. .
Page 1 of 3 123 LastLast
  1. #1
    Hello my name is Ollie.

    I have come to the preliminary conclusion that I will have to build a CNC router and thus enter the seemingly confusing realm of how exactly to go about it.

    I am a Joiner (most of the time) and Cabinetmaker (when I get the chance) and have a small but useful joinery shop at my disposal.
    The reasons for needing a CNC router are many. Sometimes I need to make one offs and copies of existing woodwork for listed buildings and stuff, often this will require getting spindle moulder knives custom ground which adds time and money, other times a great deal of hand work. A CNC router would be able to make one offs for this type of thing assuming I can draw it up on fusion 360 of course.
    I also have many ideas for small furniture projects and more art based things I would like to do. Also I can see use cases for making simple but accurate jigs for work holding and gluing up. Also advantages if making several versions of one component in different woods for example. Also perhaps some luthiery.
    I think a machine of about 1000mm by 750mm cutting area will be plenty for what I need.

    My thought process has been as follows. First look around the internet/ youtube etc.

    x carve or shapeoko 3 look fun and cheap....Rejected due to the madness of trying to keep a machine accurate whilst running open wheels on an aluminium extrusion. My sliding carriage on the spindle moulder has steel bearings on a steel guideway and it is constantly getting gummed up with dust and sap. Oh and belts look like a nightmare. So nope.

    Stepcraft looks better as it has enclosed guides and not just open wheels, also leadscrews for better accuracy, I also liked the fact that it has options for vfd spindle with software control. .... Rejected due to poor reputation from actual owners, slow movement speeds and other reported problems.

    Chinese Machines 6090 etc. Look good and cheap, proper ballscrews and supported rails.........Rejected due to being a bit of a lottery and many reported electronics and mechanical faults.

    High z Tseries look good at first........Rejected due to the fact that they seem very expensive for what you get.

    It was at about this point when I was beginning to think that I might as well just build my own.

    Sorotec, cnc router parts and Damen CNC kits. They seem to have nice kits but the price soon skyrockets once you add it all up....... Rejected, too expensive for me in one go.

    Other machines from reputable manufacturers. ...........Rejected mainly on cost front. £6k to start with plus spindle etc etc.


    Actively tried to gather funds to buy this one actually it was quite close to me too. http://www.mycncuk.com/threads/11610...640X-x-160Z%29


    Then I started looking at the possibility of making my own, found examples on this forum and here I am.

    I am hoping to be able to do the build in stages so I don`t have to fork out such a massive wodge of cash at once.

    Perhaps.
    Stage 1 build the frame probably from Aluminium extrusion and plates.
    Stage 2 get ballscrews and hiwin rails installed.
    Stage 3 Get stepper motors and drivers installed, Mach 3 etc.
    Stage 4 Inevitable faffing about, tuning everything, hair pulling out, wishing I never started and copious swearing.
    Stage 6 Make some stuff.

    I am not under the illusion that this will be achievable for £1500 and as a joiner I am painfully aware of the cost of tooling.
    However I am hoping that for maybe £5500 I will be able to make a very capable machine, better value than if I went and spent £5500 at Sorotec or something.
    Also I am prepared to take a bit of time with it maybe a year from start to finish.

    Please feel free to encourage or dissuade me as you see fit.

    I would like to know if this is a reasonable plan. My main worry is that I could end up with an unfinished project after sinking money into it like many a kit car build, this is not an option.

    Cue confused posts about the pitch of ballscrews and if I need closed loop electronics or not.....


    Sorry for the enormously long post.
    Cheers.

  2. #2
    Neale's Avatar
    Lives in Plymouth, United Kingdom. Last Activity: 21 Hours Ago Has been a member for 9-10 years. Has a total post count of 1,725. Received thanks 295 times, giving thanks to others 11 times.
    Good bit of background there, but you missed out a big detail - what sort of size machine are you looking at? And if it helps, my welded steel frame machine with a cutting area of something like 1500x750 cost around £3K to build.

    Good luck - you've come to the right place, but be prepared for the flood of (sometimes conflicting!) opinions!

  3. #3
    Quote Originally Posted by Neale View Post
    Good bit of background there, but you missed out a big detail - what sort of size machine are you looking at? And if it helps, my welded steel frame machine with a cutting area of something like 1500x750 cost around £3K to build.

