584,802 active members*
4,926 visitors online*
Register for free
Login
IndustryArena Forum > WoodWorking Machines > DIY CNC Router Table Machines > 1st Build, will this be a decent Design?
Page 1 of 5 123
Results 1 to 20 of 96

Hybrid View

  1. #1

    1st Build, will this be a decent Design?

    Hello everyone,

    up until about 2 months ago I'd never knew DIY CNC building was a thing, but now I know it is, I really want to 'have a go,' I'm a complete newbie though so I know there are loads to get my head around 1st!!

    I've been playing around with different designs for a few months now, researching and gaining inspiration and this Design is what I've come up with, I've decided to start to buy some of the parts, but I'm wondering if this Design would be good/strong enough, I wanted to build the Gantry so it extends out (Cantilevers) to give me maximum cutting on the X-Axis, the footprint is 1000mm x 850mm (and I'm hoping the cutting area will be approx 860mm x 830mm) this is more than enough for the main purpose of this machine which is to make cabinets, I would like to be able to cut Aluminium but mainly just hardwoods as well as MDF, etc.


    I have two 4A Nema 23's on the Y-Axis and one 4A Nema 23 on each of the X-axis and Z-axis. All the Liner Rails are Hiwin HGR15's with 16mm Ballscrews. The sub Frame is 2020 Black Ionized Aluminium Profile attached to 2080 Black Ionized Aluminium Profile as the main Frame, the Gantry is made of two 4080 C-Beams, I've rotated the top Beam by 90 degrees to give me extra room for the drag chain. All the Yellow parts in the Visualization will be 3D Printed Parts.

    sorry for the quality, I had to reduce it to be able to upload it.

    Cheers for looking, any advice/help would be much appreciated.

    Cheers,

    Phil

  2. #2
    Join Date
    Mar 2003
    Posts
    35538

    Re: 1st Build, will this be a decent Design?

    I've decided to start to buy some of the parts, but I'm wondering if this Design would be good/strong enough,
    Buying parts too soon is the worst thing you can do, and will cost you much more down the road.

    Everyone asks if their design is "good enough". The problem, is that everyone's expectations are different. What's good enough for some, may be totally unacceptable to others. You really can't build a machine that's too rigid. The better you build it, the more acceptable it will be to you. Budget always comes into play.

    You want to be sure you get screws with 10mm pitch. The common 5mm pitch screws are a poor choice for a router. 2010 screws are probably the best choice.

    3D printed parts are just not strong enough imo for a machine of this size. Look at all the builds on this forum, and you won't find any machines using 3D printed parts like that.

    What drives do you plan on using?

    Is that a single linear rail for the Z axis? If so, that's not going to work well.
    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)

  3. #3
    Join Date
    Jul 2018
    Posts
    6248

    Re: 1st Build, will this be a decent Design?

    Hi Phil -

    If you have a 3D printer then that's a good way to make prototype parts but they are not stiff enough as a general comment to work long term. The other issue is that you can design printed parts that can't be made conventionally or economically. So if you prototype printed parts keep in mind how the real parts will be. eg billet aluminium, laser cut metal etc. If you must go the printed way I suggest you look at carbon fibre filament that would be the best shot at present. Unless you have a desktop metal printer? (its on my wishlist)

    The single rail for Z will work initially but will develop a wobble quite quickly as the balls settle in and wear. Been there done that, the Z definitely needs two rails.

    The twin gantry design is poor, better off using a large single beam then split it into two. It's not stiff enough. And lastly as Gerry points out don't buy anything until you have resolved your design and costed the parts and found a supplier. Everything is connected and if one thing changes then you have to change many things. The forum is littered with machines that get changed and changed and changed because of this. This blows the budget and makes a one year project take 3 years....

