Welcome to Our Community

Some features disabled for guests. Register Today.

Expanding a desktop CNC

Discussion in 'General Talk' started by Fractured Tom, Oct 15, 2021.

  1. Fractured Tom

    Builder

    Joined:
    Oct 15, 2021
    Messages:
    14
    Likes Received:
    0
    Hello, I'm new to this forum and I have a question. I have a Millright M3 desktop CNC. I'm wondering if it's feasible to expand the working area while keeping the motors and electronics I already have? Basically, just build a new frame, I guess? I'm thinking 2' x 3'. Assuming I can reuse some of my components.

    If so, might I ask if someone here could point me in the right direction for a plan to move forward with?
    Or maybe there's a better forum for my question. I don't know. Been looking around for a while and Openbuilds seems like the best place to start.

    Here's the tech info on the M3.: MillRight CNC M3 Kit Bundle | Mill Right CNC | Affordable & Reliable CNC Machines

    Thanks!
     
  2. Rick 2.0

    Rick 2.0 OpenBuilds Team
    Staff Member Moderator Builder Resident Builder

    Joined:
    Dec 20, 2013
    Messages:
    2,852
    Likes Received:
    1,524
    The problem you're facing with the Millright system is it only uses 3 stepper motors and they're more than a tad bit on the weak side. Trying to pull around a 2'x3' platform with a single NEMA 17 stepper is unrealistic. It's also unlikely the system can handle larger motors nor be adjusted to drive 4 steppers.

    Sell it and put the money toward a proper system. You'll be happier down the road.
     
    Peter Van Der Walt likes this.
  3. Fractured Tom

    Builder

    Joined:
    Oct 15, 2021
    Messages:
    14
    Likes Received:
    0
    Thank you for replying. I understand that it's not a workhorse. However, I have it and I would like to get more out of it. In your opinion, how much larger could I realistically expand it? Right now, it's about 9.5" x 9.5". Could I go 20" x 20"?
     
  4. Rick 2.0

    Rick 2.0 OpenBuilds Team
    Staff Member Moderator Builder Resident Builder

    Joined:
    Dec 20, 2013
    Messages:
    2,852
    Likes Received:
    1,524
    20x20 might be possible if you use a light enough spoil board. I’d throw you some design suggestions but I’m 1500 miles from my computer and I won’t be back to it for another week. The main issue is stabilizing the bed in a center pull system.
     
  5. Fractured Tom

    Builder

    Joined:
    Oct 15, 2021
    Messages:
    14
    Likes Received:
    0
    Rick, I would definitely appreciate your design suggestions, if you really don't mind. Also, my use of this machine is limited to simple wood pieces and parts. No heavy, complicated or detailed cuts. However, if my needs change, I know I will need to upgrade. Looking forward to your ideas after you get home. Safe travels!
     
  6. Rob Taylor

    Rob Taylor Master
    Builder

    Joined:
    Dec 15, 2013
    Messages:
    1,470
    Likes Received:
    746
    So the problem here is that you've got a CNC with:
    • NEMA 17 steppers
    • DRV8825 drivers
    • Belt drive XY axes
    • Belt drive Z axis(?!)
    None of these are much good for anything beyond a diode laser etcher, to be honest. To upgrade this, you're gonna have to cut your own new set of plates. and keep them under 10x10" square. Some may have to screw together to go larger, but that's not a huge problem.

    The XY axes are both V-slot, and therefore quite easy to expand- just buy longer V-Slot extrusions. Probably around 500-700mm. Bear in mind the X axis- the gantry- will get quite floppy by itself with a router hanging off of it if you keep it just as a 4020 piece. The reason it's rigid now is because it's short, not because the metal part is particularly strong. An upgrade to a piece of C-Beam for the gantry would be in order (part of the reason I said before you need new plates- this has been engineered for its size, you can't just scale it up and expect it to work the same). Mounting it "flat" would let you keep the same carriage wheels as are on the current machine.

    You need to convert all three axes to screw drive, which luckily is quite easy and shouldn't really need tensioning at relatively short runs of 20-30". The reason this can just barely manage to get away with belt drive is because they're so short- the shorter the run of a piece of belt, the less it can stretch, and the less deflection you'll have under load. That won't be an option when your belts are much longer. You *could* convert instead to a belt "rack" drive, sort of a pinch-drive dual belt system that wouldn't use a full loop of belt, and therefore would technically be the same length as the current belts. Personally, I'd say just converting to screw drive would be way easier and you can design it right into the new plates before you cut them. You can keep it as center-driven on the moving bed, too. The Z might be a little tricky, but it's doable with an angle bracket or something. The Z rails as-is are probably fine.

