The Form 2: From Prototype to Production

In the previous post in this series we focused on how we torture tested the Form 2.

The Form 2 is the most advanced desktop 3D printer ever created. But as the saying goes, you don’t build the most advanced 3D printer in just one day. In fact, it took more than a half-dozen complete system iterations over the course of 18 months.

A peek inside the Form 2

We believe in sharing some of our experiments and learning as we built the Form 2. As engineers and designers, this is what we would want to see. In this post, we intend to pull back the curtain on our prototyping process and walk you through how we iterated toward the Form 2 design.

Our iterations took the form of complete system prototypes. Some iterations solved particularly difficult engineering issues; others focused on industrial design, while some of the last iterations focused on manufacturability, fit, and finish. Every version had one thing in common — they were all working printers, and each iteration built toward the final product we shipped.

Proto 0 — Hacking the Form 1

Goal: Explore a new print process and resin system.

Right from the beginning, we knew we wanted to focus on print reliability. To see if we could make that happen, we built mock-ups of a new printing process. The first prototypes used one of our most plentiful resources, Form 1 3D printers. Each engineer received a Form 1 to hack with the goal of improving on the peel mechanism in the Form 1+. The ideas were wide-ranging and involved modifications from attaching some 80/20 aluminum extrusions to the printer, to complete rebuilds of the structure and internal mechanisms.

An early slide mechanism
A rotational peel mechanism

It was also while building Proto 0 that we began to test ideas that we wanted to be key features of the Form 2, for instance, an automatic resin management system. For inspiration, we looked to 2D printers, tearing ink cartridges apart trying to understand the decades of engineering wisdom and learnings that have gone into designing those systems.

An early prototype of our gravity fed resin dispenser

Proto 1 — (Occasionally) Works Like

Goal: Build the first working prototype with the new Form 2 architecture

In the first few months of prototyping you wander, explore, and search for the best ideas. But eventually you have to start to converge. As with most engineering decisions at Formlabs, we used the process of elimination to select the core architecture concepts most likely to survive.

Proto 1 was the first attempt to put together a machine that would become the Form 2. Systems such as the wiper and the slide mechanism were first used in action with this prototype. One early simplification and exploration was to have the slide peel displace the whole length of the tank as can be seen in the photo below. Proto 1 was the last prototype that relied heavily on legacy Form 1+ components, including the build platform and the “brain” components.

Proto 1 hard at work printing

The wiper went through numerous iterations during Proto 1, getting swapped out every few days for a new design. This component was an integral part of our new printing process and it needed numerous improvements. We went from a top-down insertion wiper to a design that is closer to what is shipping today. The design is both insertable and preloaded against the silicone tank surface.

Evolution of wiper and wiper insertion design

Prototype 2 — Prints Like

Goal: Create a machine that could print parts not possible on the Form 1+

We learned a lot from building Proto 1. We learned that the sliding peel mechanism only needed to move a portion of the full width, and we added a separate drive mechanism to decouple the wiping from the sliding motion. At this point, the wiper was showing strongly positive results that would improve overall print reliability.

First version of the new build platform with the cam handle on the carriage.

By now, the system was getting more complicated. With the wiper and cartridge actuators, in addition to the slide axis and Z-axis actuators, we were up to 4 motors. We also wired in more limit switches and our first passes at resin level sensing and temperature sensing. Our kludge of Form 1 electronics was becoming complicated with a network of laptops and phidgets helping to drive the system. This meant it was time to build our first Form 2 electronics board.

As we developed the first electronics board, we could finally begin to picture the shape of the machine. It was time to start thinking about how large the printer might be and make some crucial decisions before we went any further. One key interaction that affected the shape of the machine was where and how the user would insert the resin cartridge.

Variations on cartridge placement — in the build volume and the final “backpack” concept. Note the mini printer in between the big prototypes.

Proto 3 — Feels Like

Goal: Develop a “feels like” version of the final machine.

After a few weeks, a fully functional Proto 2 s started to look like a Frankenstein 3D printer. Proto 3 would integrate our learnings into the package we wanted. Certain systems began to stabilize while others remained in flux. To get us to our Proto 3 goal, we needed a design. But making a design that is instantly recognizable while being efficient, manufacturable, maintainable and as functional as possible is a delicate task.

Early industrial design sketches for the Form 2
A couple of Form 2 concepts that never made it

We migrated the electronics into the body and added a display to the printer. An extrusion that was designed as the rigid backbone of the printer became housing for the resin cartridge system. Cosmetic shell work was also done at this point in the process.

Proto 4 — Looks Like

Goal: Get as close as possible to the final design to prepare for manufacture.

By the fourth prototype, the rate of change hadn’t slowed down (if anything, it was still speeding up) but core print mechanisms only needed small tweaks. Proto 4 was the last prototype before we committed to tooling, so we aimed to do all testing with high-fidelity machines

Final industrial design of the Form 2

Proto 4 was also when the software development became instrumental to success. PreForm versions designed for the Form 2 were shared internally and a growing embedded software team developed the software forWi-Fi, responsive touch screens, interlocks and more.

Proto 4.4 — almost there, towards tooled parts

Proto 4.4 was the last prototype before we started ramping up our manufacturing machine. The goal was to have few mechanical differences from Proto 4, but all of the parts were produced by their final suppliers as they would be for mass production. Injection molds, extrusions, and die casting tools were built, first samples were shot, and first parts were assembled. The Proto 4.4 result look a lot like the machine we’re about to start shipping, but many of the parts are lacking proper finishes, are warped, or out of dimensional spec. For example, the injection molding tool for the orange polycarbonate cover requires a high polish. In the picture below, you can see the cover is still rough. It was important to test general fit and shape before we committed to this costly and time intensive step.

A fully functional Proto 4.4

From Many Prototypes to One System

We hope this behind-the door look at our development process gives you an idea of some of the thrills and challenges we faced in building the Form 2. The design process can be difficult for a machine with as many interrelated systems as a stereolithography 3D printer, but we figured out a prototyping process that got us to a successful finished product. We did this by working on our most important goals early, iterating on our successes, always setting new objectives to reach for and most importantly, constantly pushing for integration.

Stay tuned for the next episode, where we are going to talk about our early release program and beta testing. And if you found this article interesting please let us, the engineers at Formlabs, know what else you would like to hear about Form 2.