    Don't be shy of making a plywood machine for your first one. Extrusions are very dear (especially black) and have many downsides for a newbie. Personally I wouldn't use them for any machine. Ply is easy to work with, modify and learn from. The machine size you are talking of is a good starter machine. Once you have the ply machine going you can make your next machine parts from that. You won't need 4A motors and 3A motor systems are cheaper as you can get integrated controllers in 3A which are infinitely easier to wire than separate components for a newbie. Cheers Peter

  4. #4

    Re: 1st Build, will this be a decent Design?

    Thanks, guys for your replies I've only purchased an assortment of Screws, 90-degree brackets, the Motors and the Linear Guides/Blocks so far so nothing too extravagant yet. I'm happy to use something else instead of the Printed parts, I just didn't think those Printed parts would really affect the Machine itself as they are mainly just holders for the Motors.

    The thing I'm really not sure of, is the electrics, Gerry you asked what Drivers will I be using and all I can say is I have absolutely no clue at the moment on any of the electrics it's just over my head when I start reading the electrical terms etc... This is the main reason I signed up to the site to try and get my head around the Electric side of things, but there's just so much info I cand see the wood for the trees!!

    Thanks again for your replies, I appreciate the time you've both spent to comment and I will take what you've said onboard.

    Phil

  5. #5
    Join Date
    Mar 2018
    Posts
    70

    Re: 1st Build, will this be a decent Design?

    I would at least use closed Loop Steppers. Better go with servos.

  6. #6
    Join Date
    Mar 2003
    Posts
    35538

    Re: 1st Build, will this be a decent Design?

    Quote Originally Posted by millhouse_ View Post
    I would at least use closed Loop Steppers. Better go with servos.
    There's no reason you need either of those. Open loop steppers are used by the vast majority of people, and work just fine.

    What is your overall budget?
    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)

  7. #7

    Re: 1st Build, will this be a decent Design?

    I've not looked into Servo Motors purely due to reading that they are considerably more costly than standard motors and my budget is tight...

  8. #8
    I’m running ihsv57-30-18-36-01-BY from JMC and they have incredible amounts of torque and speed. Also ultra quiet and I found them on aliexpress for $100 each with integrated drivers. Seems fairly competitive to steppers after you buy drivers especially compared to closed loop steppers and these are actual AC servos driven by an internal vector drive. It’s a dc to AC servo.

    Quote Originally Posted by BigFellaPhil View Post
    I've not looked into Servo Motors purely due to reading that they are considerably more costly than standard motors and my budget is tight...

  9. #9
    Quote Originally Posted by Uguessedit View Post
    I’m running ihsv57-30-18-36-01-BY from JMC and they have incredible amounts of torque and speed. Also ultra quiet and I found them on aliexpress for $100 each with integrated drivers. Seems fairly competitive to steppers after you buy drivers especially compared to closed loop steppers and these are actual AC servos driven by an internal vector drive. It’s a dc to AC servo.
    Here is a photo of my benchtop router. I designed it with the idea when I have the space I can move the z axis and all the electrical hardware to a bigger table some day. It can machine steel at slow ipm and aluminum without issue. Though it’s running I do plan to install Y axis hiwin rails (the weak point) and extend the length to 48 inches with a cast Alcoa plate. I can tell you in retrospect that slotted tables stink, mdf stinks, unless all you cut is wood. The t slots won’t ever be as flat as you want if you do any type of precision work. Considering router are not really that they can do very well setup good. My custom z axis and gantry are stout and using all my physical force cannot flex it where round rails will flex easily. I used cast Alcoa for all the parts I made. Cost was around $2/lb for drops at the local aluminum yard.. I bought most everything a year ahead and pre planned my build. The only thing I changed was rid the steppers for servos. The rest has been part of my design plan from the get go. If you can try and keep a machine usable and upgrade it along the way to minimize downtime and gain experience on it you will have more enjoyment versus it sitting there waiting for your ideal build to get finished.

  10. #10
    Join Date
    May 2005
    Posts
    3920

    Re: 1st Build, will this be a decent Design?