    If the steppers are grunty- and I know even NEMA17s can be- then sticking with the current electronics would be ok for the time being as long as it's not gonna be a full-time aluminum platemaker- especially since screws have a better mechanical advantage than belts. I'd be looking to upgrade the drivers pretty quickly and the steppers not too long after that though, to be honest.

    I'd build the new "base" as a square of 6020 V-slot with the two rails for the Y axis in the middle, and then connect the gantry column plates to the side beams where they can be rigidly attached. At this point, you may have noticed... You have almost a completely new machine. That's what happens when you engineer for a different scale of problem though.
     
  7. Fractured Tom

    Builder

    Joined:
    Oct 15, 2021
    Messages:
    14
    Likes Received:
    0
    Rob, thank you for your insight. From my other readings across the web, the NEMA 17 steppers and belt drive are good choices at 20x20, maybe 18x18 would be even better, for milling wood. That's all I'm using is wood. I'm also using a spindle to cut, so maybe weight won't be too much of an issue with an expansion. However, I agree that the driver is not the best, but it is adequate for now.
    I know it may be hard for the CNC-experienced folks to get what/why I'm doing what I'm doing. As a woodworker, I don't get why some people want to stain Oak to look like Walnut?
    Anyway, please understand that I'm not dismissing anything you've said. You know way more about CNC's than I. I want a bigger, better machine. However, I need to make some coin first. If I can expand my current machine a little I can make the money for a proper upgrade.
     
  8. Rob Taylor

    Rob Taylor Master
    Builder

    Joined:
    Dec 15, 2013
    Messages:
    1,470
    Likes Received:
    746
    What kind of spindle? The more runout you have, the more of a problem it's going to be. If you use larger belts that you can tension more and can deal with milling forces, you're probably going to end up deflecting the frame and pushing something out of square. Hardwoods can have strong grain and be grabby- we've seen plenty of *frame* deflection problems on here, never mind drive deflection. Even hobby extrusion CNC builds aren't reeeaaally up to what we try to make them do, never mind something the next tier down.

    I don't think it's odd at all though. Some projects have constraints, both practical and for fun. I think you're probably overconstraining yourself on some of this, and may end up causing yourself more frustration or disappointment than necessary.

    I have a roughly 18x18" laser build- 4020 extrusion, NEMA 17, belt-drive: Portable Diode Laser Cutter. Different setup, but quite similar capabilities. I absolutely would not put any kind of milling spindle on there except maybe a high speed, 300W, Dremel-equivalent, unit that only takes 1/8" PCB milling tooling. It is quite rigid, but I also gave it welded 1/4" aluminum plate gantry columns- not MDF. It's getting some upgrades here soon too, when I get a chance to install them. The point of that build was to keep the price down while still being quite capable. Kind of the opposite of my latest build, but it was a lot of fun.
     
  9. Peter Van Der Walt

    Peter Van Der Walt OpenBuilds Team
    Staff Member Moderator Builder Resident Builder

    Joined:
    Mar 1, 2017
    Messages:
    13,749
    Likes Received:
    4,070
    Selling onward a (still working) machine that doesn't fit your needs, and putting that toward a LEAD1010 makes a lot of sense once you think about it.

    Expanding a machine is non trivial. Likely opening a money pit that ends in scrap pile
     
  10. Rob Taylor

    Rob Taylor Master
    Builder

    Joined:
    Dec 15, 2013
    Messages:
    1,470
    Likes Received:
    746
    I wouldn't discount the *learning process* of heavy modification or scratch-building (which I think this would have to be to get closer to a useful machine than merely a functional one). Something more akin to a C-Beam Machine if they prefer the moving-bed style. I think something about as functional as a LEAD1010 can be built for half the cost... But only if the person is willing to really throw themselves into machine design and CNC electronics- for months, potentially years. It's absolutely not worth the time cost if you just need a functional machine to do work on.

    It's not a simple task- my laser still needs electronics alteration work based on things I short-changed it on a few years ago. But I wouldn't change the education I got from it for anything. That said, I'd been manually machining for a little while there and gotten my Youtube degree from literally hundreds, maybe thousands of hours of Tom Lipton and Adam Booth. You can't Google what you don't know, you have to have a starting point.

