

    Default

    Quote Originally Posted by huleo View Post
    Well, lets assume 3 of us work blind folded so it takes us a whole year to control one machine.

    Lets then relook at things. How much is a Siemens 840d kit? Assuming all we want is the control unit and software?
    Last I checked, I think we are paying in the neighborhood of $55-$60k for the entire electrical control system package. This includes the 840Dsl, servos, motors, and all other electrical devices (electrical hardware, new cabinet back plane, switches, door switches, safety relays, sensors, PB's, wiring, terminals, breakers, contactors, wire ends, wire tray, panduit, etc) on the machine for our in-house retrofits (5 axis).

  4. #24
    Join Date
    Sep 2009
    Location
    barcelona, spain
    Posts
    2,062
    Post Thanks / Like
    Likes (Given)
    409
    Likes (Received)
    1123

    Default

    Seems to make sense.

    Quote Originally Posted by Tonytn36 View Post
    Last I checked, I think we are paying in the neighborhood of $55-$60k for the entire electrical control system package. This includes the 840Dsl, servos, motors, and all other electrical devices (electrical hardware, new cabinet back plane, switches, door switches, safety relays, sensors, PB's, wiring, terminals, breakers, contactors, wire ends, wire tray, panduit, etc) on the machine for our in-house retrofits (5 axis).

  5. #25
    Join Date
    Nov 2001
    Location
    WAPELLO, IA USA
    Posts
    6,395
    Post Thanks / Like
    Likes (Given)
    29
    Likes (Received)
    1788

    Default

    Quote Originally Posted by Tonytn36 View Post
    Last I checked, I think we are paying in the neighborhood of $55-$60k for the entire electrical control system package. This includes the 840Dsl, servos, motors, and all other electrical devices (electrical hardware, new cabinet back plane, switches, door switches, safety relays, sensors, PB's, wiring, terminals, breakers, contactors, wire ends, wire tray, panduit, etc) on the machine for our in-house retrofits (5 axis).
    It does depend some on the SIZE of the motors. I priced a fanuc retrofit installing a new 21i, all motors, drives, spindle motor, etc, the whole ball of wax for a 3 axis vmc that ran 30 NM motors- motors and drives alone were over 60k. The control another 40-45k if I recall. This had a reasonably sized table, 60 x 30, big screws, high inertia, about 1000 ipm, I think a 2500 lb table load. In smaller machines I don't doubt your price, but it all depends. One can't buy a 496 blown alcohol top fuel engine for what one can buy a run of the mill factory reman 350....... depends what you desire.

  6. #26
    Join Date
    Dec 2003
    Country
    UNITED STATES
    State/Province
    Washington
    Posts
    4,351
    Post Thanks / Like
    Likes (Given)
    1876
    Likes (Received)
    3351

    Default

    Quote Originally Posted by Tonytn36 View Post
    Last I checked, I think we are paying in the neighborhood of $55-$60k for the entire electrical control system package. This includes the 840Dsl, servos, motors, and all other electrical devices (electrical hardware, new cabinet back plane, switches, door switches, safety relays, sensors, PB's, wiring, terminals, breakers, contactors, wire ends, wire tray, panduit, etc) on the machine for our in-house retrofits (5 axis).
    I bet you also invested some time putting drawings together, ensuring motor flanges and shafts properly integrated into the existing axes, commissioning, getting all your M-codes and I/O mapped together, HMI and MCP linked, setting up your 5-axis coordinate system, compressor functions, limit switches axis zeros, etc.

    Having similar machines and only needing to do the above once so you could leverage that NRE over several systems is a big advantage and smart way to do things.