    A thought just occurred! The power supply issue can be completely eliminated by buying stepper drives with built in supplies ready to be wired up to line voltages. These drives will cost a bit more but it does eliminate a whole bunch of design effort and can save panel space.

  11. #11
    Quote Originally Posted by wizard View Post
    A thought just occurred! The power supply issue can be completely eliminated by buying stepper drives with built in supplies ready to be wired up to line voltages. These drives will cost a bit more but it does eliminate a whole bunch of design effort and can save panel space.
    Isn’t it 50hz 380v in the UK? There should be plenty of psu’s that plug in over there to 48v DC or whatever you choose for under $100usd. Just a quick look at aliexpress there is this US $408.64 10? Off | 12V/24V/36V/48V/50V Switching Power Supply Input 110V 220V 380V AC to DC LED Smps 4000W Adjustable Power Supply
    https://s.click.aliexpress.com/e/swzP0I4g. That’s $400 however I’m sure you can find them for less. eBay has a 380v AC to 24v dc for $20 so whatever is needed there should be something compatible so he can use what he needs.

  12. #12

    Re: 1st Build, will this be a decent Design?

    I would say around £1300 - £1500 would be my Max i could go higher but it's only really a hobby Machine at the end of the day so it seems a bit daft to go too much over that, but if I have to, to get it working correctly I will

    Phil

  13. #13
    Join Date
    Mar 2003
    Posts
    35538

    Re: 1st Build, will this be a decent Design?

    The best bank for the buck would probably be these drives, with a 60V DC power supply.


    https://www.omc-stepperonline.com/di...-stepper-motor
    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)

  14. #14
    Join Date
    Aug 2009
    Posts
    230

    Re: 1st Build, will this be a decent Design?

    There's no reason you need either of those.
    Have you forgotten the all important bragging rights?

    Nah... I agree with ger21... for a first machine; as long as you don't overload the steppers or under-size them for your machine, open loop will work fine.

    I've not looked into Servo Motors purely due to reading that they are considerably more costly than standard motors
    They are. They are also more difficult to get setup properly. You should, however, look into them just for your own education.

    Re: 3D Printed Parts.
    Stepper motors can get rather warm during normal operation; and most 3D printer filament gets rather soft at elevated temperatures; PLA being the worst in this regard. At best this will manifest as cutting inaccuracy; at worst your motors will fall off because the 3D printed mount melted.
    If you look on eBay (and presumably elsewhere like Aliexpress or BangGood) you can find a variety of different motor mounting adapters; the search terms I used just now were "nema 23 stepper mount". However; make sure you read the listing carefully; there are some people selling 3D printed plastic motor mounts; you will want to avoid those.

    Re: Electronics.
    There's two related parts of the equation that need solving: the controller, and the motor drivers.

    For controllers, at this end of the scale, most people either go with a Grbl based solution; or use Mach 3 or Mach 4. Personally, I'm using Grbl; and have a version of it of my own (I do not recommend starting out with my version of Grbl, however, just because it's still in development and probably has bugs in it - use the official distribution instead). Masso and UCCNC are systems that I've seen mentioned online & in YouTube videos; however I don't personally have any experience with them so I can't say what their strengths and weaknesses are.

    Grbl runs on Arduino based hardware - either an actual Arduino (or clone) or a board that integrates all the components of an Arduino, and some other requisite hardware (either buffer chips & screw terminals, or the stepper drivers themselves). The advantage of an all-in-one board like that is that it is, well, all-in-one; just hook up your motors, set the drive current, and plug it in to your PC. That's also the disadvantage - if you fry something, the whole board needs to be repaired or replaced; in contrast to a modular system where you can just replace the part that got fried.

    For an example of an all-in-one board, look on eBay for "Grbl control board"; many of the listings are all-in-one (or very close to all-in-one; as they come as a package with both the motherboard and stepper driver modules). These do however have another disadvantage; in that they are based on the ATMega328... and the current version of Grbl has literally used up all the flash memory on the '328... there is no room for adding any features.

