Multiple Setup Build

This topic contains 23 replies, has 7 voices, and was last updated by  atmonaut 2 months ago.

Viewing 24 posts - 1 through 24 (of 24 total)
  • Author
  • #46873


    Hey all, starting a build of my 2nd MPCNC, with a specific goal in mind of trying to make a machine a little better for learning multiple setup CAM operations, and hopefully a bit more effective for aluminum milling.  It will be 24″x24″, with a goal of having a cutting area of 14″x10″, the primary work-holding will be using a drill-press vice, and possibly a machinists vise in the future.  I’d love to get some feedback or suggestions, I plan on experimenting with this build a bit to get a more rigid Z axis, and I am willing to give up a lot of cutting area to get there.  This is a work in progress so stay tuned.

    Original 32″x32″ MPCNC (I will be breaking this one down and placing the new machine in the center of the table, already center cut to 24″x24″.  I use a vinyl shower curtain for dust control.  This machine runs a Dewalt 611, I have a 400w brushless  spindle on order for the new machine to try out.

    Original MPCNC 32"x32"

    Examples made on the original machine.  Tolerances generally around 0.1mm, but the Z axis definitely is not square.

    1/2" Acrylic Parts

    Acrylic Carving

    Printer is chugging away making new parts for the J version.

    I’ve ordered 0.065″ 316 Stainless tubes, all 24″ length, and some 80/20 1″x2″ for the frame.

    New Frame

    316 Stainless ruining my band-saw blade in 30 seconds.

    Bandsaw Fail

    Pipe cutter was much more effective.

    So I have legs cut now, waiting on parts to finish on the printer and hardware to arrive via amazon.  Stay tuned as I experiment with the rest of the build.  I’ll try to update this page with anything that I try, whether it works or not.




    I am interested to see your direct comparison of the 400W spindle. From what I understand is the dewalt is 600W (better) and spins faster (worse). The dewalt should also have a shorter thicker neck(?) with larger bearings. So if you can do a check with the same chip load per tooth and see what comes of it I think most of us would owe you!





    The 611 dewalt is a very beefy router.  It will be my fallback in case the 400w is not strong enough.  Here is the spindle I ordered.  I generally run adaptive or trochoidal cuts, I almost never run slots on my current machine, I’ve found that to work far better.  It wouldn’t be fair to compare the current machine to the new one due to the (hopefully) improved rigidity, so I will mount up both spindles and document the results here for everyone.




    1 user thanked author for this post.


    I cut my SS with a hacksaw. I basically made a miter box for it by screwing two 12″ long boards to some 3/4″ plywood, using some pipe to space them apart. Then I used my table saw to make a square cut 1/8″ deep in the top. I put in the pipe and then used the hacksaw through the rest of the boards and pipe.

    I don’t have a very good pipe cutter. So it would wander instead of cutting through.



    I was afraid the hacksaw would go dull just as quickly as the bandsaw blade.  I found for the pipe cutter that if I tighten it a bit and it wanders, to just tighten down a bit more and reverse the direction and it would straighten right up.  Definitely less work than a hacksaw.

    Has anyone tested the deflection of their rails?  I have a dial test indicator and was thinking about testing the conduit versus the stainless steel.  However, I don’t have any left-over conduit, and the original machine has 32″ rails, and the new has 24″, so I don’t know if my test would be useful to anyone else.  It would be neat to do some sort of standardized deflection test of various rail types and lengths.



    That would be cool a deflection test at 24″ would be good. I did a simple one at 6′ when time was hard to come by. I have not revisited it.

    Found in the first section of the FAQ’s,



    For a decent test though would include engaging the steppers and measuring bit deflection while located in the center is going to give much more meaningful results. This will test if the flex is coming more from the rails or the parts. I have 1 part on the chopping block I see as the largest weakness and it has to be fixed before I look for other rails.

    With a bit tricoidal milling 6-8mm deep on a larger build in aluminum with conduit rails has me feeling all warm inside.

    Stainless is obviously better but for the price many opt out. How much of a difference would be cool to know.


    I think the Dewalt router will probably be better for aluminum milling than the 400W brushless. Reason being that the dewalt spins way faster.

    It’s supposedly not a good thing to run faster on aluminum, but when I look at the results Kevin Lopez had versus the ones I had, I think that the obvious difference in finishing and smoothness came from his higher rpm.

    The good thing about the brushless spindle is that it will be way quieter.

    Anyway, nice build, keep us posted!



    I just receive that same model of spindle for my MPCNC, but mine is 500 watt, waiting for my 3D printer to do a mounting bracket. The aluminum one that come with the unit is to wide to work properly. Also re=designing aluminum corners to be able to change format size in a flash. Will show pictures went all done and working.



