How is everyone building a DIY CNC router system that maintains cutting tolerances? - Page 2


Page 2 of 11 FirstFirst 12345 ... LastLast
Results 13 to 24 of 122

Thread: How is everyone building a DIY CNC router system that maintains cutting tolerances?

  1. #13
    Member
    Join Date
    Jan 2005
    Location
    USA
    Posts
    1938
    Downloads
    2
    Uploads
    0

    Default Re: How is everyone building a DIY CNC router system that maintains cutting tolerance

    Quote Originally Posted by JFL4066 View Post
    Say if you nested 20 of the same part, such as a 6" x 6" square with 1/2" rounded corners and a hole drilled at the corner radius center point. Will the features in all 20 parts measure to within +/- 0.002"?, +/- 0.005" of the CAD specs? Assume generally accepted speed and feeds for wood and aluminum.

    In my experience, assembling precision machined/cut parts, especially the size and length of a CNC machine, requires precision setup instruments. A new machine generally requires commissioning from the OEM to "tweak" and adjust for each sub assembly to be square, flat, and true to the required tolerance. Especially after shipping. I see no mention of this when anyone is assembling parts from a kit or built by themselves from scratch. Everyone says to use high precision rails, bearings, etc. But assembly is more critical. When scratch building a base for example, what techniques is the builder using to make sure each corner is square, flat, and in the same horizontal plane. What are the generally accepted tolerances for those?
    I would venture to guess that most people running homemade routers don't even have the tools to measure if their parts would meet your specs.

    A little background on myself. I worked as a machinist for years before going back to school and becoming an engineer. Just so you know where I'm coming from.

    I would also say for your specs, ballscrews is your best choice.

    As for your scenario. If you want the best accuracy, you would get that from dual loop feedback using glass scales. Doing this the positioning accuracy would be to whatever the following error limit is set to for the axis in question. This is user defined and would affect things like contouring because the drive can lag/lead up tot he following error limit while in motion. For things like positioning to drill a hole you would get glass scale accuracy because the system will try everything it can to get to the exact programmed position and on a well tuned system it would get there.

    Having said that, you can also get very good positioning accuracy using ground screws. Many hobby machines use rolled screws with +/- 0.004" per foot, which sound like wouldn't fit your needs. In reality the rolled screws are better than this, but that is there allowed error. I'm not sure the specs on ground screws but they are much better and much more $$$. You could "map" the rolled screws which is software compensation for lead errors, but would need a way (glass scale) to map your screws to determine the error first. You would be looking at double ball nuts for this because backlash would need to be eliminated for your specs.

    From there it all comes down to machine rigidity. Profile linear rails and steel subframe are going to be best, but something less can stiil be good if you watch your feeds and speeds.

    Everyone says to use high precision rails, bearings, etc. But assembly is more critical.
    Well this depends on why you are using those components. Profile rails for example will provide more precision, but also more rigidity. Many people likely use these for the rigidity rather than the precision. I may want to make something fast but not requiring tight tolerances, so as long as it rolls smooth and is stiff it is good.

    For you desires, assembly will be somewhat critical but I think you are overanalyzing. A router is a really a pretty simple machine. Level the base checking those 2 rails are both level (Machinist level), make sure your rails are straight and parallel (tight music wire check), Do the same for the gantry, check it for square to the first axis (granite square), repeat for the Z.



  2. #14
    Member
    Join Date
    Apr 2016
    Location
    United States
    Posts
    738
    Downloads
    0
    Uploads
    0

    Default Re: How is everyone building a DIY CNC router system that maintains cutting tolerance

    Quote Originally Posted by JFL4066 View Post
    Say if you nested 20 of the same part, such as a 6" x 6" square with 1/2" rounded corners and a hole drilled at the corner radius center point. Will the features in all 20 parts measure to within +/- 0.002"?, +/- 0.005" of the CAD specs? Assume generally accepted speed and feeds for wood and aluminum.