    An example of a more modular approach, and the one I recommend should you decide to go the Grbl route; is one you (having a 3D printer) are probably somewhat familiar with (or at least have heard of): an Arduino Mega 2560, plus a RAMPS 1.4 board, plus step-stick format stepper driver modules. The reason I recommend the Mega based solution for Grbl over the '328 is that for just a few dollars more you can get 8 times the flash storage and 32 more IO lines; enabling future expansion of your setup. Also, the RAMPS board has slots for 5 stepper drivers, and a bunch of other IO ports that are useful when setting up a CNC.

    If you go with Grbl, you can use inexpensive stepper driver modules that plug directly into the RAMPS or CNC shield board; and that are good up to about 2 amps. However, if you need more amps for your motors, you will need external drivers; and will need to run three signal lines to them from the sockets on your board (so, four wires in total to each driver - step, direction, enable, and ground). Wires that fit the sockets are available in the usual places (eBay, etc) sold as "dupont wires". If you go this route, I suggest getting an assortment of these, as well as an assortment of plastic shrouds for them (look for "dupont connector housing").

    Another option for Grbl, instead of a RAMPS board; is what is sold as an "Arduino Screw Shield" (however, do make sure that the one you get fits your Arduino - the ones labeled "nano" don't fit the regular Arduino form factor). This has the advantage of having screw terminals for each IO line; but has the disadvantage of not having any of the amenities of a RAMPS board (but, if you are doing everything externally anyway due to larger current requirements, you wouldn't be needing what the RAMPS offers anyway, so...).

    If you have a suitable PC, you can use Mach 3 as the control system. If you have a PC that's less suitable, you can use Mach 4, and use a device like a SmoothStepper or PMDX-414 (neither of which I personally have any experience with). Mach 3 can work with a parallel port equipped PC (subject to some limitations - certain PCs have BIOS/hardware issues that cause timing issues); as well as some breakout boards that handle the step pulse generation (like a SmoothStepper). However, Mach 3 is pretty much an end-of-life program; all new development is being done on Mach 4. Personally, I have only used Mach 3 with a parallel port; and I am migrating all my machines away from it, to my Grbl version.

    The hardware and software for Mach 3/4 is much more expensive than a Grbl based system. It does have more features than Grbl; but for the description of your goals for your machine; I'm not sure you'd ever use them (and certain ones I know you won't use, like lathe threading; because it's not a lathe, it's a router).

    For example, at the moment; PMDX lists a PMDX-414 breakout board + Mach4Hobby license for $322 USD.
    In contrast, eBay has a listings for an Arduino Mega 2560 clone for under $11 USD; RAMPS 1.4 boards for under $5, and 2A stepper driver modules for under $2 each.

    For stepper drivers for either a Grbl based system with larger motors, or a Mach based system; there are a wide variety of options, ranging from GeckoDrives (a little expensive, but quite a nice product; especially the V series; I have 6 of them myself) to TB6600 based modules, to the variety of modules by Chinese companies like LeadShine or Longs Motor. I personally only have experience with G203V drives from GeckoDrive; however while looking up current prices for this post; I spotted some TB6600 based modules for under $7 USD each; so I think after I post this I'll pop back over to eBay and grab a couple just to mess around with (seriously - a Big Mac combo meal at McDonald's is more than what they want for that stepper driver!). As long as the driver can supply the required current for the motor; any of them should work OK (assuming that there's no design flaw in the module - I have heard of some issues with TB6550 based modules a while back; apparently the designer didn't fully understand the data sheet from Toshiba).


    The official Grbl repository for the '328 version is here:
    https://github.com/gnea/grbl/

    The official Grbl version for the Arduino Mega 2560 is here:
    https://github.com/gnea/grbl-mega/

    The official Mach 3/4 website is here:
    https://www.machsupport.com/

    And just in case you were curious (again, I don't recommend actually starting with this, because I'm still working on it... and it might turn your dog into a geranium ) my version of Grbl-mega is here:
    https://github.com/bdurbrow/grbl-Mega


    I hope that gives you someplace to start.

