I switched from a Rambo controller to the Jackpot for my MPCNC. It has been so much better. You can easily fine tune your endstop pulloffs to get your MPCNC square.
1 Like
I thought I was close to square. Measuring with a tape measure on a 6 in squareā¦.
When I try to line up mirrored handle scales with three screws, it is clear I am close, but not there. One solution is to drill out the screw holes a little bigger than needed for the number 2 screws. But I would rather tune up the machine. Thanks. My understanding is that ball screws are more precise than belted machines.
Thank you. I will look into this too.
With a 16T GT2 pulley, the machine moves 32mm for every rotation. With a 1.8 degree stepper motor, thatās 0.16mm or 6 thousandths of an inch per step. With microstepping, weāre usually well below that. Thereās a limit to how far down you can go because of the stick-slip motion that starts to occur, but we can easily command the machine to position itself within 0.05mm or a thousandth of an inch. The precision of the belted motion system isnāt going to be a limited factor for us at all.
Ballscrews are ātechnicallyā more precise, as in a single turn on a screw is less distance traveled than a single turn on the pulley of a belt, but in practice thatās meaningless for us. In some cases ballscrews will actually be worse because, especially with cheap ballscrews with single nuts etc, they can have a considerable amount of backlash. Belted systems donāt have backlash but may not be as rigid. We donāt need that rigidity as much because weāre not putting the machines under huge loads (single digit kgs of force).
Either way, this isnāt the point of this thread. Please make another post and we can discuss it there 
2 Likes
The wants of tng MPCNC was the point and I suggested the parts I see on high end benchtop cnc leading to higher precision and accuracy such as ball screws, more rigidity, etc Right? When I mentioned my skew issues, I said I knew I had to dig into it. I was not asking you to dig into it. That is absolutely not the point of the thread.
3 Likes
The Lowrider has now evolved to something that it wasnāt just a few years ago. Itās actually quite startling to compare the original to the current.
As mentioned above, itās clearly moving down into the MPCNC territory, but I do think thereās room to evolve in the same manner that the Core1 is an evolution of the Mark2 (or even earlier) Prusas.
Iām thinking about a blank sheet. Use the Core One as a model - swap out the print head of course, perhaps use ply, mdf or folded ACM as the ultimate structure and enclosure.
One size for reasons noted above, and really, really resolved through a tediously long design development program. Because of the small overall footprint (less than 30" square) so it can sit on a normal table, sadly the conduit will have to go, but thereās a point of difference anyway.
Why not use 10mm rods and bearings?
Someone mentioned why it shouldnāt be a core xy but Iāve forgotten.
But is the router too cumbersome for the smaller format?
Itās a packaging problem. See above. You can get really nice spindles in small packages (note the H3 machine above)- but you donāt find those conveniently sitting on the shelf at your local home improvement store.
Making a coreXY mount with sufficient rigidity to move Z up and down with a ~100mm travel seems to be the challenge with that format. (as it is with the LR4, just in a different part of the machine). Z axis mill rigidity is a tough problem.
Itās curious to me that the one person yet to weigh in on this is @vicious1. I suspect that re-thinking the MPCNC would be refreshing, but exhausting at the same time.
Maybe he was hoping you wouldnāt tag him and he could ignore this thread. 
3 Likes
No such luck I am following along and trying not to sway any of the discussions just yet.
I have piles and piles of sketches. Random parts made in CAD just to try out ideas. I am ALWAYS thinking about these things.
9 Likes
Linear rails would help make the solution more compact, price should stay relatively reasonable in smaller lengths, keep the GT 2mm belts, could use plate or tubes inside a printed gantry to stiffen the assembly. Could be a like a baby LR4 with the Z still on the core/carriage.
A baby brute?
1 Like
I agree with this, but I was thinking more about how bringing multiple ideas from other machines together might help solve for some of the things weāve always assumed as limits.
Since my original response I canāt help but think that the MPCNC seems to focus on that smaller rigid footprint.
This point is arguable, intentionally. But many of the uses of such a small footprint machine overlap with a MiniMill of some sort. I know that the trucks are larger, but I look at the core of an MPCNC and wonder what it might look like if it moved the X and/or Y of the work surface instead of the milling tool. Because of the 3D printed aspect, I doubt it could solve the footprint concern, but it might have a cool reprap type upgrade path where printed parts could be replaced with aluminum, which might eventually be replaced with steel. All milled on the machine. (Iām thinking like 4 pieces of conduit for each bed axis for rigidity). Handling the āthroatā of the machine with a highly rigid structure is something I havenāt solved. In my mind I keep coming up with something that looks roughly like a bed slinger printer with the milling tool doing Z and X, and the bed doing Y.
And then I look at my Lowrider and realized I just reinvented it.
1 Like
Ryan will ultimately decide if linear rails, or leadscrews or corexy make sense. Iām happy to offer my opinions. But really, I trust him to determine how to get towards the requirements. And if anyone has any novel ideas, I think we would all be interested in hearing it.
I am hoping this conversation helps give Ryan āpermissionā to push off in a particular direction. Maybe also spark his imagination or feed him some motivation.
It seems like one of the hardest choices is going to be deciding if the future would be a more accessible (and cheaper) version or a more expensive (with higher cost parts). I think both directions are interesting.
The most interesting revelation (to me) is how the machine can change now that the LR is more capable. What features would make the MPCNC better as a desktop CNC than the LR?
3 Likes
In other words, why bother making a new MPCNC instead of just focusing on the LR? What would make someone choose the TNG over a LR?
I have my preconceptions of why but I am extremely interested to hear everyone elseās input.