    In my experience, assembling precision machined/cut parts, especially the size and length of a CNC machine, requires precision setup instruments. A new machine generally requires commissioning from the OEM to "tweak" and adjust for each sub assembly to be square, flat, and true to the required tolerance. Especially after shipping. I see no mention of this when anyone is assembling parts from a kit or built by themselves from scratch. Everyone says to use high precision rails, bearings, etc. But assembly is more critical. When scratch building a base for example, what techniques is the builder using to make sure each corner is square, flat, and in the same horizontal plane. What are the generally accepted tolerances for those?

    I am building my own extrusion-based 4 x 4 CNC. I understand your concerns about assembly and took great pains to assemble mine as precisely as I could

    I'm the guy wmgeorge mentioned who had issues with the extrusions. Virtually all extrusion-based machines in the U.S. use extrusions from 80/20. IMO, 80/20's cut tolerances are less than optimal. For cut at the ends, the tolerance is .002" per inch. So, for a 1.5" x 3" extrusion, you are looking as .003" across the short 1.5" dimension, and .006" across the long 3" dimension. Extend that over a 5'+ extrusions length, and you are going to be way out of square on assembly. 80/20 also has a cut length tolerance of .015". Quite at bit, IMO and makes a precision assembly pretty much impossible. I found that 80/20 doesn't always stay within their published tolerances. Some cuts were off more than .002" per inch. I didn't try to determine whether 80/20 stays within the length tolerance. Since I was milling anyway, it wasn't important to me. My mission was to get everything right.

    I solved the 80/20 tolerance issues with my benchtop mill and an auxiliary support structure. I milled the ends of the extrusions square and milled them for length. Rather than repeat the process, here is a link to my build thread, which is still open and ongoing: https://www.cnczone.com/forums/diy-c...machinist.html

    The CNC Router Parts Pro machines are pretty popular, but unless CNCRP has some special deal with 80/20, or mills the extrusions themselves, (which I doubt), their machines are bound to suffer from the tolerance issues. Most folks apparently don't have a problem with tolerance issues and probably don't even know they exist. I decided to build my own, because I wanted a machine that was as accurate and repeatable as I could make it, and substantially beefier than the usual kit machine.

    Most first time CNC buyers wouldn't know where to start with designing a CNC. I was one of those guys, when I bought my first kit. However, once I saw how it went together, designing my own turned out to be fairly easy. I've had to make modifications on the fly, and have run across some unanticipated issues, like designing hard stops and limit/homing switch/sensor flag placement. I'm not an engineer or a machine designer, so for me, those kinds of issues were inevitable. So far, I haven't encountered any unsolvable problems. Long story short, I wouldn't discount the idea of building your own.

    Keep in mind that extrusion-based machines are never going to be as rigid as a welded steel. However, using larger extrusions, you can make as machine that significantly stiffer than the kits.

    I tried the Fineline Saturn 2, but the one I bought was so badly put together that it wasn't worth the effort to even try to get up and running. Others have reported similar problems, albeit in less detail than I documented. From the best I can tell, the earlier models were fine, but any semblance of QC went away over a relatively short period of time. Needless to say, I cannot recommend the Saturn 2. Here's a link to my experience: https://www.cnczone.com/forums/finel...cnc-posts.html,

    My assembly was accomplished with a precision straight edge, dial indicators and a large precision machinist's level. The whole process proved to be very fussy and time consuming.

    Gary




  3. #15
    Member JFL4066's Avatar
    Join Date
    May 2019
    Posts
    21
    Downloads
    0
    Uploads
    0

    Default Re: How is everyone building a DIY CNC router system that maintains cutting tolerance

    Thanks Gary and 109jb for the replies. I did watch an extensive blog of a guy on the internet that built a very substantial 48" x 96" machine with 3 x 3 steel tubing with welded and machined flanges on the ends for assembly. He then used a tight wire and depth mic system to get the rails within half a thou flat. then he epoxied them in and then dowelled the flange connections. You are only as accurate as what you measure with. Precision ground rails are only as flat and straight as the structure/tubing you are fastening them to. The structure needs to be stronger than the rail so you pull the rail straight and not the other way around.
    Gary, you have experienced exactly what I have seen in other cases with extrusions. Frankly, most people are building machines just so they can watch the router move in 3 directions. LOL. When you are finished with your machine, what kind of tolerances will it be capable of in the 3 axis directions?
    All of the electronics really don't mean anything unless the machine itself is built with precision.
    So my assumption so far, majority of people are building/assembling machines based on the accuracy of the cuts of each piece. They are not assembling the pieces on a known reference surface to ensure the assembled part is flat, straight, and square. No one really knows how repeatable the machines are.