    If you're not a dogged autodidact, I'm not sure how well the non-kit path will work out. I know some people have posted here having bought kits, even, not comprehending the amount of work that goes into CNC- whether hobby or commercial, it's never, ever, plug'n'play- and realizing they also had a money pit that ends in a scrap pile. It really depends on the individual.
     
  11. Fractured Tom

    Builder

    Joined:
    Oct 15, 2021
    Messages:
    14
    Likes Received:
    0
    Rob, it's a 400w, DC, air-cooled. 1/8" collet. So, yeah, Dremel type. It's fine for the work I'm doing, currently. For example: I recently cut a bracket decoration for a mantle I'm building. The decoration measures 6" X 6" square, cut from 1/4" Poplar, a soft wood. For this project, it would have been more efficient if could cut more than one at a time.

    Question: What if I leave the width of the machine as is and only constructed a longer moving bed?
     
  12. Fractured Tom

    Builder

    Joined:
    Oct 15, 2021
    Messages:
    14
    Likes Received:
    0
    Yes it does. If you know anyone who has one that I can borrow indefinitely, I promise I'll take good care of it.
     
  13. Fractured Tom

    Builder

    Joined:
    Oct 15, 2021
    Messages:
    14
    Likes Received:
    0
    I agree with all of this. I did buy a kit and it wasn't easy. It was/is a positive challenge.
     
  14. Fractured Tom

    Builder

    Joined:
    Oct 15, 2021
    Messages:
    14
    Likes Received:
    0
    Let me add this to the discussion. Below is a parts/features list for the M3. Which of these parts could I re-purpose if I was able to build a 20"x20" Open Builds frame for use as a CNC/Laser combination?

    – Laser cut stainless steel bearing plates
    – V slot rails on X&Y axes with acetal V wheels
    – Linear rods on Z Axis with linear bearing blocks
    – High torque NEMA 17 stepper motors
    – Steel motor mounts for the X & Y
    – 9mm fiberglass reinforced GT2 timing belts
    – Mechanically advantaged belt driven Z Axis for fast lifts
    – Space efficient moving bed design
    – Pronged tee nuts in the bed for easy work holding
    – Generic Uno, USB controller board
    – Grbl compatible stepper driver shield and DRV8825 drivers
    – 24V Volt, 120 Watt power supply
    – Spindle mount and Router mount

    I know I may be all over the place here, but I'm really just trying to get all my options together and keep cost down. I should have been born into money because I have expensive work hobbies!
     
  15. Rick 2.0

    Rick 2.0 OpenBuilds Team
    Staff Member Moderator Builder Resident Builder

    Joined:
    Dec 20, 2013
    Messages:
    2,852
    Likes Received:
    1,524
    I looked over the assembly instructions this morning and this seems fairly simple based on what you have to work with. As I said previously I won’t be able to offer much assistance until I get back home next week but in the mean time there is a bit of information I need you to gather.

    1. What are the dimensions of the front plate? (WxHxT) Get as accurate as possible please.
    2. What is the spacing of the Y-axis rails and what is their length?
    3. What is the width of the X-axis carriage assembly?
    4. Would it be possible to re-drill the X-axis carriage plate and move the 2 top wheels up 20mm? I’d recommend increasing the X-axis rail to at least a 20x60 to accommodate the additional length.

    One more question, would you have any aversion to picking up another stepper motor and another 8825 driver? I’m thinking that turning this into a fixed bed system would give you better results and could probably be done for about the same amount of money.
     
    Peter Van Der Walt likes this.
  16. Rob Taylor

    Rob Taylor Master
    Builder

    Joined:
    Dec 15, 2013
    Messages:
    1,470
    Likes Received:
    746
    Somewhat short answers to provide Googleable terms...

    Is this the X carriage? Don't see anything else matching this description. Might be reusable if you stick with 2040 or if you use C-Beam mounted sideways.

    They'd be short if you're trying to substantially add machining volume, but might be usable for some reinforcement here and there, columns, that kind of thing.

    I actually like the Z axis, I'd try to re-use it. The very small overhang from the gantry thanks to how small the Z axis assembly is is actually pretty good. Part of the way this machine functions at all.

    I mean... Maybe. They're not spectacular. But for a 400W mini mill, they might be enough.