  7. #27
    Join Date
    Dec 2007
    Location
    Southeastern US
    Posts
    6,079
    Post Thanks / Like
    Likes (Given)
    652
    Likes (Received)
    2403

    Default

    Quote Originally Posted by WILLEO6709 View Post
    It does depend some on the SIZE of the motors. I priced a fanuc retrofit installing a new 21i, all motors, drives, spindle motor, etc, the whole ball of wax for a 3 axis vmc that ran 30 NM motors- motors and drives alone were over 60k. The control another 40-45k if I recall. This had a reasonably sized table, 60 x 30, big screws, high inertia, about 1000 ipm, I think a 2500 lb table load. In smaller machines I don't doubt your price, but it all depends. One can't buy a 496 blown alcohol top fuel engine for what one can buy a run of the mill factory reman 350....... depends what you desire.
    This is for a custom machine I would equate to a 8" chuck sized lathe. However, we are also classified as a global MTB with Siemens (and Fanuc and Beckoff) so we get MTB pricing. Fanuc is more expensive than Siemens.

  8. #28
    Join Date
    Dec 2007
    Location
    Southeastern US
    Posts
    6,079
    Post Thanks / Like
    Likes (Given)
    652
    Likes (Received)
    2403

    Default

    Quote Originally Posted by motion guru View Post
    I bet you also invested some time putting drawings together, ensuring motor flanges and shafts properly integrated into the existing axes, commissioning, getting all your M-codes and I/O mapped together, HMI and MCP linked, setting up your 5-axis coordinate system, compressor functions, limit switches axis zeros, etc.

    Having similar machines and only needing to do the above once so you could leverage that NRE over several systems is a big advantage and smart way to do things.
    Yes, all of that has already been engineered. The price range I gave above is only the electrical control hardware package, doesn't include any of the assembly/build labor or supervisory engineering / commissioning time.

  9. #29
    Join Date
    Dec 2004
    Location
    uk
    Posts
    11,835
    Post Thanks / Like
    Likes (Given)
    3153
    Likes (Received)
    5159

    Default

    Linux CNC has been dead nuts reliable for me, have had 0 maintenance issues since installing it. That said, i have the 5 or so programme files backed up on another pc and could have another old PC reconfigured - installed as a controller in under a hour. Price a fanuc board or deal with there wnky back up batteries if you wish, i personally don't get why anyone spends thousands on a control with non of the interfaces and memory - capabilities that come std on a £200 PC

    Im in the process of making a gang tooled cnc lathe that’s also running Linux, its getting controlled via a 7I76E remote - networked FPGA card, am aproaching the point of makeing the final ball nut bracket for the Z axis and the controls part wired, just been too busy on other stuff, this years been my best ever.

    One big advantage of a linux CNC retrofit is you can add and modify as you go. Get it working with your current drives, use the machine with manual tool changes, get the tool changer - PLC side sorted then continue on using the machine, upgradeing a axis servo and drive is but a couple of hours down time. Adding a 4th or 5th axis is but 20-30 lines of code and a few simple changes (like telling it it now has thoes axis), Hell same for adding the 6th, 7th, 8th and 9th axis. You don't have to go from one extreme to the other in the retrofit. You very much can bite it off chunk by chunk as time allows and very much continue on upgradeing as far as you wish.

    Time wise i would budget a couple of days ahead of time sorting out drive requirements and settling on your control source with linux, highly recommend the Mesa board rout too. The actual gut and retrofit does not take that long, i did my mill in a fortnights spare time around a 12 hour 6 day a week job. Depending on your in-house skills, parts pile and resources that may be faster or take a lot longer.

    Can not stress enough though, support is not a issue, you won't need to spend hours on a help line trying to find out why it does not work down the road because you will just know it inside out.

  10. #30
    Join Date
    Mar 2017
    Country
    CHINA
    Posts
    1,695
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    737

    Default

    Quote Originally Posted by Tonytn36 View Post
    Fanuc is more expensive than Siemens.
    Fanuc in China is a good 30% cheaper than the same Fanuc in the US. Of course they won't service it in the US but .... hmmm.

    We ran into the same thing with Mazak, where they flat refused to service a machine originally sold in the US. I thought that was strange ... no matter where it was sold, you do the service you get a new customer. You don't do the service, they go somewhere else and don't buy your stuff in the future.