  4. #16
    Community Moderator ger21's Avatar
    Join Date
    Mar 2003
    Location
    Shelby Township
    Posts
    34409
    Downloads
    0
    Uploads
    0

    Default Re: How is everyone building a DIY CNC router system that maintains cutting tolerance

    They are actually very repeatable, as repeatability is somewhat independent of accuracy.

    Because repeatability is rather easy to attain, as is flatness, for most people, the only requirement is to get it square, which can easily be done by making some test cuts.

    You may be surprised how easy it is to get a machine that's remarkably accurate and consistent. If you need ultra high precision parts, that all need to fit together very well, than a router is probably not the right tool for you. If you need everything to always be within .005" or less, it get's a lot more complicated.

    Gerry

    UCCNC 2017 Screenset
    http://www.thecncwoodworker.com/2017.html

    Mach3 2010 Screenset
    http://www.thecncwoodworker.com/2010.html

    JointCAM - CNC Dovetails & Box Joints
    http://www.g-forcecnc.com/jointcam.html

    (Note: The opinions expressed in this post are my own and are not necessarily those of CNCzone and its management)


  5. #17
    Member JFL4066's Avatar
    Join Date
    May 2019
    Posts
    21
    Downloads
    0
    Uploads
    0

    Default Re: How is everyone building a DIY CNC router system that maintains cutting tolerance

    Quote Originally Posted by ger21 View Post
    They are actually very repeatable, as repeatability is somewhat independent of accuracy.

    Because repeatability is rather easy to attain, as is flatness, for most people, the only requirement is to get it square, which can easily be done by making some test cuts.

    You may be surprised how easy it is to get a machine that's remarkably accurate and consistent. If you need ultra high precision parts, that all need to fit together very well, than a router is probably not the right tool for you. If you need everything to always be within .005" or less, it get's a lot more complicated.
    I guess what I'm trying to determine is, what is the acceptable accuracy and consistency everyone is seeing in an 'average' DIY and Kit CNC router. I'm a numbers guy. LOL.



  6. #18
    Member
    Join Date
    Apr 2016
    Location
    United States
    Posts
    738
    Downloads
    0
    Uploads
    0

    Default Re: How is everyone building a DIY CNC router system that maintains cutting tolerance

    Quote Originally Posted by JFL4066 View Post
    Gary, you have experienced exactly what I have seen in other cases with extrusions. Frankly, most people are building machines just so they can watch the router move in 3 directions. LOL. When you are finished with your machine, what kind of tolerances will it be capaIble of in the 3 axis directions?

    To be perfectly honest with you, I don't know what tolerances to expect. CNCRP advertises + /- .005" accuracy or better. Given the care I've taken with my build, I expect to beat the CNCRP accuracy. CNCRP also advertises +/- .002" repeatability. I should get as good or better.

    I explored using epoxy for leveling the profile linear rails and gave up on the idea. Seems that most haven't had much luck with it. I wonder about those who seem satisfied. Was the epoxy truly level, or did they put it down and just call it good. No way to know for sure. IMO, epoxy would be a last resort.

    On the question of the linear rail mounting surface conforming to the linear rail vs the linear rail conforming to the mounting surface, in most cases, it will be the latter. The rails are no so rigid as to cause steel tubing or at least 1.5" X 3" tubing to bend. If we were talking about 1" x 1" extrusions, my answer might be different. However, I am unaware of folks trying to use anything so small, so I don't see it as an issue.

    CNCRP mounts their Y axis rails on the sides, so you have cross brace extrusions to keep the surfaces flat (within the 80/20 cut tolerance constraints). The gantry is an 80mm X 160mm cross section, which is very stiff. A linear rail rail is going to be what does the conforming.