    Possibly, though they point in the wrong direction for screws. You'd have to mount them differently.

    9mm is better than 5mm, but I still wouldn't, personally. Though you could get a lot more tension with a proper machine frame, the added resolution with screws is nice too.

    I don't understand why this was ever on there, it's terrible. Just no.

    "Space-efficient" is debatable- you need the additional space for the bed to move into, that a fixed-bed machine doesn't need. But moving beds can be faster, more rigid, more precise, and you can "hide" the extra size in a lot of desktop-scale situations (like on the Ender 3). Splitting up your motion chain *is* actually preferable. The question is what are you putting on the bed, and is your system capable of flinging it around at high speed? You need the power and the workholding.

    I've suggested this to others before, I like it. Bed's too small for an upgraded machine, but I'd stick with something like this, a quasi fixture table.

    If it works, I wouldn't replace it. If it's sketchy, I'd replace it with a branded Arduino version. Or even look toward future 32-bit upgrades and consider something like a Teensy 4.1 for grblHAL or something. Not immediately, but once you get more familiar with CNC controls, you might want the extra features or more advanced G-codes.

    These are the thing we see the most complaints about out of anything besides OpenBuilds hardware support (obviously). If it's working, don't rush to replace it, but if it's not a name-brand Protoneer board it's probably not great and DRV8825 drivers are basically just an upgrade for a4988s. Mostly not good, probably not capable of running NEMA 23s either. This is where you'll have to try and see, really. Steppers/drivers kinda go hand in hand to some extent. DM542 drivers are the sort of NEMA 23 hobby standard, they're like $30-40 each if you get somewhat branded ones (and you should, stepper drivers are a minefield). Depends where you go with it, but I'd foresee some lost steps and disappointment if you stick with NEMA 17 as you try to push more out of the machine.

    Fine for NEMA 17s as long as you're staying with those. An upgrade would need an upgrade, to what would depend. I'm personally paranoid about PSUs and only buy Mean Well. Current machine build has a $250 PSU for it. o_O

    Probably fine, hard to tell without a more rigid machine for testing but I'm sure it's ok.

    It may be that Rick can come up with something you can magic out of your box of scraps too. I agree about modifying the X carriage.
     
  17. Fractured Tom

    Builder

    Joined:
    Oct 15, 2021
    Messages:
    14
    Likes Received:
    0

    Okay, I think I have this correct. I used MM because it seemed more accurate. Let me know if something isn't right or doesn't seem right and I'll re-measure.
     

    Attached Files:

  18. Fractured Tom

    Builder

    Joined:
    Oct 15, 2021
    Messages:
    14
    Likes Received:
    0
    @Rob Taylor Having only used the M3, I have not had any issues with the belt-driven Z. How are they usually built?
     
  19. Rick 2.0

    Rick 2.0 OpenBuilds Team
    Staff Member Moderator Builder Resident Builder

    Joined:
    Dec 20, 2013
    Messages:
    2,852
    Likes Received:
    1,524
    Thank you. If you would please also provide a good square-on photo of the side with dimensions (WxH) that would be helpful. Thanks.
     
  20. Fractured Tom

    Builder

    Joined:
    Oct 15, 2021
    Messages:
    14
    Likes Received:
    0
    @Rick 2.0 This spec sheet might be of interest. Overall dimensions at the bottom.

    "Would it be possible to re-drill the X-axis carriage plate and move the 2 top wheels up 20mm?" I believe this would be possible.

    Machine Classification: Fixed Gantry, Moving Bed, 3 Axis
    Movement Range: 10.25” x 10.25” x 2”
    Cutting Tool: DeWalt DWP611 Router, 1.25 HP, 16,000-27,000 RPM, 1⁄4” collet or a 400w Spindle w1/8" collet
    (Purchased the holder for the DeWalt and I have a DWP611, but I chose the spindle due to noise. However, I can migrate to the router if that makes a different with a possible expansion)

