584,858 active members*
4,947 visitors online*
Register for free
Login
IndustryArena Forum > WoodWorking Machines > DIY CNC Router Table Machines > Design Feedback For 8020 Ballscrew and Linear Rail Design
Results 1 to 5 of 5
  1. #1
    Join Date
    Apr 2005
    Posts
    43

    Question Design Feedback For 8020 Ballscrew and Linear Rail Design

    I'm currently slapping together some of my ideas and taking some inspiration from what I've seen around the communities.

    My goals are higher precision(<.001 precision), rigidity, ease of construction(bolt together with some not too much fussing, medium cost(may become high) cost. Minimal "custom" parts, and if so, "simple" design(flat plates requiring machining on one side only). Like sized hardware throughout)As much as possible)

    That being said, the design is based on an 8020 3060 (3"x"6" profile). 20mm Hiwin HG linear rails, 16mm X 5mm pitch ball screws, and direct drive servos on the ball screws. I've opted for the "XYYZ" (twin ball screw and twin servo) type setup so there is no cross member underneath in case I want to make the table a plasma cutter.

    The main plates are 1" aluminum(possibly 3/4")(mic-6/jig plate) and can be machined from one side. They are designed to be symmetrical for left and right sides, but I may move away from that idea.
    Attachment 385294

    I'm looking for any feedback on the design thus far. Any ideas on how to maintain rigidity and precision but to reduce some of the cost? Any alternative 8020 sections? Alternative linear rail manufacturers to Hiwin?(no junk!) Alternative Rail and Plate Configuration? Should the X axis 8020 frame be mounted below the Y axis frame? The plates are quite tall and the raw material is expensive.

    I hope to take some community feedback and iterate the design.

    Maybe a configuration like this one:
    https://www.youtube.com/watch?v=ingI-Q8J0Zk

  2. #2
    Join Date
    May 2005
    Posts
    3920

    Re: Design Feedback For 8020 Ballscrew and Linear Rail Design

    First off what do you mean by precision? 0.001" by itself doesnt mean anything. Do you mean flatness of the axises over a foot, squareness over a foot, positioning to one thou.?

    I ask because extrusions simply are NOT that straight or flat. They can be very good out of the box but the warranted values arent thst good. Some vendors such as Misumi offer ground surfaces options but im not sure what sort of precision they are speccing.

    The next problem you run into is machine size which you didnt mention here. However for and machine where the cutting tool comes into contact with the work piece you will get significant deflection. Well significant if your goal is 0.001" precision. You end up with a significant engineering problem if you want to build a sizable machine that maintains 0.001" precision.

    A lot of machines are built with extrusions but i believe most builders just settle for what they get. In many cases such a router is good enough. In the end you have to decide what you really want.

    One east avenue to machine precision is to use square steel tubes and epoxy level the X axis mounting surfaces after frame assembly. This would take care of you biggest axis and with the right epoxy result in a very flat working plane for your linear rails. Epoxy leveling solves the problem of how do you achive two coplaner rail mounting surfaces in a home shop to a high degree of precision in a very limited shop environment. This still leaves the other axises but there are many options there.

    I guess the problem i have is you ability to build a machine precise to one thou with extrusions, this is hard to do in a machine shop. In this context im talking about selecting any point in 3D space and getting there under load or not Even ball screws arent accurate enough on a large machine.

    Beyond that if you need a plasma table build a plasma table. Accuracy isnt a big deal and you have no cutting loads to speak of. The need for a water tank underneath also puts into question the wisdom of trying to do a multipurpose machine.

  3. #3
    Join Date
    Apr 2004
    Posts
    5728

    Re: Design Feedback For 8020 Ballscrew and Linear Rail Design

    Like Wizard, I'm skeptical about any home-made router claiming .001" precision, no matter how it's defined. But there are things you can do to increase rigidity. A machine based on a moving gantry, especially one that's suspended on high side plates, is not going to be as rigid as one using a fixed bridge/moving table design. A bridge can be as heavy as you want to make it; the more massive the better. You could use steel tube instead of that expensive aluminum extrusion, which would be more rigid. And you'd save a set of rails; you'd only need one set to mount the table on. It's also a lot easier to make two rails line up and slide together than 4.

    I don't know why you need to build walls out of aluminum extrusion either; the main advantage it has is low weight, but that's not necessary with non-moving parts. It would be better to make the base frame out of something that's easier to adjust to squareness; in the design above, if it's a little bit out, there's no apparent way to compensate. I'm not sure about the direct-drive servos either; servo motors tend to run best at higher speeds, so most people using them run them through some kind of reduction, which also increases their effective torque.
    Andrew Werby
    Website

  4. #4
    Join Date
    Apr 2005
    Posts
    43

    Re: Design Feedback For 8020 Ballscrew and Linear Rail Design

    @wizard Machine size is appx 60"x48". I was pretty vague with tolerance and precision. After reading your answer, and taking some consideration, I think something like .001" precision over 12" XY would be ok.(square/straight/size/repeatability) Ideally.