    Good luck - you've come to the right place, but be prepared for the flood of (sometimes conflicting!) opinions!


    Thanks Neale

    It is good to know that it is a feasible goal at least.

    I have a (very rough and subject to change or advice) spec list of the following.

    Y axis of 1000 to 1200mm
    X axis of 600 to 750mm
    Z of 200mm

    Hiwin profiled linear rails on every axis
    Ballscrews
    Stepper motors (not sure best size 23/ 34 )

    Proper spindle ie software controlled 2.2kw /3hp sort of size with decent sized collet for 12.7 mm router bits and end mills.
    Limit/ homing switches.
    Some kind of auto height sensor for when changing bits and setting up.
    Very good dust extraction.

    I think I might as well aim for something that will machine Aluminium even though it will be mostly for wood.I think woods like Macassar Ebony and Snakewood are nearly as hard as Aluminium anyway and sometimes i need to stabilise wood with epoxy too.

    I would like accuracy and decent speed.

    I am happy to buy parts direct from China (if supplier is reputable or recommended) as I am in no rush. Willing to adapt my size a bit if it is easier to use `standard` sized ballscrews etc rather than having them customised.

    I am unsure if I should use a single central ballscrew for the Y or one each side.

    I really have little clue about the electronics side of it regarding stepper drivers and the like but would like it to be able to run without being attached to the computer all the time if thats possible.
    Very much open to suggestions regarding the electronics.

    The machine must not be too enormous as I may need to move it, I am thinking a sort of `desktop` type with a Separate frame underneath maybe with wheels.
    I have been thinking of using heavy duty alloy extrusion and plate as I have no welding skills but can probably manage to cut and shape alloy with some of the kit I have.


    Thats all I have so far ?

    Cheers

    Ollie

  4. #4
    Hi Ollie,

    Welcome, First good job on rejecting those that you did and you were spot on with your reasons.

    Provided you have the skills and patience then it's possible to build quite a nice machine with the budget your thinking. However, it does take time and careful prep/design along with will-power not to rush out and buy stuff premature. From what I've seen so far with your research and decision making you seem to have the makings of successfull builder so with little help from us chosing the right components I'm sure you'll be fine provided your confident in your abilty's regards the actual building. Be honest with your self here because like most things which look easy there is more to it than appears.!

    Good luck and don't be afraid to ask any questions you have, no matter how stupid they feel and DONT BUY ANYTHING WITHOUT ASKING or knowing it's 100% correct for your needs.

  5. #5
    Quote Originally Posted by JAZZCNC View Post
    Hi Ollie,

    Welcome, First good job on rejecting those that you did and you were spot on with your reasons.

    Provided you have the skills and patience then it's possible to build quite a nice machine with the budget your thinking. However, it does take time and careful prep/design along with will-power not to rush out and buy stuff premature. From what I've seen so far with your research and decision making you seem to have the makings of successfull builder so with little help from us chosing the right components I'm sure you'll be fine provided your confident in your abilty's regards the actual building. Be honest with your self here because like most things which look easy there is more to it than appears.!

    Good luck and don't be afraid to ask any questions you have, no matter how stupid they feel and DONT BUY ANYTHING WITHOUT ASKING or knowing it's 100% correct for your needs.


    Thanks JAZZCNC, for the encouragement and words of warning.

    I will ensure I don`t go on a spending spree without checking first.
    My plan is to design the machine on Fusion 360 until it is going to definitely work as intended (this will take a while i think as I am new to the software). Once design has been finalised (and checked over on this forum) I will begin to gather parts and then get cracking on building it.

    And now let the stupid questions begin.......

    When beginning to draw up a design, is it better to find the sizes of linear rail and ballscrew that are readily available and size the frame to those existing components. Or just draw the machine out to the desired size and worry about those after?

    Are Nema 23 steppers a good fit for a machine of this size, 1200mm maximum assuming a twin ballscrew design on the Y axis?


    Is there a best practice order of design? Z axis first, gantry (X) first or any other methodology?


    Enough questions for now I think.