2 Likes
Have been wondering the oppositeā¦ What features/capabilities is LR missing that Primo has that couldnāt be addressed via LR mods, e.g captured gantry using linear rails or mashup of Primo (or similar) concepts. Note how LR4 rail/bearing rollers are separately moddible than YZ plates chonks.
Never assembled MPCNC Primo, have only printed parts for a friend.
I like how easily LR can be scaled up or down in Y. X can be scaled too without buying a whole new machine, just takes a bit more effort.
1 Like
For me, I think the question is exactly that: Whatās the actual goal? Is it just another machine with different kinematics but significant overlap or is there a set of requirements that are currently not being met by the LR4?
Personally, I notice there are VERY few people building LR4s in anything smaller than full sheet or 1/2 sheet kinda size, so I wonder if maybe thereās a focus on the MPCNC being the āsmall oneā that might be more of an implicit bias than an actual situation.
1 Like
I have a hard time understanding why someone would choose an MPCNC over even an LR3. From my perspective, there seem to be 3 major differences that matter:
- Itās a bit cheaper
- The rails are captured
- Option for more Z
To me the cost difference wasnāt a huge deal since either still seems like a bargain compared to other options out there.
Iām not sure I have a good understanding of the benefits of captured rails. It seems like it lets the router push the bit down instead of relying on gravity? It also seems to make it more flexible for storage. It would be pretty cool if you could store a CNC on one of those flip tables so it was out of the way when not using it.
The more Z option is a rigidity compromise. I also feel like the increased Z of the LR4 negates this benefit a bit.
I like the idea of something in the 12-18" usable area as suggested above. I also like the idea of trying to maximize the ratio of total size to usable size. Right now according to the calculator, you can get more usable space in the same footprint with an LR4 than an MPCNC. With something in that size, I think shipping a full kit may become a more reasonable possibility. I also think in this size, trying to use things like EMT that you can buy anywhere might be negated. It seems like various types of linear rails are available online at reasonable prices. It seems like trying to find pipe/tubing sizes that works internationally might be more trouble than itās worth if the price for other options is reasonable.
At that size, I feel like the purpose changes. I think it becomes more appropriate for things like engraving or laser usage or maybe milling PCBs (not that you canāt do those things on an LR). I kinda want to mill some brass for a branding iron kind of thing. It would be cool to put it in the back of my car and take it somewhere to make something for my dad or a friend. Technically you can pull that off with an LR but Iām unlikely to do that. Perhaps at a smaller size, it attracts a different group that may find a full sheet LR intimidating.
This is around the size of a 4040 CNC and it seems like thereās room for a cheaper option.
It would be cool to integrate a DIY rotary axis. Iām not sure what Iād do with it but it sounds fun.
I donāt know whether it makes sense to have a separate MPCNC or not. I think this group is a bit biased in that weād happily build one whether it makes sense or not.
Whether or not this goes anywhere, this is still a fun discussion.
5 Likes
Logically (as Ryan points out), they are two halves of the same coin.
There doesnāt have to be one CNC for everyone. But we also donāt want to keep supporting two if the ven diagram is completely covered by one machine.
Maybe I am being sentimental and I am just trying to keep it alive. I definitely think Ryan has the design flexability to throw out what we know and go nuts with a new design. I am hoping for something as revolutionary as when he first made the LR1. Flip the sketch upside down and go for it.
3 Likes
I totally agreeāIām hoping for the same thing! I still remember how excited I was seeing LR4 up close at Open Sauce ā24.
Iām still curious about which features the LR might be missing relative to MPCNC and other comparable CNCs (although that strays from this topicās focus)ā¦ I brought up captured rails only as the first example that came to mindānot because itās a priority. Since Iām not a mechanical engineer, Iām mostly following along to see what people propose.
If Primoās captured rails allow for more force in certain tasks for some makers, there could still be some LR4 mod options that donāt require captured rails. For instance, you could stack grid-finity containers on the YZ plates and fill them with any heavy material that fits your budget. I offer this idea semi-seriously to show there are ways to compensate for any perceived or real downward force weakness compared to the MPCNC, without needing to mod a LR4 to support captured rails.
It might be useful to test a small LR4 versus a typical MPCNC to collect data and guide the design.
Would having a single machine design that can easily (and affordably) scale up or down make attachments and modifications more widely compatible? Would selling more of one unified design simplify documenting, supporting, sourcing and potentially lower costs, whether produced directly or through outsourcing kits? 
Thanks for starting this topic!
2 Likes
Orobs opinon ā¦ Skip if you like. Tldr: mpcnc needs to be the welcome mat for noobs.
I jumped into this mix about 3 years ago with an mpcnc build because it was sub $500 and the LR3 was in development so it wasnt an option at that time. I was looking for something faster than a Maslow but the lowrider price and size were too much. It didnt work out like originally planned, but was later reconfigured it for something else. There are many mpcnc builds similar to mine: too big, but proceeding anyway. I believe because it is the only gateway to entry at that pricepoint. My mpcnc became very useful when i cut it down 55% (30x40" ā 18x30") and used it for foam carving. Many others build a lowrider once they believe printed parts are reliable.
Is it possible that the mpcnc holds its place because it is still the entry level machine, and not just the small system? It is a hopeful compromise of economy and value. I dont think the first time user build rate will be as high if it is just a small machine. It isnt ideal in those scaled up situations (not stiff enough to be big or big enough to be the āeverythingā machine) but it is functional and educational. As a first or second exposure to cnc, it is a good experience.
TNG needs to keep this āfirst-timerā vibe.
4 Likes