    Quote Originally Posted by adama View Post
    Can not stress enough though, support is not a issue, you won't need to spend hours on a help line trying to find out why it does not work down the road because you will just know it inside out.
    There was a guy at Westinghouse who would pull controls off machines getting scrapped and put them on other machines, write the executive, do the interfacing etc. That's because Westinghouse controls don't grow on trees and he was sharp. But if he ever got hit by a bus or something, the whole factory woulda been up Poop Creek

    I don't think they do that anymore .... I would think a Linux control is great for an individual but less so for a factory.

  11. #31
    Join Date
    Nov 2001
    Location
    WAPELLO, IA USA
    Posts
    6,395
    Post Thanks / Like
    Likes (Given)
    29
    Likes (Received)
    1788

    Default

    When I was looking at Linux the biggest turn offs for me:
    1. On a v m c no tool I'd numbers- had to go strictly Fanuc style with g43 hxx. I hate re arranging the same tools in a machine between programs. Call a tool Id on a acra 2100 or osp p300 and it find that tool regardless of pocket.
    2. I asked about twin spindle lathes and dual ( or triple) path. Like an 18i TT. Got blank stares and a lot of ummmmmm, we don't know.
    3. Asked about tool change wizard or template. Again a bunch of " every machine different ".

    A guy in Kansas retrofit some big cinci 5 ax machines, and those worked out for him. Linux can do a lot but it also can't do a lot.

  12. #32
    Join Date
    Dec 2008
    Country
    UNITED STATES
    State/Province
    New York
    Posts
    7,745
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    2061

    Default

    Quote Originally Posted by huleo View Post
    We have a couple older machines that are worth recontrolling. I am not immediately confident Linux is the answer because it seems to written by non-machinist types that may not understand what makes a great control. I think about anything can "make it move", but its the little things and special features that make them jam.

    If we upgrade a couple machines, one of the priorities is the take a close look at turning up speeds as much as possible. I realize there are physical and sometimes electrical limitations to what can be done. We have one machine that runs Yaskawa drives/servos/spindle. It runs 787ipm rapids and 8K spindle. We want to take a close look at where the limits are. I know ballscrews and thrust bearings could get hot if pushed too hard but I seem to remember a lot of the limits of 20yrs ago had more to do with the ability of the motion controller to read the encoders? I know we checked on one machine and the MTB most certaintly pulled every bit they could from the system at the time, but I want to determine if today's tech can allow a bit more from the same amps and servos?

    I realize machines were built with "everyday/all day" in mind, and they seem to take some objective away. For instance, rapids and accelerations are fixed whether you have 100lbs or 2000lbs on the table, and that can change the load to the servos considerably. As well, spindles seem to be fixed to run "safe" at full speed all day long.


    What we are looking at is possibly we can improve acceleration for our own needs or possibly we cannot escape acceleration limits, BUT, max rapid speed could be turned up because we can now read encoders 10x faster? Maybe our machine is fixed with 2G accel to account for max table load, BUT if we are light, maybe we can push to 3G accel, or remain at 2G but push rapids from 787 to 1000?

    Regarding the spindle, maybe we can't push the spindle to a higher "safe" limit, BUT maybe we can consider a "little more" with a certain duty cycle in mind? Usually higher RPM is used with smaller tools and maybe we only need that extra speed for one tool, for 10min/hr.

    Now, I intend to fully inspect the mechanical limitations, but I want to look more at the electrical limitations. Where might we find those limits? Is the switching fast enough in older amps?
    .
    on my toy cnc at home the limit is the computers ability to keep up. i found a better and newer computer can go faster. that said no shop normally wants to do the research and design making their own cnc. most prefer buy a cnc already made that just works with no or very little adjustments

  13. #33
    Join Date
    Dec 2004
    Location
    uk
    Posts
    11,835
    Post Thanks / Like
    Likes (Given)
    3153
    Likes (Received)
    5159

    Default

    Quote Originally Posted by WILLEO6709 View Post
    When I was looking at Linux the biggest turn offs for me:
    1. On a v m c no tool I'd numbers- had to go strictly Fanuc style with g43 hxx. I hate re arranging the same tools in a machine between programs. Call a tool Id on a acra 2100 or osp p300 and it find that tool regardless of pocket.
    2. I asked about twin spindle lathes and dual ( or triple) path. Like an 18i TT. Got blank stares and a lot of ummmmmm, we don't know.
    3. Asked about tool change wizard or template. Again a bunch of " every machine different ".

