74 Seconds: an engineer aims to break the 3D printing speed ​​barrier

Jan Roetz, an independent engineer and YouTube creator, has just set a new record in 3D printing. His self-built printer, with a name that is anything but accidental, printed a 3DBenchy model, the widely used test boat, in just 74 seconds. For comparison, an average FDM printer needs anywhere from 30 minutes to over an hour for the same task. Minuteman needs just over one.

This is not a lucky result or a one-time demonstration. Jan worked on this achievement for over a year. More than 365 days of consistent engineering work, dozens of iterations, and solutions to problems previously considered insurmountable physical barriers of FDM printing.

Who Is Jan Roetz and Where Did This Project Come From

Jan Roetz operates under the brand Roetz 4.0 and has been creating content about CNC machining and 3D printing for years. He is not an academic researcher or an employee of any corporation in the additive manufacturing industry. He is an independent engineer with a workshop full of lathes, milling machines, and printers, treating them all as a single working environment.

The Minuteman project grew from one specific question: how far can FDM printing be pushed if you stop respecting conventions? Jan searches for the answer publicly, documenting every step on his YouTube channel. Every iteration of the machine and every problem that arises becomes another episode. It lets viewers watch alongside him as a real engineering process unfolds in practice.

The goal he set for himself is simple and specific: print a Benchy in under one minute. Hence the name of the printer. Minuteman.

SpeedBoatRace, the Context Behind the Record

To understand the significance of a 74-second result, it helps to know what SpeedBoatRace is. It is an unofficial but widely respected competition within the 3D printing community, with the goal of printing a 3DBenchy in the shortest possible time. The rules are strict: PLA filament only, a defined layer height, model dimensions must be maintained within acceptable tolerances, and the entire print must be recorded with a visible timer in the frame.

Benchy is not a random model. It was designed specifically as a test of a printer’s capabilities. It features overhangs, bridges, a cylindrical chimney, and fine interior details. Each of these elements reveals a different weakness in the machine or the material. That is exactly why it has been the standard benchmark of the FDM community for years.

Jan took the competition literally. And started from scratch. The project did not begin with a record. It began with a Benchy in five minutes, gradually working down to lower and lower times, ultimately reaching the 74 seconds mentioned above. Each new iteration of the machine solved one specific problem that had emerged in the previous version.

Solving One Problem at a Time

In the beginning, the biggest bottleneck was the speed of the toolhead movement. Classic CoreXY printers move the printhead along the X and Y axes while the bed drops along Z. This works well at standard speeds, but at extreme acceleration, the heavy toolhead begins generating enormous inertial forces that literally tear the model apart during printing.

Jan decided to reverse this logic. In Minuteman, the printhead stays stationary. The bed moves, across both horizontal axes simultaneously, driven by four motors mounted at its corners. This means the heavy, multi-input extruder does not need to be violently accelerated and decelerated with every move. Inertia drops dramatically and movement speeds increase.

When Movement Is Not Enough

Once the movement problem was solved, another emerged: material flow. A classic single-nozzle extruder has a physical limit on how much molten plastic it can push through in a given unit of time. At very high speeds the print becomes under-extruded because the hotend simply cannot melt filament fast enough.

Jan designed a printhead with four BondTech extruders coupled together. Four strands of 1.75 mm filament feed into a shared melt chamber, and from there the material exits through a single nozzle. These are not four independent nozzles printing simultaneously. They are four inputs into one reservoir, which multiplies flow rate without requiring any changes to the slicer architecture. Simple in concept, but very difficult to execute.

Once that was solved, the cooling problem appeared. With a layer time of 0.4 to 0.6 seconds, PLA does not have time to solidify before the next layer is deposited. The model deforms, loses its dimensions, and in extreme cases melts from the inside out.

In response, Jan designed a circular air channel acting as an air knife, surrounding the nozzle from all sides simultaneously. But compressed air alone was not enough. He added dry ice to the system, cooling the air below zero before feeding it into the channel. Colder and denser air removes heat far more effectively than a standard cooling fan. The result of all these solutions combined: 74 seconds and a model that holds its dimensions.

How Minuteman Works, the Specifications

Minuteman is a machine built differently from anything available commercially. The motion architecture is based on a moving bed rather than a moving toolhead. The bed moves along the X and Y axes, driven by four independent motors. The toolhead moves only along the Z axis. This means the mass of the moving element is significantly lower than in the case of a full toolhead with multi-input extrusion.

The bed rides on an air bearing rather than mechanical rails. This eliminates friction, play, and vibration that at extreme movement speeds could undermine positional accuracy. The extrusion system consists of four BondTech extruders feeding material into a shared melt chamber. Effective flow rate is many times higher than in a standard hotend, without requiring custom profiles in the slicer.

Cooling is handled by a circular air channel with dry ice in the circuit. The air is chilled below zero and delivered to the nozzle from all sides simultaneously. The material used in the record print is PLA, in accordance with SpeedBoatRace rules.

The Goal Ahead and the Broader Significance of the Project

Jan does not hide the fact that 74 seconds is not the finish line. The goal remains unchanged: under one minute. The full technical documentation for the project is available for free on his Discord, and his Patreon exists solely to fund the next iterations of the machine. His work can also be followed on the YouTube channel mentioned above.

It is worth looking at Minuteman through a wider lens than a single record. The history of 3D printing is full of solutions that started as hobbyist experiments and ended up as standard features in commercial machines. Pressure advance, now built into every popular slicer, was for years a feature found only in advanced home-built rigs. Input shaping, which allowed Bambu Lab to print at speeds unavailable to competitors, also traces its roots to academic and hobbyist experimentation.

Quad extrusion with a shared melt reservoir, dry ice cooling, air bearing motion. None of these things will show up tomorrow in a printer from a shelf for a thousand zloty. But all of them could get there within a few years. And by then, no one will remember that it all started with one engineer and a boat printed in 74 seconds.