Recent build, dust management, and printed inserts

Hi folks! I had come across the MPCNC a few years ago, thought it was cool, but promptly forgot about it. I stumbled across it again just after Christmas 2021, and am now in the final stages of commissioning a Primo build. The design and the all of the printed parts came together very smoothly on my Ender 3 (although I had to print the core three times to get a base that didn’t warp / delaminate from the bed halfway through the print…), and I’m really impressed by all of the effort and care that has so obviously went into the system!

One suggestion for the V1 store - I bought the end-stops from V1, but I had a really hard time finding the right crimp connectors to connect to those switch terminals. I made do, but I might suggest that if Ryan can find the right crimp-on connectors for the limit switches, I for one would have happily paid a few more dollars for the 8 connectors I actually needed to save the trouble of visiting multiple stores to find the right ones and have to buy a big pack…

As seems obligatory, here’s some photos of the build. I really enjoyed looking through all of the builds on the forum and the ways that everyone seems to put their own touches on the system, so I hope that this is interesting for others as well! I went with a 590 x 320 mm working volume and the standard Z-axis height.
.
I needed more workspace anyway, so I went for the torsion box design out of 3/4" plywood & 3/4" MDF for the top; the plywood was way overkill, but was the same price as the thinner stuff, so why not…

The router is going at the back-left corner, with the rest of the table left open for other work. I wanted a nice top surface (unlike my normal “throw some 2x12s on a trestle and call it a day”), so I added a 1/4" lip around the edge with some cherry to protect the MDF edges from fraying and nested some high-density particleboard in and around the elevated MDF spoilboard and mounting area. The HDF will pry out if I ever want to replace it.

The Primo build itself came together pretty cleanly; squaring was a bit of a pain, but I think I got it close enough - with the dual-endstop and the correction on the homing, I get basically perfect squares, and the plotter-mode works!

I added some minor customizations to the build itself - mostly just added a few cable supports that clamp inside the truck and gantry rails on the X1 and Y1 axes for a first attempt at cable management.

I had the most fun with the electronics, wiring, and control system. I went with the SKR Pro from Ryan, and so needed a place for the screen as well as a place for the board itself. The workshop is pretty far away from my office, and I wanted some way to pass gcode over to the system without hauling SD cards around, which meant I needed something with wifi that I could either remote into from the desktop or run one of the web systems or similar. Raspberry Pi was the obvious solution, but about the time I was making the plans and buying everything, there were no Raspberry Pi’s to be had for love or money… I had a 2005-vintage Compaq convertible laptop (one of the early ones) lying around collecting dust, and I managed to find a version of ubuntu that still supported the 32-bit processor. The kids had (for some reason that made sense to them) pulled all of the keycaps off of the keyboard, so I’m just using it as the monitor in effect with an external keyboard and mouse. I mounted the laptop, the TFT screen, and the SKR pro all on a swinging arm that I can pull over to access the cabling in the back, swing around to get it out of the way, or leave it where I can work on the screen. A few printed parts and brackets, some hinges and plywood, and we’re ready to go!

I’ve really liked the repetier host software for controlling the system - I use the jog interface a bit, but more often find myself typing the G0 commands for absolute positions manually as a more convenient way to do things…Being able to have a full PC (even an old one) for directly managing the cnc out in the shop has been really nice so far.

I intended to close off the back and fully enclose the board, but haven’t gotten to it yet; I’d need a fan probably if I did put a lid on it, and would need to do a better job with the cables as well…


I put the charger for the laptop and the power supply for the SKR & steppers in there, to keep the cables as out of the way as possible. I have a separate box with AC line voltages that breaks out a single overall power input through a switch, with the SKR & the Makita router through an E-stop switch, as well.

I did the first recommended cuts on the foamboard, and almost instantly decided that I couldn’t try cutting anything else until I had some sort of solution for the dust - even the foam particles were getting everywhere.

For the dust collector, I took the lower mount for the makita and modified it to have slots for magnets and some retention notches, and designed an almost-annular intake nozzle for my tiny shopvac that snaps in place to the Makita lower mount with a combination of some pins and the magnets and has the hose run off horizontally, in a direction that doesn’t interact with or block anything in my build. Add a printed dust brush (just PLA), and no more dust problems! Because of the way that the makita mount is shaped, this dust intake doesn’t reduce any of my vertical working volume - everything but the brush is still at or above the level of the collet.


I’ve seen a few examples of dust shoes with mechanically secured brushes, rather than the quick and dirty superglue option that I went for out of impatience to see if it worked - that’s probably a future design improvement to allow easier replacement of the brush.

Not so much because I needed it, but because I thought they were cool and was interested to see how much it would actually help, I designed and printed a cyclone dust pre-filter as well and installed that inline with the vacuum. I completely eyeballed the shape, design, and geometry, but it seems to work; I don’t get much at all in the vacuum canister, and just about everything ends up in the coffee can…If I start making too much dust, I’m sure I’ll need a bigger dust container, but for now the coffee can and lid make it easy to empty.

