So I haven’t built an A-frame printer in 3 years. Other than the moronic spirit-level-and-plumb-line method, how are people going about squaring their axes these days?
I’ve been looking through docs, and I see recommendations for using a carpenter’s square to square the Z rods along the X/Z plane, but nothing to square them along the X/Y or the Y/Z plane other than measuring the distance along the edges of the machine to where the rods intersect.
Skepticism of the precision aside, this seems like it would lead to heritable error: If you build the machine from parts printed on a machine that is not square, the child machine will be a parallelogram that exhibits the same non-squareness and pass it on to any parts printed on it.
Despite the use of rigid square extrusions, the Mendelmax is not immune from this effect because it’s made up of metal-to-plastic joints rather than metal-to-metal.
Wallace (like the printrbot, which inspired it) was particularly vulnerable to this, and it’s one of the many things I’m trying to fix in Wallace 2.
I was just looking at the original Wallace design; I’m excited to see what you have in mind for the sequel!
I was also worried about the negative effects of inheritance lately when working on improving the Y axis of my MG Prusa; seems like any part that I’d print to improve the tram would probably, to some degree, inherit the error I’m trying to correct.
The Sells Mendel had adjustment points everywhere they were needed to compensate for non-square RP parts (assuming they were close enough to square to allow the machine to be assembled at all). This was also true (although less so) of the Prusa Mendel, although it sacrificed the Sells Mendel’s tolerance for non-parallelism (I tried to introduce parts to correct this at the time, but they didn’t catch on). In both cases, it’s unnecessarily difficult to do the squaring (not to mention reinforcing that squaring in all the wrong directions). Many of the more recent designs and “upgrade” parts make this adjustment impossible.
My plan for Wallace 2 will bring back the adjustability and tolerance for imperfect replication in new ways that I’ve recently come up with. If it works the way I plan, it will be impossible to print or assemble the parts wrong, while the adjustment that is necessary is obvious and straightforward. This, of course, is in addition to the unprecedented configurability that I’ve previously discussed.
I was reading through the assembly guide for the RepRapPro Huxley and it suggests that you first level the machine using shims of some sort and a spirit level, then use a plumb line to square the z axis.
I plan on re squaring my zaxis using this method and definitely will use it for the Huxley that I’m building with my friend.
You know, I have been thinking of the same problem. I always thought that typically it won’t make a difference to the printer since everyone seems to be printing well without squaring the rest of the axes.
@Whosa_whatsis , it is significantly less moronic than all the guides that state to simply use a plumb line and would require the whole building you are in to be level.
I just recently was reminded by a series of tall rectangular prints that my MG Prusa has squareness issues. In my case, I have the build plate parallel to the X axis, but neither is square to the Z axis. In fact, my Z axis is not perpendicular to the XY plane in either direction.
Carpenter squares and plumb lines are only a first step. After that, print two tall, rectangular parts. turn one 180° and place it back to back with the other. The difference between the two will be measurable as twice the out-of-squareness of the machine. Make adjustments as necessary.
I think this was inherited in part from the parent machine (I believe MG was using an UP machine then, if not now). The Z axis is overconstrained, as is the bad habit with RepRap machines. So when I built it, I was too preoccupied with getting things to run smoothly to remember to make sure everything was square. I currently have plans to slot the X ends to allow the top bushings to float, eliminating the overconstraint so I can make it run square.
I don’t think there is a way to avoid this merely by careful assembly, or by the use of any affordable hand tools. I think this will remain as just another calibration step for any new machine. I agree that adjustability should be designed into the frames for this purpose. I know that is how it’s done for manufacturing machine tools, even with the expensive manufacturing and measuring tools available to those budgets.
The mendel90’s design seems to ensure the Z-axis is square based on the interface between the back of the z-axis steppers and the bottom plate - and the printed parts that hold the smooth rods for the axis are printed in such a way that I think they overcome (at least minor) misalignment of the parent printer.