There is something very satisfying about a tiny CoreXY machine moving this fast. Fail Fast! has a Voron V0.1 running at 1,000 mm/s, with the first layer at 250 mm/s and acceleration set to 50,000 mm/s². The toolhead, in its orange housing, sweeps across the bed in rapid test moves without extruding, and at these speeds it becomes a blur above a dark hexagonal print. Layer height is 0.1 mm with a 0.4 mm nozzle, which makes those velocities even more striking.
The machine itself is the familiar small-format layout: black aluminum extrusions with clear acrylic panels, and the visible linear rails and belts of a compact CoreXY. On the toolhead sits a Dragon HF with an aluminum Volcano heatblock, a CHT nozzle, and a custom spacer with a blade in the middle. The motors run on 48V, the filament is eSun ABS+, infill is set to line, a pattern the maker found among the quickest, and the fans are managed manually for now.
The resonance work is what stands out to me. Early Y-axis testing found peaks at roughly 25, 80 and 120 Hz, and that spread forced the use of 3 hump EI, whose smoothing noticeably changed the geometry of parts. After stiffening the frame this morning, testing shows one peak near 80 Hz with 0% residual vibrations and 0.03 smoothing at 17k acceleration. The maker calls that a huge win for print quality at speed, and the approach of adding stiffness and weight to move the peak heights and positions seems like a smart way to attack the problem.
Where it stands now: the print still looks bad, even though no steps are skipped and flow and part cooling are sufficient. The whole effort is a work in progress aiming for usable output at these speeds, with a plan to rerun the exact same gcode to compare before and after. The specific stiffening ideas have not been described yet, so that part remains an open question.
Project by Fail Fast!. Watch it on their YouTube channel: