I’ll never forget the day I first met MakerBot. It was August 1, 2012 when he*—a bright, shiny first-generation Replicator—arrived at our Cambridge, MA, office, greeted by screams of delight by a throng of fans. I must admit, I was a bit intimidated and star-struck: MakerBot’s reputation preceded him. He was a rockstar in the DIY community, a true maverick of a machine, ushering in the “Wild West of 3D printing” among our sedate sea of MacBook Air laptops running Adobe InDesign. All we had ever made here before were PDF files, but with MakerBot humming cheerfully in the lounge next to the kitchen, that had all changed. We were now maker-magicians, spinning ABS thread into gold.
At first, it was hard to get any quality time with MakerBot. I’d come into the office in the morning, and he’d already be surrounded by three or four groupies, who were browsing the catalog at Thingiverse, selecting a fresh set of STL models to print: from Mario and Batman to Mayan Robot.
But MakerBot didn’t just allow me and my coworkers to print out other people’s models; he offered us the promise of designing our own plastic masterpieces. He came packaged with the open source software ReplicatorG, which provides a nice GUI for doing simple modifications on existing models (scaling, rotating, etc.). ReplicatorG isn’t a tool for constructing models from scratch, however, so I also started experimenting with other 3D rendering applications like Blender, MeshLab, and OpenSCAD.
I was interested in the possibilities in transforming 2D photos into 3D models that MakerBot could print, so I started experimenting with a Python tool called img2scad, which can convert a JPEG image file into a .scad file (convertible to a compatible STL file with OpenSCAD) by transforming each pixel in the image to a rectangular prism whose height is directly proportional to how dark/light the pixel is. When this SCAD model is printed, the output is a photograph embossed into a sheet of plastic. Pretty cool—although, in practice, the results were somewhat lackluster since much of the detail captured in the subtle shading differences among pixels in the source JPEG didn’t get preserved in the conversion to prisms.
I wanted to take things to the next level and actually make 3D replicas of photographs, which is possible by hooking up an XBox Kinect to a Windows machine and using ReconstructME to do a 360-degree scan and convert to STL. I volunteered to be the first guinea pig of this process, which entailed sitting motionless in a swivel chair while one of my coworkers spun me around in a circle at a rate of 0.0000000001 miles per hour, and another coworker aimed the Kinect at my head.
Here was the result:
My plastic doppelgänger looks just like me (white and inert), except he has no eyes or mouth.
Anyway, six months into his residence at the office, MakerBot had become my new BFF. But then a series of unfortunate events occurred that started to test the strength of our bond. One of the qualities I prize most highly when entering into a relationship with a piece of machinery is its dependability, and as time wore on, MakerBot just kept letting me down again and again and again. Here were my three main gripes.
Gripe #1: MakerBot was wicked** slow and glitchy
We live in an age of instant gratification (“You mean I have to wait five seconds for that web page to load? Screw dat!”), and MakerBot just took too @$(&@# long to complete a print. Waiting five hours for your Yoda feels like an eternity; you can play approximately sixty rounds of Candy Crush Saga in that same timeframe (although arguably, staring blankly at the MakerBot is equally intellectually stimulating).
To make matters worse, I’d estimate MakerBot’s failure rate fell in the range of 25%–33%, which meant that there was around a one-in-three chance that two hours in, your Yoda print would fail, or that it would finish but once it was complete, you’d discover it was warped or otherwise defective.
Some of these glitches could be attributed to human error. For example, you forgot to configure the Energy Saver setttings properly on the laptop feeding data to MakerBot via USB, and the laptop went to sleep, causing MakerBot to time out as well. Or you neglected to set the temperature of the MakerBot platform hot enough, and partway through the print, the plastic stopped sticking to it, potentially causing the model to tip over and/or MakerBot to spray a bird’s-nest jumble of plastic thread over it. But other glitches were less foreseeable and preventable, such as molten plastic getting gunked up in the extruder midway through the printing process, resulting in the MakerBot nozzle dancing futilely above your half-finished creation.
Overall, MakerBot was rather finicky and unforgiving when it came to preparation and configuration for printing, which segues nicely into my next gripe.
Gripe #2: MakerBot was wicked high-maintenance
Because of the relatively high chance of errors (see Gripe #1), my colleagues and I developed our own rigmarole of preflight checks to try to mitigate the likelihood of #makerbotfail on any given print. These steps included:
- Use ReplicatorG to preheat the extruder to be used to ~235 degrees Celsius.