With the dust collection in place and confirmation that the system works fine on the foam, I felt brave enough to start cutting some wood, but immediately started worrying about how to clamp down the workpieces…even though the spoilboard is meant to be sacrificial and replaceable, I still have some sort of an aversion to just throwing in wood screws willy-nilly… some painters tape works a lot better than I thought it might, but I still wanted something better. The simplest option was to put in at least a couple of threaded inserts so that I wasn’t beating up on the wood too much, but…no inserts without a run all the way to the store…I started wondering tonight how well PLA might be able to serve that purpose. I’ve had pretty good luck in previous projects at printing threads directly into my parts, and the MDF is pretty soft…I gave it a try.

For an instant-gratification initial test, I think that they worked pretty good. I whipped something up that looked right to the eye in Fusion360, with an 8mm hex head, metric “thread cutting” threads (which looked coarse enough and yet solid enough to hold), and an internal M4 hole. I sized the outer diameter so that I could cut a 9.5mm hole with the mill itself, or use a 3/8 drill.

The socket drove it right into the MDF, and the screw goes in and out. I was initially concerned with the part unscrewing on me, and so was planning on including a sort of hexagonal retaining ring around the insert to keep it from rotating and unscrewing once it was seated, but I’m not sure I would really need it after all; we will see. I couldn’t pull the screw+insert out of the MDF (even with only a couple of threads showing), and the top of the plastic is a couple of mm beneath the MDF surface. Ideally I’d install these from the bottom rather than the top surface, but I’ll start from here. I’m sure it will wear out much faster than a metal insert, will be easier to cross-thread, etc., but I’m intrigued that even positioning and mounting features like this can conceivably be printed and be satisfactory for some purposes. Has anyone else tried 3D printed spoilboard hardware, any lessons learned?

I’ve really enjoyed building this, and I haven’t even started any of the projects that I intended it for yet…I can’t wait to see everything this tool can do! A big thank you to the community and V1 for this awesome design, as well!

12 Likes

Nice setup you have there… and it’s already getting dirty… like it! Thanks for sharing.

I think the 3/4 ply frame on your table is not overkill; I used 2x4 KDHT Douglas fir for my table, with 3/4 CDX ply for my top/bottom sandwich, and a layer of 3/4 MDF on top that serves as a spoilboard/padding for T-track… all glued and nailed. No such thing as overkill on a cnc lol!

I also like your idea for printed inserts. I can’t count how many times I have had some project come up, and had everything ready to go except one trivial part like that. Having a printable solution means those will never end up being “that one part” as long as you have filament on hand for it. Seems like they would hold up pretty well for what they are too. Even if the inner threads wear out, at least you can print 'em in bulk cheaply and have a bunch of spares for that. Print em in nylon and they’ll probably last very, very long.

The dust can is maybe a bit small for some jobs. MDF makes such fine dust, lots of it will bypass a cyclone anyways… a 5gal bucket often lasts all day with MDF… just check your bag more often. Aluminum, an mpcnc has to cut it so slow it would take a day to fill up that coffee can. OTOH, if you’re mostly milling hardwood I’d go with 5gal; chips/shreds can build up fairly fast with hardwoods, where 2gal might get frustrating to deal with. If you get a streamlined production of such hardwood items going, a 55gal drum may even be reasonable.

1 Like

Wow what a fantastic build. Looks great!
I have always loved a torsion box designed table. Used the idea in the past and so happy with the results. I did a formal furniture project that I veneered with walnut. Looked like a beautiful solid walnut top (except for the 2 end grain sides)

So light but so very rigid.

I made my primo table (massive overkill) with TWO 3/4" Baltic Birch laminated together so 1 1/2" heavy slab.

I also love your scrape wood collection! I have the same issue of not throwing any pieces of hardwood away. Great for little projects.

I am busy building a dust collector that I found on this site.

I do like your cyclone also. I have a purchased Dust Deputy set up in my wood shop that really made me a true believer and will need to maybe try printing one like yours.

-Craig

1 Like

Thanks! I am glad that the table is as solid as it is. I didn’t go for the T-track yet - in the last several days, I used the inserts and some slotted wood beams machined on the router itself to add some adjustable clamping capability. For the moment, all of the mounting points are outside of my working area - I’ll see how long I can work with that.

I’ll see how long the coffee can lasts - I’m sure I’ll need to upgrade to a larger bucked depending on which projects I get to. Mostly I’ll be working with hardwoods.

Thanks Craig! I never like to throw away a piece of wood if there’s any chance I could use it…and I’ve found value in the smallest of little scraps in the past…

With lumber prices as they are, why waste…

My source material for the MPCNC in the near future is going to be the stack of wild cherry lumber sitting above the scrap shelves; a buddy and I took down an oak, wild cherry, and pine tree off his property and had a Sawyer come over to mill it. We’ve cured it for about a year now, and gotten the cherry planed and cleaned up in the fall. A lot of projects are queued up for this stuff - time is the challenge.

1 Like