I began work on Mk. I back in 2016. I finished printing the rollers the day Ryan announced the new (and hugely improved) version. I continued printing the current set of STLs I had, and figured I would circle back around some day. In the mean time I had a kid, switched jobs, moved twice, and made it through a couple hurricanes. Mid 2019 Mk. I completes it first shaky test cut. Not long after I noticed a lot of parts were starting to crack. I pushed everything into the corner of the garage to be fixed later. Last week I started taking everything apart to asses the damage. MK. I suffered complete structural failure with only a few test cuts under its belt. Turns out a few years of heat cycling in the garage does a number on PLA+.
Box of failed hopes and dreams. All parts are cracked or obsolete. There was talk of getting a filament extruder at work so maybe I can reuse all of this.
Back to Mk. II though. All the rollers are complete and are substantially better. I think I’ll actually be able to tension the belts this time. The fit on the conduit is much better too. The burly design truly is a lot beefier. It’s also good to see that acme threaded rod, end stops, and individual stepper drivers are all standard items now. I really think Mk. II will make it past the test cut stage. Great job on the redesign Ryan! After two full mpcnc prints I think my printer may need a rebuild next.
Conduit is slightly oval shaped. It must be indexed to the bearings rather than tightening the bolts to clamp down on it.
Print everything in one color. Each color filament has a different amount of fillers to get the desired color. It seems like the “gold” filament cracked worse than the black.
If Ryan comes out with an update just print it.
One construction tip to finish with. It seemed like a few people used to complain that it was hard to cut the conduit straight. Put a pipe clamp on it and use that as a guide. It clamps square to the pipe and the saw will follow down the side of it. Learned this one years ago while watching Stacey from Trucks! on Power Block TV.
So…
I have this project I have been working on for about a year that is almost done. So I am not announcing anything within the next few weeks but boy you have some timing! Well get that thing dirty and get some use on it than maybe I will have something for you to start printing.
With that said. Makes me feel pretty good that even though you have not had the best of luck with my stuff you seem to enjoy it enough to keep going with it. Actually makes me feel pretty good. Thanks for sticking around dude!
The time and money I’ve put into this makes me want to see it through. There will be dust! Besides, the mpcnc is still the cheapest and easiest entry into a non-desktop sized machine.
I was able to get enough printed to have a usable machine again. My printer blew out an idler pulley while printing the feet, but replacements will be here Monday. Pretty sure I spend more time fixing my printer than I do printing. There are a few more cracked parts on the legs but nothing critical. I’m still having some issues with fitment on the z axis rails. If I ever need to reprint the center section I’ll scale down the xyz plates by 1% in x and y. I think the current conduit is from Lowe’s so maybe I’ll try some from Home Depot to see if it fits any different. Maybe the idler caused slop in the x axis and made the parts too big?
To-do list:
Cut stuff!
Shorten the z rails. I originally had longer legs but cut them in half to help with rigidity.
Buy wire and hook up the end stops.
Reprint the z axis plates.
Cut more stuff!
Look into 32bit boards. I remember it being mentioned that the z travel speed is limited by the 8bit boards. The SKR looks interesting especially because it can do wifi and have 6 drivers.
Project ideas:
Dust collector similar to Marius’ design, but smaller. I have a full size dust collector but I’m not going to run that for hours on end. https://www.youtube.com/watch?v=tREMRr81olg
Going off of hannesb82’s post I may have figured out my z-axis fitment issues. My z is ~.5mm wider at the tool plate than it is at the top. Tomorrow I’ll sand down the tool plate to see if I can get everything parallel.
You have a lead screw now so that doesn’t come into play. All-thread had like 4000 steps to cover what a leadscrew does in 100, all good now. SKR is a complicated board to setup, only do that if you enjoy working with the firmware…a lot.
I have a single start lead screw so I’m still up in that 4000 step range. I bought this when people were just trying out lead screw and we weren’t sure if the steppers could handle 4 start. I’ll order some 4 start today.
I’ll skip the SKR for now since the 4 start will take care of things. I’m eventually building some sort of coreXY printer so I’ll get a 32 bit controller for that. I’m a developer/application analyst, and have done T3 support and QA in the past. Setting up firmware won’t be a big deal.
You can also reduce the number of microsteps to take some effort from the cpu. But the stepper still needs to rotate 4x faster with a 1 start, so you may have reduced torque at higher speeds. I would also consider a 24V PSU.
But it sounds like you’ve switched to a 4 start anyway .
4 start has been ordered along with a flex coupler.
I was able to get the z axis rails a little more parallel but it didn’t help much. My conduit is averaging 23.33mm which is ~1% undersized. By the time everything is tightened enough for the bearings to make contact you end up with a lot of flex. This is what caused Mk. I to fail. I’ll try some washers between the xyz and xy parts for now. I need to bring my calipers to Lowe’s/Home Depot and measure what they have.
Tip for shortening the z: try to see if you are able to cut the z screw in two halves. That way you can make a second z carriage, to easily swap tools (you would need an extra motor for this of course)
I would try a to rotate your conduit to see if it is an oval.
Check you printers calibration, X, Y, Z, Extrusion…and most importantly all the diagonals. I think this is a print issue. If you are able to clamp the Z mounts and corners without them touching I think the conduit is fine.
Used a new roll of filament for the center so I calibrated extrusion and printed some test cubes before I started. Rollers and corners don’t hold the conduit until just before touching. The conduit is slightly out of round and undersized. I’ve twisted it to take up as much slack as possible before tightening everything. Need to take my calipers or a piece of the gantry to the store and try all the conduit until I find pieces that fit.
What about your rollers, when inserting the conduit with no tension on the tension bolts are all bearing making contact? With no tension they should fit, if they do not I do not believe 0.17mm of on the conduit is the issue. I know it seems like the obvious problem but it generally is not.
Did you print 10mm test cubes or the 100mm on all three axis test parts? 10mm tells you nothing on the order of accuracy we need. What about the diagonals (as in, is your printer square)?
The step calibration is close but z is off a little. My xy parts are 1mm to tall. I printed washers to account for this. This week I need to take the whole printer apart to replace the x-axis idler. Everything will be recalibrated after this.
I did find an easy solution for the conduit. I cut a fruit bowl into strips and used them to shim the bolts towards the conduit. So far everything is fitting snug with the tension bolts loose!
A more permanent solution (besides getting everything printed correctly) would be to put a captive nut and 6-32 screw perpendicular to the tension bolts. They could be tightened against the side of the tension bolts to help take up slack against the conduit.
All bearings that can be shimed have been. The gantry fits tight now and is square without tightening any of tension bolts. It feels much more rigid now! I’ll temporarily wire everything with the current wire while I wait for some shielded wire.
.