- Detach the filament guide tube from the extruder and remove the filament from the tube.
- Manually apply pressure to the filament to push it through the extruder.
- If plastic thread is emitted from the extruder nozzle, proceed to step #5. Otherwise, turn off the MakerBot, disassemble the extruder apparatus, clean off any plastic buildup from the components, reassemble the extruder apparatus, and start over at step #1.
- Hooray, plastic is being extruded from the MakerBot nozzle! Now see if the MakerBot can automatically extrude plastic all on its own, without you jamming it through the machine. In ReplicatorG, click the “forward” button next to “motor control” to initiate a feed through the extruder. If plastic does not feed through automatically, return to step #3. Otherwise continue on to step #6.
- Click “Build” and pray.
We discovered that by going through this preflight process (which takes anywhere from ten to ninety minutes to complete), we could decrease the likelihood of a printer failure by one or two percent.
Gripe #3: MakerBot just plain breaks
After we used MakerBot for about three months, he stopped squirting plastic from his right extruder, no matter what pre-print precautions and preparations we took (see Gripe #2). After examining the extruder mechanism and doing some research, one of my coworkers pinpointed the problem: the Delrin plunger was worn out.
The Delrin plunger is a small, black cylinder whose purpose is to apply pressure against the molten plastic thread to help force it through the MakerBot extruder nozzle. However, apparently this pressure gets applied via a grinding process, which slowly erodes the plunger material until it eventually is no longer able to make contact with the plastic.
But luckily, the MakerBot store offered replacement Delrins for sale, and they were only $6, so I just ordered another one. This served us well for another three months or so, until the Delrin plunger in the left extruder wore out as well. I went back to the MakerBot store, but this time I found that the Delrins were out of stock (and they continue to be unavailable to this date, possibly because first-generation Replicators are no longer being sold).
I was really bummed. One of the two Delrin plungers in the $2,000 MakerBot was busted, the $6 replacement parts were no longer available, and eventually the other Delrin plunger would wear out, and I’d be left with the equivalent of a fried toaster.
But then I had an idea: the Delrin plunger was just a chunk of plastic. What if MakerBot could print a replacement plunger, and regenerate itself back to health like a starfish? And sure, enough, there was indeed some good news: Thingiverse came to the rescue with a model for a plunger replacement. But then some bad news: the replacement plunger just didn’t work. Whether that was because the STL model was not an accurate replica of the bona fide Delrin plunger, or because the actual plunger that MakerBot printed was defective (see Gripe #1), I don’t know. Either way, we were out of luck.
I did some research into some more elaborate contraptions MakerBot could print as replacements for the extruder apparatus, but they required screws and springs and other stuff, and upon further reflection I realized I was not cut out to be a MakerBot repairman.
Not long after the second Delrin plunger failed, I decided it was time for MakerBot and me to go our separate ways (the majority of my colleagues had already jumped ship many weeks prior). Nearly every time I would attempt to print something on MakerBot, I was met with frustration or disappointment because of some snag or another. Dealing with MakerBot was time consuming and depressing, and I decided I just didn’t need that kind of negativity in my life. I deserved better than that!
But I really miss making those plastic tchotchkes. I’ve considered investing in a Replicator 2 (the follow-up to the first-generation MakerBot Replicator), but right now, it still feels too soon to be getting involved with another 3D printer.
I was excited to hear last week of Stratasys’s $400 million acquisition of MakerBot Industries. I hope this move presages further investment in creating high-quality, low-cost 3D printers for the average consumer who wants a reliable device for printing models that doesn’t require a DIY approach to maintenance and repair. The first-generation MakerBot Replicator felt too much like a prototype, as opposed to a proven, refined piece of hardware.
I look forward to the day when 3D printers are as cheap, ubiquitous, and easy to use as their 2D inkjet printer counterparts. But for now, every time I pass by MakerBot, abandoned in a corner of the office lounge, I feel a slight pang of guilt and regret. I’m sorry we couldn’t make it work, MakerBot, but I’ll always remember the great times we shared.***
* Our MakerBot’s name is Rob Roboto, and since gender-normative naming is de rigueur for 3D printers, I have concluded he is male.
** FYI, in Boston, we use the word wicked as an adverb—in place of very or really—to indicate emphasis.
***Since I originally published this piece on Medium, MakerBot support has reached out to me and has suggested we try using the Replicator 2 Drive Block Hardware Kit in place of the Delrin-based extruder apparatus. They have been kind enough to offer to send us a couple free kits, and I’m open to giving them a try.