  15. #15
    Join Date
    Mar 2003
    Posts
    35538

    Re: 1st Build, will this be a decent Design?

    If you have a suitable PC, you can use Mach 3 as the control system. If you have a PC that's less suitable, you can use Mach 4
    Not sure what you mean by "suitable", but Mach4 requires a more powerful PC than Mach3, just because it's modern software, rather than the 10+ year old Mach3, which was originally developed on Windows 2000.


    I do not recommend Mach3 at all, as it's obsolete at this point. Mmach4 will typically cost ~$400 to get started. I recommend UCCNC, with their AXBB-E controller. UCCNC gives you the features and flexibility of more expensive controls, at a very good price.

    To me, GRBL and all the low cost and open source stuff is more suited to small, very low cost hobby machines.
    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)

  16. #16
    Join Date
    Oct 2004
    Posts
    94

    Re: 1st Build, will this be a decent Design?

    Regarding 3D printed parts. It depends on the filament used. While PLA wouldn't be suitable for higher temperatures, ABS, PETG or a nylon variant w/carbon fiber may well be. If nothing else, 3D printed parts are helpful in prototyping because it's easy to try it, revise, and reprint.

  17. #17
    Join Date
    Aug 2009
    Posts
    230

    Re: 1st Build, will this be a decent Design?

    ger21 --

    Upon re-reading my post... yeah; I think it could stand a little clarification. Sorry about that! So...

    These days, practically speaking, in this context, 'suitable for Mach 3' translates to 'old', and probably gathering dust in a closet somewhere.

    More specifically, for parallel port use, you'd need a machine that passed the Mach 3 timing tests, and, well, had a parallel port. There were a lot of machines from back then that would not pass the timing tests; laptops were the worst in this respect; but a lot of business class desktops with remote management in the BIOS would fail also.

    which was originally developed on Windows 2000.
    Which is what I was running it on... at the time, I specifically went out on eBay and bought a copy of Windows 2000 just for this purpose. FWIW, I think I still have it around here somewhere....

    I do not recommend Mach3 at all, as it's obsolete at this point.
    I don't use it anymore; but I still do see it used (various forum posts & YouTube videos lead me to this impression). Perhaps it's just "marketplace inertia" causing this?

    There is one use case that I can think of for it today; but otherwise - yes, it's basically an end-of-life product. That one use case I can think of is this:
    1) You have a suitable PC gathering dust in a closet somewhere.
    2) You are already going to be using stepper drivers that can be directly interfaced with a parallel port using a $3 DB25 breakout board. It helps if the stepper driver has optoisolation built in.
    3) You don't have the money for a full Mach 4 (or UCCNC or Masso or... etc) setup yet.
    4) Grbl based solutions won't do what you want.

    Then you could give it a try, with the demo version first. If it doesn't work out you are only out the $3 for the DB25 breakout board; and some time spent fiddling with it.

    I recommend UCCNC,
    I haven't used it myself, so I'll have take your word that it's good stuff.

    To me, GRBL and all the low cost and open source stuff is more suited to small, very low cost hobby machines.
    Well... the X-Carve uses it, and that's available in a size that's slightly larger than what was described here in the original post. I suppose it depends on what you are expecting to get out of the machine as to how suitable it is - it's definitely less feature-packed then the other options mentioned here; but I can say from experience that it does work. It's also available for less than $20 in hardware costs... so I guess you get what you pay for?