    Dui, the spindle may or may not work well with aluminum,  I am hoping it does, but I have my doubts being only 400w.  Moe, keep me posted on how your spindle testing goes.  I am starting to redesign the feet to work with this 80/20, will post files up when I am done in case anyone else is interested in using 80/20 for a frame.



    My latest aluminum stuff has been with a 300 watt dc spindle.  It’s slow, but it works.



    Quick update, I am busy drawing new parts in fusion and testing them out but I have finished planning out my build.  I’ve designed some new feet that can be attached to the 1″x2″ 80/20 extrusion with 4 bolts and t-nuts.  It’s going to be a pretty customized build but still using all the same motion components.  I spent the weekend with a set of 2 foot long calipers measuring the frame dimensions and getting it squared up.  I am going for a platform on the Z axis, to hopefully get a spindle that stays perpendicular, and can support things like vacuum hose and cutting fluid attachments without giving up precision.  Most cuts will be taking place just underneath the rails after I put a vise on the bed.



    On a build that size (the recommended size and smaller), you should not need any extra stuff to have great cuts. You are giving up a ton of work space with double rollers and I really don’t thing there will be much difference. You are making a very large machine and cutting down the usable space drastically, but if you just build a smaller machine my way I think it would be better, less moving mass and shorter rails to flex?

    I can’t wait to see the results but I think you should try in my way first and compare it to that way. I am really worried you are going to say “this works great here are my improvements” and a ton of people will read that and think they are necessary, when really it might work just the same as a regular build. It would be best if you were able to say my cut speeds in aluminum went from Xmm/s to Xmm/s but my build had to be this much larger to have the same usable space.

    Does that make sense? I am not trying to be off putting just trying to keep things easy. I love that you are trying it but a baseline would make this a really cool experiment…and could possibly lead to a redesign if it is drastically better.




    Sorry I am wrong, flawed logic. Either way we learn a ton. Keep at it, sorry for being paranoid.



    Ryan, this build was intended to be an experiment, and I will try out both the 525 middle assembly and this one to ensure a fair comparison.  I’ll reach out and get your advice on the 525 assembly to ensure its properly tuned before that test.  This is my second machine, so there are a couple of items I am really interested in based on prior experience, primarily the tram (perpendicular) and deflection at the endmill, I would like to be able to cut deeper and move through material faster with less chatter.  A big thing to keep in mind is that I’ve completely eliminated any MDF or plywood from the build, and have 1″/0.065″ stainless rails, which increases the cost considerably.  I am giving up a massive amount of cutting area, and the moving mass will be much higher, I can already see some potential pitfalls with this approach regarding binding on the parallel Y rails.  Since most of my printing is done, I can go ahead and start printing the 525 middle assembly, and switching between the two assemblies should be relatively easy, I have enough bearings to build both assemblies and swap them out.



    Made some progress today and have the platform nearly completed.  Final cutting dimensions are just over 11.5″ x 9.5″, so the platform takes away a significant amount of cutting area, and it is quite heavy.




    Still need to add homing switches and clean up the wiring, build an electronics enclosure, and add the drain pan, but the machine is functioning well.  I’ll continue to update here, especially once I finish tuning and have some time to make parts.



    First aluminum test:






    Did some dimensional tuning yesterday and got it within 0.04mm over 100mm on the X and Y axis.  Noticed that I am drawing parallelograms by 0.2mm over 100mm, definitely going to have to solve that, but thats for another day.

    Today I installed a drain pan.  Now I won’t get coolant all over the wood table and hopefully will help keep the chips under control.  My father was super helpful and had the guys at his shop build this for me out of 20 gauge steel.


    Hoisting up the machine to add the drain pan, routed a 2 inch hole in the table for the drain.

    Clamping down the work piece on parallels while tightening the drill press vise to make sure the stock stays flat.

    Running a 2d adaptive cut to face off the top of the stock.  3/16″ two flute HSS end-mill. 12000 rpm, 1000 mm/min, 1mm DOC, stepover 1.905mm, feed per tooth 0.0416 mm.

    This cut was super loud, I was shocked how much noise this machine is capable of.  The 400w spindle is super quiet, but when this cut started, every part of the machine started screaming.  I was manually spraying flood coolant on the bit with a squirt bottle the entire duration of the cut.

    The cut was stopped before I could finish the facing operation, the end-mill was gradually being pulled out of the collet throughout the cut, so it got deeper and deeper as the cut spiraled inwards.  After I stopped the machine, I checked the depth of the cut at the point of failure was 3.10mm.