    Thanks

    Ollie

  6. #6
    Neale's Avatar
    Lives in Plymouth, United Kingdom. Last Activity: 21 Hours Ago Has been a member for 9-10 years. Has a total post count of 1,725. Received thanks 295 times, giving thanks to others 11 times.
    Quote Originally Posted by Ollie78 View Post

    When beginning to draw up a design, is it better to find the sizes of linear rail and ballscrew that are readily available and size the frame to those existing components. Or just draw the machine out to the desired size and worry about those after?

    Are Nema 23 steppers a good fit for a machine of this size, 1200mm maximum assuming a twin ballscrew design on the Y axis?
    Which would you like to do? I started my machine around 3 years ago, and at that time I was nervous about ordering from China. I designed my router to the size I wanted, worked out rail/ballscrew dimensions, found a uk supplier who sold standard sizes, and redesigned to use the next off-the-shelf size up. Hiwin rails cut easily with an angle grinder but ballscrews need fiddly end machining which is beyond most people at home. These days, I would probably spec exactly what I wanted and order from someone like BST Automation in China.

    Nema 23, 3Nm or (better) 4Nm, would be fine for X and Y for this size machine. Important parameter is motor inductance, so check back before buying! But at least you have a frame size for planning purposes.

  7. #7
    Quote Originally Posted by Ollie78 View Post
    When beginning to draw up a design, is it better to find the sizes of linear rail and ballscrew that are readily available and size the frame to those existing components. Or just draw the machine out to the desired size and worry about those after?
    Design the machine you want then have the components made to size. It's highly likely because of budget you'll buy Rails and ballscrews from China and they will machine ballscrews to your spec. Same with rails, they'll cut to your exact size or like Neale says just buy little longer than needed and cut your self.
    The important part is choosing the correct diameter and pitch to suit your machine. Depending on length you'll also take into consideration end bearing type but don't worry about that yet your still some way off worrying about those things.

    Quote Originally Posted by Ollie78 View Post
    Are Nema 23 steppers a good fit for a machine of this size, 1200mm maximum assuming a twin ball screw design on the Y axis?
    Yes perfectly fine but need to look closely at the spec because not all steppers are equal and come in various flavors ie: 4 wire 8wire.
    Like Neale indicated the inductance rating is important but this rating changes depending how the phases are wired so you'll need to understand this to get full picture. Without getting too technical I'll explain quickly the differences and why one type/setup suits Router better.

    The phases of Steppers can be wired in 3 ways depending on how many wires motor come with. These are Bi-polar series, Bi-polar parallel, Uni-polar. With Modern drives only the first two of those are used so won't explain Uni-polar.

    Series Wired motors generaly provide high torque but require high voltage to achive high RPM. This is because the inductance is much higher.

    Think of inductance like pushing car with sticky brakes and the people pushing being the voltage. The more sticky the brakes the more people needed to push car the same speed.
    So lower the inductance the less voltage will be required to achive same speed as high inductance motor, or put another way, with same voltage the motor will spin faster if inductance lower.

    So series wired motor are generaly best used for machines like Mills or Lathes which don't require high feed rates(rpm) but do require more torque due to heavy design.
    Also because series wired motor uses roughly half the current of Parallel wired motor smaller lower cost drives can be used.

    Parallel wired motors are pretty much the opposite. The inductance is half that of series wired so require less folks to push, down side being when going uphill those folks will need to be stronger. In real terms what this means is that double the current will be required when wired parallel.
    In performance terms what this allows is that higher RPM can be achieved and they hold the torque further up the RPM range, down side being there is less torque at lower RPM.

    Now this is where it gets little confusing to people because motors come in 4 wire and 8 wire(often called Hybrid).
    4 wire motors are fixed by the manufacturer and can be wound either series or parallel, often they are series wound.
    8 wire motors can be wired in either way by joining wires in certain order. (they can also be wired Uni-polar)

    So cutting to the chase for Router builder your better using 8 wire motor with low inductance when wired Bi-polar parallel. This setup when combined with correct voltage(65-70Vdc) gives best performance.

    Quote Originally Posted by Ollie78 View Post
    Is there a best practice order of design? Z axis first, gantry (X) first or any other methodology?
    Like any good structure it all starts with solid foundations. Design from the bottom up and each will form the other.
    Last edited by JAZZCNC; 01-05-2018 at 07:32 PM.