    I mounted my Y axis rails on top, but I used 3" x 3" uprights sandwiched between 3" x 3" Y axis rails, which is very stiff.

    I bought Hiwin rails. I found them to be surprisingly flat and straight. The 25mm rails were very stiff. The 20mm (on the Y axis) were also stiff, but not nearly as stiff as the 25mm.

    Gary





  7. #19
    Member
    Join Date
    Sep 2005
    Location
    United States
    Posts
    1549
    Downloads
    0
    Uploads
    0

    Default Re: How is everyone building a DIY CNC router system that maintains cutting tolerance

    Quote Originally Posted by GME View Post

    To be perfectly honest with you, I don't know what tolerances to expect. CNCRP advertises + /- .005" accuracy or better. Given the care I've taken with my build, I expect to beat the CNCRP accuracy. CNCRP also advertises +/- .002" repeatability. I should get as good or better.

    I explored using epoxy for leveling the profile linear rails and gave up on the idea. Seems that most haven't had much luck with it. I wonder about those who seem satisfied. Was the epoxy truly level, or did they put it down and just call it good. No way to know for sure. IMO, epoxy would be a last resort.

    On the question of the linear rail mounting surface conforming to the linear rail vs the linear rail conforming to the mounting surface, in most cases, it will be the latter. The rails are no so rigid as to cause steel tubing or at least 1.5" X 3" tubing to bend. If we were talking about 1" x 1" extrusions, my answer might be different. However, I am unaware of folks trying to use anything so small, so I don't see it as an issue.

    CNCRP mounts their Y axis rails on the sides, so you have cross brace extrusions to keep the surfaces flat (within the 80/20 cut tolerance constraints). The gantry is an 80mm X 160mm cross section, which is very stiff. A linear rail rail is going to be what does the conforming.

    I mounted my Y axis rails on top, but I used 3" x 3" uprights sandwiched between 3" x 3" Y axis rails, which is very stiff.

    I bought Hiwin rails. I found them to be surprisingly flat and straight. The 25mm rails were very stiff. The 20mm (on the Y axis) were also stiff, but not nearly as stiff as the 25mm.

    Gary

    Are you about ready for the first cut Gary? Curious as to how it does.

    Retired Master Electrician, HVAC/R Commercial. FLA Saturn 2 4x4 CNC Router Mach4 Kimber 1911 45ACP


  8. #20
    Community Moderator ger21's Avatar
    Join Date
    Mar 2003
    Location
    Shelby Township
    Posts
    34409
    Downloads
    0
    Uploads
    0

    Default Re: How is everyone building a DIY CNC router system that maintains cutting tolerance

    Quote Originally Posted by JFL4066 View Post
    I guess what I'm trying to determine is, what is the acceptable accuracy and consistency everyone is seeing in an 'average' DIY and Kit CNC router. I'm a numbers guy. LOL.
    What's "acceptable" can vary greatly from person to person. That's why you can buy a machine for $1000, or a similarly sized machine for $10,000.
    Different levels of acceptable.

    Almost 15 years ago, I built a machine from MDF and plywood, that uses skate bearings on pipe for linear bearings. I have no idea how accurate it is? But if I cut multiple parts that need to fit together, they'll all be within .005 of each other. If I cut 10 differerent parts that all need to be 10" long, they'll all be the same size.

    I know it's not perfectly square at the moment, but it doesn't matter for what I currently use it for. I know it won't cut perfectly round holes, but again, they're close enough for what I need.

    Gerry

    UCCNC 2017 Screenset
    http://www.thecncwoodworker.com/2017.html

    Mach3 2010 Screenset
    http://www.thecncwoodworker.com/2010.html

    JointCAM - CNC Dovetails & Box Joints
    http://www.g-forcecnc.com/jointcam.html

    (Note: The opinions expressed in this post are my own and are not necessarily those of CNCzone and its management)


  9. #21
    Member
    Join Date
    Apr 2016
    Location
    United States
    Posts
    738
    Downloads
    0
    Uploads
    0

    Default Re: How is everyone building a DIY CNC router system that maintains cutting tolerance

    Quote Originally Posted by wmgeorge View Post
    Are you about ready for the first cut Gary? Curious as to how it does.
    Thanks for asking, Bill. Not too far off. I need to update my build with some discussion and photos. Right now I working on hard stops and sensor flags. Then, it's on to making up a spoilboard. I am going to replace some of my cabling to my steppers and have to rewire plugs for P&F sensors. I want better wire (Igus) to my steppers and I used XLR connectors for everything, including my proximity sensors. P&F sensor wires come with aviation connectors.

