I made some successful, but small prints on my new 3D printer, then started to become ambitious. I wanted to print a really huge Stanford bunny. Everything looked great for the first several hours, but the print failed. This was a bit heartbreaking since I thought I had finally put all of the problems with my first printer behind me. It turns out (as usual) that there was a good explanation for what was happening.
My Z stage was failing to lift the extruder at some particular point up the Z axis. I knew that the system was challenged lifting the extruder and electronics out on a long level arm using a linear stage not intended for vertical lifting. Putting one finger on the stage as it tried to lift could cause missed steps. I determined that this was the problem when I noticed pieces of insulation and Kapton tape from the extruder stuck in the print around the area where it failed.
One solution that I tried and later rejected was adding a counterweight on a pulley to reduce the burden on the Z stage. This seems in principle to be reasonable, but I did not like depending on a shelf above the printer to hold the pulley.. My compact printer was suddenly not very compact anymore. Ultimately, this did not provide a reliable solution. I could still stop +Z motion pretty easily.
The better solution that I implemented was adding a second Z stage behind the first, and mechanically connecting the two stages together. Two motors lifting the stage is of course more powerful, but the back-to-back linear bearings also significantly reduced the play in the lever arm holding the extruder. Unfortunately, friction is also significantly increased, so without other changes to the system, this change in itself still did not fix the problem.
Ultimately, and I guess obviously, the one change that can overcome an over sized load on a linear stage is to reduce its speed. Since slowing down the Z axis has a very minimal effect on print time, there is no real penalty to turning this speed down significantly. I also changed the Z stage to half step and adjusted the firmware accordingly.
Once I made this change, it was actually a challenge to make +Z motion fail by pressing down on the stage as it moved. In the end, I had a much more rigid apparatus holding the extruder, and confidence that it’s motion would not fail.
To save time, I printed a standard size Stanford bunny.