  8. #8
    In terms of the machine design process it is fairly iterative and you need to go around the loop many times as every choice has a knock on effect that ripples through the design.

    Assuming you are building a fairly standard machine with moving gantry and raised X axis then I'd suggest you draw the spindle with your favourite tool in the collet and place the tip about 150mm above the top of the bed plane.

    Draw the Z axis plate that holds the spindle and make it say 300-350 mm tall by 150 wide. The bottom of the plate should not be lower than the main spindle body. Add the Z rails to the back of this plate making them at least 300 mm long and ideally matching the height of the Z plate.

    Add the lower carriages to the Z rail placing them no lower than the main body of the spindle. In the CAD program slide the Z axis (spindle, Z plate and Z rail) down until the tool hits the bed plane. The lower carraiges should remain in place.

    Add the upper Z carraiges near the top of the rail, leaving 10mm or so if you want a bit spare to face the bed.

    Draw the Y axis plate which hold the Z carraiges on the front side. On the rear of this plate add the Y carraiges making sure they do not clash with the Z carraiges and that it can be assembled!

    The Y rails can then be drawn to fit the Y carraiges and the gantry drawn.

    Add horizontal plates on the ends of the gantry and fit the X carraiges to the underside. This dictates the X rails position.

    Draw the frames to hold the X rails and then add vertical supports under these frame sides to connect down to the bed frame.

    Add the steppers, ballscrews, ballnuts, and ballscrew mounts and the fun starts. Things get tweaked and moved to get it to work. I would say this bit takes the most time so hang in there.

    Finally plenty of advice on hand on this forum to guide you. Best of luck!
    Building a CNC machine to make a better one since 2010 . . .
    MK1 (1st photo), MK2, MK3, MK4

  9. #9
    Thanks for the great responses to my questions.

    I was thinking of starting with the Z or the spindle as you suggest routercnc. I can see what you mean about everything affecting everything else which was partly where my difficulty starting.
    This gives me a good start.
    Great tip about setting the height from the bed to the tool straight away.
    I will remember Jazzcnc`s comment about starting with the strong base too....

    Do you do suggest sort of rough drawing on fusion (or whatever software ) first just using plain rectangles etc, to get the shape and then do a proper one with the profiles of the extrusion and the carriages etc, or just go straight in with the full detail.

    I have been attempting to gather dxf files of aluminium extrusions and the Hiwin stuff. does anyone have a source to a library of this sort of thing.


    Thanks

    Ollie

  10. #10
    I went to the Hiwin website and they have a selector program where you can select the carraige type the rail type etc and it will output a solid model in a range of formats. These are 3D models in a range of formats (e.g. step) not dxf which I believe are only 2D line drawings and not what you want.

    When I used the Hiwin tool the model it exported was one entity and the carraiges could not be slid around once in the CAD program. I was Ok with that as I used simple surrogate rectangles and cylinders for the design and only added holes and details at the end. Saying this you do need to think about where the bolts will go so they don't clash when you model the holes at the end and have to make changes.

    CAD for extrusions are usually on suppliers website but for initial design just use a rectangle with outer dimensions to match a commonly available size.

    If you use steel rectangular hollow section in the design make sure you model the outer corner radius as leaving them square makes it look like you can bolt things right up to the edge.
    Building a CNC machine to make a better one since 2010 . . .
    MK1 (1st photo), MK2, MK3, MK4

Page 1 of 3 123 LastLast

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Similar Threads

  1. can someone please help with hole dimensions
    By reefy86 in forum Solidworks
    Replies: 38
    Last Post: 20-02-2016, 12:44 PM
  2. Hole Radius
    By benkat in forum Machine Discussion
    Replies: 0
    Last Post: 03-03-2014, 05:29 PM
  3. SBR16UU Centre hole.
    By cambesol in forum Rails, Guideways & Bearings
    Replies: 3
    Last Post: 25-11-2013, 05:39 PM
  4. HELP! Problem drilling 2.39 dia hole in Ali casting
    By woolley2002 in forum Machine Discussion
    Replies: 4
    Last Post: 04-09-2013, 09:06 AM
  5. 6mm shaft to 6.35mm hole?
    By JuKu in forum Metalwork Discussion
    Replies: 3
    Last Post: 15-12-2012, 02:29 PM

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
  •