    A guy in Kansas retrofit some big cinci 5 axis machines, and those worked out for him. Linux can do a lot but it also can't do a lot.

    Don't know when you looked but you can set it up to do as you want, you may well have to write a bit of code, but you should be able to find examples, fundamentally tool number does not have to be pocket number in linux cnc anymore, they use to but use too is key, its advancing constantly still. But you do have to configure it and the default does not come that way, but its still little more than finding and copy - paste code though. When the tool gets placed in a given pocket the tool table gets updated to that effect, hence you know were everything is. Its really not a hard problem to solve and has very much been done several times now.

    Dual or tripple path you have to understand less people have done that, but the control can do multi axis at once, lots of people are using it to control multi axis robotic arms, weather there all on the same spindle - tools or different is upto you. Again the exact how you want it to work is something you have to settle on. That also very much has implications in how you programme it too. More than a few commercial machines effectively split the different turrets - spindles off to 2 almost separate controllers, linux does not have to do that.

    Same goes for tool changes, there’s some out there, but your going to have to write your own ladder logic to some degree, every machine is different in how they do it and what is done through simple mechanical switching, pneumatic timers etc and what is controlled on the PLC. Then also how the tool chain - carrosel is indexed, anything less than a absolute encoder on it and you very much have to write it into the homeing sequence on power up. To me this is not a hard problem, but you do have to do it. people are pre staging tools though and there’s some bloody out right neat approaches to changing tooling on even non tool changing machines.

    Key thing to realize with LINUX CNC or any of the other retrofits, unless some one else has done that exact machine model before, YOU have to work it out yourself. In nearly all cases you can find things to base it off, but just like making a cake, you have to play chef with the ideas and get the correct combination. With linux in particular there’s way way more retrofits out there at the smaller simpler end of the machine scale, not the high end multi spindle lathe end of the market place. Hence you have a lot less prior art examples to look at for things like multi spindle lathes for instance.

    Remember though, when it comes to linux its free, no one is paid to develop it, hence no one cares weather or not its plug and play for your given application. The commercial retrofits very much have paid engineering support already solving your problems. To a degree both work much the same its just a time versus money thing. If you have the time not the money retrofit on the linux route has some great things going for it, if your short on time but have money and need a working machine going out and buying one probably makes way more sense unless you need - want something unique.

  14. #34
    Join Date
    Oct 2013
    Location
    New Jersey
    Posts
    709
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    583

    Default

    Simple question you need to ask. Are you a machinist or electrical controls and programmer guy? Because unless a machine is making chips you are dumping money into a bottomless pit. I am a mechanical designer, electrical controls guy, and PLC programmer and built my own home CNC machine. I would never recommend a piecemeal retrofit to anyone without a ton of experience doing this.

  15. Likes motion guru, adama liked this post
  16. #35
    Join Date
    Dec 2004
    Location
    uk
    Posts
    11,835
    Post Thanks / Like
    Likes (Given)
    3153
    Likes (Received)
    5159

    Default

    ^ yeah you have gotta be totaly comfortable working on servo stuff weather that be drives or the servos and doing the sums to spec em. Electronics + PLC wise, you need that knowledge too. Then there is the software side, helps to be at least at the level your doing your own PC network stuff with zero difficulty and know how to find your own answers online when you do hit a snag.

    So if your a competent electronics engineer, have the machining skills of at least a partially competent tool maker, the mathematics of a papered engineer and at least enough PC savy to hack the pentagon, the only question left to answer is why the hell are you wasting your skill set restoring old cnc's???

    All joking aside, you really do need a broad skill set to pull it off. Above all you need the problem solveing to be able to find your own answers.

  17. Likes motion guru, triumph406 liked this post

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
  •