    Personally, I'm in a bit of an odd position, so what's best for me may well be not best for anybody else... I'm a programmer who dabbles in hardware; and having full access all the way from the G-Code interpreter down to the hardware is tremendously appealing to me. I'm working on the AVR platform right now; but intend to transition to the ARM platform as soon as the official ARM version is out (I've decided not to try to port the existing version of Grbl to an ARM based Arduino; as according to Chamnit, he's completely refactoring Grbl for Arm; and as soon as he does release it, my porting work would be obsolete). One of my motivations in doing the work that I have on Grbl is I wanted to not have a PC hanging off of each CNC machine I have... no PC cost, no PC maintenance, no PC boot time (my version of Grbl is pretty close to "instant on")... more of an "appliance" user experience. Obviously, this won't apply to everyone... or even most people. I'm odd.


    Arjay --
    I have printed PLA and PETG on my 3D printer (an i3 clone built by WanHao for MonoPrice). My experience is that PLA starts quite rigid at room temperature, and really gets soft quite quickly as temperatures rise; and PETG is slightly softer at room temperature but doesn't soften until much higher temperatures than PLA does. However, the PTEG that I have will soften enough to compromise the rigidity of a motor mount. Perhaps you have better materials than I do (mine is from eSun). I have not tried nylon or CF reinforced filaments; so I can't speak to those.

    It's certainly good for prototyping parts that don't have to be rigid or high-strength while being tested; and for some parts I've directly used the 3D prints (they weren't mounts for motors that got hot, though).


  18. #18
    Join Date
    Dec 2003
    Posts
    1206

    Re: 1st Build, will this be a decent Design?

    I have never used Mach3 and it is getting a bit old now.The positive thing is that there is a large user base to ask for advice.UCCNC seems to be getting quite a following and again,I know nothing of it.Arduino based solutions just didn't click with me when I tried to use them with just the one motor to learn.I did succeed with LinuxCNC and since it costs nothing,that frees up some money for other components.Might be worth downloading and taking a look from a live installation.

    As always with ideas for a new build it might be helpful to know what ,if any,experience with CNC machines and associated software the OP has.It would be frustrating to have a machine sitting there while having to learn how to generate programs and make parts.

  19. #19

    Re: 1st Build, will this be a decent Design?

    Thanks for all the replies, there's some great info here! Thanks to Britt for your detailed reply, it's very informative. I haven't really decided what controllers to use but mainly I've been thinking about Mach4, I will look into GRBL now thanks! My Motors are all 8 wire, Nema 23's 4amp 2.83nm 400oz.inc I must admit I don't know what a lot of this means to be honest, but I've been told they are more than enough for a 'hobby machine'? I'm also not sure if these will determine what controllers I will have to get now..?

  20. #20
    Join Date
    Mar 2003
    Posts
    35538

    Re: 1st Build, will this be a decent Design?

    UCCNC is half the price of Mach4, and easier to use. You should at least consider it before going with Mach4.


    My Motors are all 8 wire, Nema 23's 4amp 2.83nm 400oz.inc I must admit I don't know what a lot of this means to be honest, but I've been told they are more than enough for a 'hobby machine'?
    "Hobby machines" come in all shapes and sizes, and with different levels of performance. The performance you get out of the steppers is very dependent on other factors. Mainly, drives, voltage, and screw pitch.
    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)

Page 1 of 5 123

Similar Threads

  1. GME's New 80/20 CNC Build - My Design
    By GME in forum DIY CNC Router Table Machines
    Replies: 405
    Last Post: 03-05-2020, 11:56 AM
  2. First half-decent machine build attempt
    By _AR_ in forum Vertical Mill, Lathe Project Log
    Replies: 19
    Last Post: 03-16-2013, 05:16 AM
  3. Design and Build
    By Eddymeister in forum DIY CNC Router Table Machines
    Replies: 34
    Last Post: 08-18-2011, 09:33 AM
  4. Build Inexpensive CNC with decent tolerances-Share your input
    By Andy990 in forum DIY CNC Router Table Machines
    Replies: 23
    Last Post: 12-26-2009, 05:49 AM
  5. What CNC design did you build first?
    By soundmotor in forum DIY CNC Router Table Machines
    Replies: 2
    Last Post: 04-29-2006, 03:05 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
  •