    A lot of machines are built with extrusions but i believe most builders just settle for what they get. In many cases such a router is good enough.
    And this may be the case given how I want to build the machine.

    The 3" X 6" extrusion is fairly stiff, but when I run a quick deflection check, I get about .0007" deflection with the spindle/carriage weight in the center. I haven't done any twist/torsion calcs yet, so maybe .001" precision might be difficult with the current design.

    I want construction and parts to be relatively simple. Sure, I could use some heavy steel sections and fabricate something far stiffer than extruded aluminum, but that's not my goal. So, working within those parameters, what's are some preferred construction methods and designs?

  5. #5
    Join Date
    May 2005
    Posts
    3920

    Re: Design Feedback For 8020 Ballscrew and Linear Rail Design

    Quote Originally Posted by GisMo View Post
    @wizard Machine size is appx 60"x48". I was pretty vague with tolerance and precision. After reading your answer, and taking some consideration, I think something like .001" precision over 12" XY would be ok.(square/straight/size/repeatability) Ideally.
    I suspect that will be rather hard to achieve. Also the Z height becomes very important if you expect and sort of precision. Basically that Z axis is a big lever arm that twists your gantry beam

    And this may be the case given how I want to build the machine.

    The 3" X 6" extrusion is fairly stiff, but when I run a quick deflection check, I get about .0007" deflection with the spindle/carriage weight in the center. I haven't done any twist/torsion calcs yet, so maybe .001" precision might be difficult with the current design.
    In most DIY builds the gantry and/or its supports are the weak link in a machines design. To get to that 0.001" mark you will have to pay a lot of attention to engineering and you will need a very well equipped shop or access to one. I won't say it is impossible but the overwhelming majority of machines discussed here likely don't even come close.
    I want construction and parts to be relatively simple. Sure, I could use some heavy steel sections and fabricate something far stiffer than extruded aluminum, but that's not my goal.
    First before I comment I should tell your I'm very biased AGAINST the use of "T-Slotted" extrusions in the main structures of a machine tool. It certainly can be done in the router world but I'm not a fan of the costs and the long term stability of the machine.

    In any event I think your goals are in conflict, if you really want to achieve a very high level of precision you will be involved in a significant expenditure of time, money and materials to get there. As for T-slotted extrusion and simple< I'm not sure I agree with that idea. You will have far more parts to do a proper extrusion build than one from steel in my opinion. Trying to mount linear rails to T-slots can be frustrating and may require larger than required rails simply to span the slot properly.

    In any event I don't want to distract you from your intentions, however considering steel for the gantry beam really isn't a big deal. You can easily build a hybrid steel and aluminum machine. Depending upon the design the gantry beam can be pretty simple with just a few holes drilled in it.
    So, working within those parameters, what's are some preferred construction methods and designs?
    I'm drifting off here after a 12 hour shift so this is going to be quick.
    1. I'd strongly suggest going to the stickies to review all the opinions there.
    2. If you go the aluminum T-Slot route make sure plenty of gussets braces and other items are factored in.
    3. I you go with T-slots for the gantry, go with the largest cross section extrusions you can find for the gantry.
    4. Weight isn't automatically bad in a machine tools. Mass helps in controlling vibrations, it also helps to kee the machine in place preventing walking around.

    Dave

Similar Threads

  1. Machine 3 - 8020 + Linear Rail + High Speed Ballscrews + GRBL
    By jahnj0584 in forum DIY CNC Router Table Machines
    Replies: 28
    Last Post: 01-24-2018, 01:06 AM
  2. Linear rail leveling idea feedback; feasible or folly?
    By dikymon in forum DIY CNC Router Table Machines
    Replies: 4
    Last Post: 12-15-2013, 11:10 PM
  3. Ballscrew and linear rail eBay package
    By judotossu in forum Open Source CNC Machine Designs
    Replies: 11
    Last Post: 06-18-2012, 02:12 PM
  4. New linear bearign design 8020
    By Sil_x in forum Linear and Rotary Motion
    Replies: 12
    Last Post: 06-22-2010, 08:41 AM
  5. Homemade Linear Design, Feedback.
    By MrBean in forum Linear and Rotary Motion
    Replies: 22
    Last Post: 05-20-2005, 08:50 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
  •