    Gary




  10. #22
    Member Chris D's Avatar
    Join Date
    Apr 2005
    Location
    USA
    Posts
    484
    Downloads
    0
    Uploads
    0

    Default Re: How is everyone building a DIY CNC router system that maintains cutting tolerance

    There are many issues with building a machine that is accurate and true to the axes (Square and flat). Industrial machines are built using lasers and master squares for alignment and flatness. Even large, heavy cast iron machines exhibit twist along an axis which has to be taken out with "leveling" the machine.

    For accuracy, the ball screw is what will control positioning and accuracy. The cheap ball screws we use on hobby machines are rolled threads, not ground. They are very economical for a reason, they are NOT accurate like you would find on an industrial CNC machine. I would not trust the accuracy of a rolled ball screw to anything less than .003" / foot of length.

    Using an old Tool Maker rule that I believe still applies today...

    If you need +- .002" you better have an accuracy in your 'tool' that is 10% of that amount or better. Tool in this case is the machine and thus, if you want to hold +-.002" you better have a positional accuracy of +-.0002".

    Chris D



  11. #23
    Member handlewanker's Avatar
    Join Date
    Sep 2006
    Location
    Australia
    Posts
    6281
    Downloads
    0
    Uploads
    0

    Default Re: How is everyone building a DIY CNC router system that maintains cutting tolerance

    Hi, when you said you had the facilities of a workshop with machines etc, I assume you also have a decent sized marking off table to erect the machine on...….a sheet of steel welded to a bench is not suitable, but it's as good as some need it in the end.

    I would say that if you just went and dived in and did your utmost all along the line and paid attention to the fitting and calibrating of the various axes...…..something good will no doubt eventuate and you'll learn a lot along the way......especially your shortcomings and how to overcome them to make it work.

    We all had to learn by mistakes and rectifications and the courage to accept that we're not experts in all fields.

    I would hate to think that you are one of those pedantic nit pickers that think that if it looks good on a drawing it will automatically be good when you eventually cut the metal, having in the mean time.....many years etc...….agonised over higher spec tolerances you will never achieve and even higher product specs you may never afford so that it will never get built.

    You are not the best engineer on the Planet or the most skilled craftsman that ever there was when it comes to working metal.....once you accept these facts you will be at peace with yourself and proud of you achievements.

    Anything made can be corrected if it shows up a fault or two...…...trust me, we have all been down that path.

    I would think that as the longest journey starts with the first step.....propose a concept design sketch to see what it is you want to get with and you will no doubt get many suggestions as to the practicality of it.

    If it's just wood and ally parts you settle on then at least that bit is not too hard to achieve compared to steel items.

    BTW......…..better cancel the kid's inheritance 'cos you are gonna need every penny from now onwards...….have you thought of a maximum budget you can handle?
    Ian.



  12. #24
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    635
    Downloads
    0
    Uploads
    0

    Default Re: How is everyone building a DIY CNC router system that maintains cutting tolerance

    Hi JLF - You are right to think about how to measure the parts or the machine to establish its accuracy, precision and repeatability. These days this is resolved using lasers and co-ordinate measuring machines. And Chris D says aim at 1/10th the required figure to ensure you get there. (0.001" = 0.0254mm and your workshop machines should be able to get 0.01mm"??) The other addage is to build a machine of a certain accuracy you need to build all the parts on a machine that is more accurate then what you are building. As you have a machine shop this seems to be the case?... So learn up on lasers and CMM's I think you are going to need them from day dot. Peter



Page 2 of 11 FirstFirst 12345 ... 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

How is everyone building a DIY CNC router system that maintains cutting tolerances?

How is everyone building a DIY CNC router system that maintains cutting tolerances?