I suspect my thoughts on this topic will not be the most popular in this group, but please understand that they are offered, not as complaints, but solely as the thoughts from one who has been participating for several years in this community, not without some frustration.
I have now built 4 MPCNC machines: Burly, Primo, LR2 and LR3.
That said, I’ve actually been wondering whether it’s time to “bite the bullet” and just purchase a moderately priced commercial machine ($2-5K) and move on from the limitations that seem to be part of this approach.
I guess there are two main reasons for this: the difficulty of getting sufficient machine accuracy and the software (especially CAM) needed to do the projects I want to do.
My LR2 was never usable, so it evolved into the LR3, which is clearly better, but still leaves ridges on the spoilboard. I had wanted to use the LR3 for slab flattening for moderate size table tops, but it’s not yet capable.
My Burly was able to do some decent work, but the looseness in the gantry meant there was always error. It’s now converted to a laser cutter/engraver and for relatively small items (< 8" x 8") it does ok although I’m having some problem with belt wobble or something.
The Primo remains promising, although gantry slop is again a problem. I’m wondering if I just need to completely reprint all those parts as well as the trucks.
The consistent issue, for me, at least, is the need to tighten things “just enough but not too much”.
But, once one has to contemplate starting over, or mostly starting over, the “build or buy” question is back on the table.
I contrast this experience with my experience with the Prusa MK3S kit assembly. As others have said, the manual is very good and helpful, but, equally important is the fact that it’s just hard to screw it up if you are reasonably careful. The calibration procedures are well thought out and, with the exception of the bed height adjustment for specific filaments, seem to just work. Prior to the Prusa, I had about a year with a different, less robust printer, which definitely helped, but IMO, the success of the Prusa is that it is so stable. Even though it’s nominally open source, it’s completely functional when you get it assembled—you don’t HAVE to change anything.
The other area that is a roadblock for folks is the CAD/CAM issue, as several others have mentioned. In my own case, I’ve used (non-professional) CAD software for over 20 years because of my woodworking hobby, so that was a non-issue.
However, the CAM was a different story. When I looked into commercially available packages back in my Burly days, I found that many didn’t have the ability to deal with Marlin in their post processor. So, I wound up defaulting to EstlCAM. Once it’s set up, it actually works fine for standard 2D cutting and engraving since the Marlin issue is included. For relief carving, it was a much bigger challenge to figure out, but even that became manageable. But, I’m sure that it remains a barrier for many.
So, with all the above rambling as background, here are some thoughts to consider going forward:
- I think that providing “complete kits” is an option that could be useful, but may result in some of the issues others have mentioned, not the least of which is cost. Instead, you could consider 2 or 3 design options, only one of which is fully customizable. For example, a detailed set of specs and parts for a 24" x 24" Primo with dual endstops and z-probe might be one choice. The specs would include not only the necessary hardware, but also the specific wiring harnesses, belts etc, and even a recommended router/spindle. The controller would probably need to be one of a couple of options, with at least one being pretty budget friendly. (I’ve never used the V1 boards for cost reasons) The “kit” could have modules that could include the plastic parts, the hardware, the motors and controller and the electrical stuff… some, or all of the above. You might even be able to find a metal vendor who would pre-fab the tubes in the correct lengths and provide them separately as “MPCNC kits”. (This would be similar to the plastic fabricators who sell plexiglass kits for the Prusa enclosure based on the IKEA “Lack”.) This would be a big help to those Primo builders who don’t have the capability to cut steel tubing themselves.
The other option would be the BYO that would be like the current versions-- fully customizable in every way, understanding that “self-sourcing” is likely.
- The CAM issue needs to be dealt with. Good post processors for the major CAM packages need to be available and just work. And, it must be pretty transparent to the user. They should not need to try to “tweak” some inadequate version themselves, which is what i had to do with my Burly. I tried, but most wouldn’t.
As part of this, a complete tutorial for setting up and using EstlCAM might be useful as an interim solution. It could include a few simple projects that implement cutting, engraving and even relief carving. This would get people over the hump to get started.
-
Consider GRBL, if the capabilities are adequate (an issue about which I know nothing!) I use a GRBL board for the laser and it seems much more transparent than Marlin. Compare the user interaction with “Lasergrbl” compared to trying to flash the Marlin firmware using PlatformIO. Again, I have no idea of the state of the art, but, my grbl board for the laser includes wifi and a (rudimentary) phone app to control the setup. It can also operate from an SD card, although I’ve not tried that.
-
If there is any conceivable way to do so, address the “slop” issue during assembly. For me, that has been one of the most frustrating things with these units. Repeated attempts at assembly is just not fun.
There are probably other ideas that could help move the product forward, but this is already far too long. But, here’s an idea from left field…
There are a number of folks who seem to make it part of their YouTube business to review and test CNC and laser products. The quality of their work is somewhat variable, but, if you could get one of them to let you put a completely assembled machine in their shop and have them run it through its paces, you might be surprised at the results. Many of these folks have hundreds of thousands of followers, so, if their impression was positive, the results would be amazing, I expect.
I would suggest that the choice should be an LR3, that is completely ready to go, and would allow them to load a project and run it, possibly after some speed and accuracy testing. Of course, the issue of assembly would have to be explained because I doubt any of these guys would take that on. But, if the demo/test was successful, a follow up on the assembly process could feed into it.
Obviously, the risk would be a poor review result (as resulted from Tom S.) but success would likely be rewarded.
Apology for the length of this ramble. Would be better done as brainstorming with pizza and beer…