    A couple things that guys here have learned about aluminum is the trocoidal method and a single flute bit. The single flute makes a world of difference over the 2 flute in both wood and aluminum it seems. I havent tried aluminum yet on my machine.



    Yeah, I need to try to find a 3/16″ single flute bit.  I have 10 or so 1/8″ single flute carbides, and they are working well in aluminum, I am currently trying to find out how fast I can cut which is why I’ve moved up to the larger bit.  The adaptive cut I am running is from fusion 360 and has a constant engagement, similar to the trochoidal cut. This cut starts on the outside of the material and engages only 1.9mm at any time throughout the cut, I will probably have to reduce that amount to keep the bit from pulling out of the collet.


    Amazing build, I love it.

    I just have one remark: You strengthened all the axis, except for the Z one… Problem is, the main issue from my experience with the MPCNC is the strenght of the Z axis. And it is not only for the MPCNC by the way, I also have a 3040 full aluminum mini CNC and the Z axis is also the one who has the less strenght.

    You’ve made an incredible build, so, really, kudos on that. I suggest you push it a little further by adding 2 more Z tubes at the opposite sides of the current ones, mirroring them. This way you’ll have the spindle fully supported on all 4 corners, it would reduce drastically any kind of flex. You’ll need to add another motor for Z but this isn’t really a problem.

    I planned to do that a few months ago, but I gave up because of other projects and the fact that it would sacrifice a lot of usable space which I need for big 3D prints.

    The super loud noise means that your spindle is either not perpendicular to the work, or your bit isn’t adapted or your cutting speeds are off, or your rpm is in the range of resonance. When it is properly tuned and when everything is ok, it shouldn’t be very loud at all. Actually that how you know your settings are ok, when the cut is as silent as it can be and when your chips are consistent.

    Again, kudos on your build and keep up the updates and pictures!

    Great work.

    A couple things that guys here have learned about aluminum is the trocoidal method and a single flute bit. The single flute makes a world of difference over the 2 flute in both wood and aluminum it seems. I havent tried aluminum yet on my machine.

    Single flute didn’t work for me with the brushless spindle. Reason being the low RPM. I think it is perfectly suited for a Dewalt spindle who can reach 30K rpms, but not really for a brushell who only reaches 10-12K.

    According to my tests, 2 flutes and even 3 flutes worked better than single flute on aluminum. Chatter was crazy with one flute.


    You can see the results of 3 different bits on 3 different cuts I’ve made with the brushless spindle: Top left is one flute, top right is 2 flutes and bottom right is 3 flutes. 2 flutes was the winner, slightly better than 3 flutes. The cut with the one flute was horrible, I thought the machine would explode.




    Thanks, I am enjoying the build and experimenting, I just added homing switches (not the prettiest implementation), but I am hoping that will help with tool changes.

    I have a lot more testing I need to do with aluminum to get it correct, and the noise really seemed to amp up after I added the steel coolant drain pan, I am looking into various ways I can dampen the entire system.  My spindle is trammed and very perpendicular.

    Currently, my Z axis seems to be a bit stronger than default just due to the way my Z axis bearings compress on the lower end of the spindle, it holds perpendicular very well for some ridiculous cut depths.  Current tests in red oak hardwood are within 0.01 mm in X direction and 0.06 mm in Y direction at 3mm DOC (with a finishing contour pass).  The X direction accuracy seems to back up your idea that 4 tubes would help, since the existing tubes are parallel to the X axis.

    I really would like to get to a 4 tube solution, in fact I already have the tubing and extra motor ready.  My plan is to get it cutting good enough in aluminum and then mill the parts out of some aluminum plate, it’s the only way I can think to reduce the overall size of the gantry, I lose 4 inches in X and Y due to the extrusion frame around the spindle alone.  I may end up ordering some linear bearings to make it more compact.

    Great note on the end mill flutes, your single flute results look similar to mine, although I at the time I was thinking it was due to the 1/8″ end-mill bending under load.  The 3/16″ two flute bull nose I am using now definitely cuts better, I just need to tune out the harmonics.


    Tuning steps/mm:


    Making a box for my ER11 collets:


    Testing Op 1 of 4 for a custom set of dominos:


    Domino jig, this didn’t end up holding the pieces as accurately or as tightly as I need, so I will have to come up with another solution:


    Setting up for Op2 to take off the remaining stock:


    Op 3, engraving and chamfering.  This was the impetus for adding homing switches, but I also need some better endmills for engraving and for chamfering.  Will be trying this again using some different work-holding, probably after the holidays.

Viewing 24 posts - 1 through 24 (of 24 total)

You must be logged in to reply to this topic.