    Millable Materials: MDF, softwood, hardwood, plywood, plastics, FR4 copper clad boards, carbon fiber, aluminum, etc. (Milling settings vary by material, cutter geometry, etc.)
    Rapid Traverse Rates: 315 IPM (X and Y); 118 IPM (Z)
    Micro-step Resolution: 0.001” (X and Y); 0.0005” (Z)
    No-load Repeatability: Approximately 0.006” or less
    Motors: NEMA 17, 76 oz/in
    Linear Motion: V wheels on extrusion (X and Y); Linear bearings on linear shaft (Z)
    Drive Mechanism: 10mm wide fiberglass reinforced GT2 belt with 20 tooth GT2 pulley
    Work holding: Grid of sixteen pronged 10-32 tee nuts
    Firmware: Grbl 0.9, USB based, preconfigured
    Construction: MDF frame, aluminum extrusion rails, aluminum bearing plates
    Finish: Black epoxy, moisture and cutting oil resistant
    Homing Switches: Yes
    Overall Dimensions: 19” x 19” x 16.5”
    Weight: 25 lbs.
     
  21. Rick 2.0

    Rick 2.0 OpenBuilds Team
    Staff Member Moderator Builder Resident Builder

    Joined:
    Dec 20, 2013
    Messages:
    2,852
    Likes Received:
    1,524
    Thanks but a good square on photo of the side would be more helpful. I need it to get the front edge height and as I can trace it, it'll save a lot of time creating a sketchup model.
     
  22. Fractured Tom

    Builder

    Joined:
    Oct 15, 2021
    Messages:
    14
    Likes Received:
    0
    Oh, I see. The issue is I have it in an enclosure and getting it out will be problematic since the wiring is external, thus wired through the walls. I'll need a couple of days to get that image.
     
  23. Rick 2.0

    Rick 2.0 OpenBuilds Team
    Staff Member Moderator Builder Resident Builder

    Joined:
    Dec 20, 2013
    Messages:
    2,852
    Likes Received:
    1,524
    Attached is a Sketchup file of the concept. It's fully developed but not fully detailed. You'll need to visualize all the screws going through the wood panels into the ends of the v-slot.

    Basic design is 650mm wide by 1000mm long. It requires (4) 20x40x1000, (1) 20x40x500, and (1) 20x60x1000.

    Millright top.png

    The X-axis rail is increased to 60mm deep and requires a couple of short lengths of 3/4" angle to brace it as it sticks above the existing framing and thus lacks support for the top part of the extrusion. Upper holes/wheels on the back of the X-axis gantry plate will need moved up to accommodate the deeper extrusion.

    The bed is supported by 450mm lengths of 1/8"x2-1/2" aluminum bar stock and uses your existing wheels.

    Millright bottom.png

    The 20x40x100mm blocking to the side walls is held to the base 20x40s with blind connections and then screws are installed through the side walls into the ends of the V-slot to provide proper bracing of the side walls.

    Note if the Y-axis NEMA 17 proves insufficient to move the bed adequately it should be fairly easy to revise to a NEMA 23 stepper and hook a stand alone driver into the Arduino.

    You may also want to put a board under the framing to fill the gap to the table rather than relying on the existing board edges to properly support the system.

    Remember this is an untested concept. Free advise is often worth exactly what you paid for it. While it should be fine for the lightly framed system you seek, you're still on your own from here.
     

    Attached Files:

  24. Fractured Tom

    Builder

    Joined:
    Oct 15, 2021
    Messages:
    14
    Likes Received:
    0
    [QUOTE="Free advise is often worth exactly what you paid for it. While it should be fine for the lightly framed system you seek, you're still on your own from here.[/QUOTE]

    Well, Rick, I do appreciate you taking the time to suggest a possible solution. To be clear, for starters, I would/may be able to keep the two side pieces of the existing unit plus the wheels, limit switches, electronics, steppers and mounting plates?

    You've provided very good "free advice". A great place to start and scaling down a bit shouldn't be a problem either. I am assuming it's a simple matter of just buying different lengths of rail in the quantities you've already shown.
     
  25. Rick 2.0

    Rick 2.0 OpenBuilds Team
    Staff Member Moderator Builder Resident Builder

    Joined:
    Dec 20, 2013
    Messages:
    2,852
    Likes Received:
    1,524
    Yes and Yes. And as far as the lengths go the Openbuilds parts store will cut them for you. Just specify in the comments section of the order.

    The Sketchup file is easily manipulated to determine the lengths you need. The flat bar supports under the bed should be roughly 50mm shorter than half the Y-axis length.
     
  26. Fractured Tom

    Builder

    Joined:
    Oct 15, 2021
    Messages:
    14
    Likes Received:
    0
    Thanks again. Glad to hear about the cut to length option. I was wondering about that.
     

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice