baby’s first 3D printer
[This is the first in a series of notes about home-based 3D printing based on my experiences with a MakerBot Cupcake.]
In the 1980s the average person didn’t own a home computer. Those who did were likely to be gamers, hackers, tinkerers, or someone else interested in owning a computer as a hobby, not as an everyday tool. ~30 years later, computers are a part of everyday life, used for paying bills, keeping up with friends, publishing photos, and a whole host of other uses we could never have predicted back in the days of the SE and AT. We knew that home computers would change things, but we couldn’t predict how, no matter how many episodes of Star Trek or Max Headroom we watched on multi-generation VHS tapes copied from friends.
Today, 25 years after the Great Pagemaker Massacre of 1985, we’re on the verge of another massive change in how our world works. I have no idea how that change will manifest itself, but I’d like to be one of the first to find out.
I just built a MakerBot Cupcake 3D printer, which is itself based on the reprap project printers. Since the first question most people ask me is “how much did it cost?”, I’m going to start off this series of notes talking about the economics of 3D printing.
In raw dollars, the Cupcake cost a little less than my first computer, a Commodore C64 with monitor, printer, and omfg, floppy drive instead of cassette recorder, all of which set my parents back a bit over a grand. While a grand or so in the early 80s bought a fair bit more than it does now, like other home computers, you couldn’t just buy the computer. We probably spent another few hundred dollars on software, joysticks, blank floppies, that weird “computer-paper” that the printer used and so on. Most of those things came from third parties, so there was competition to keep the prices down — you weren’t locked into buying blank floppies only from Commodore.
Like the C64, one of the selling points for the Cupcake is that it’s a cheap, no-frills device. Part of the fun in having a Cupcake is the DIY aesthetic of figuring out how it works, why it works, and how to keep it working. Another not so obvious selling point, is that the Cupcake is based on opensource software and hardware. If you’re not familiar with the 3D printer market, you’re probably thinking “so? I bought a cheap PC built from parts and run linux? What’s the big deal about an opensourced 3D printer?”
Commercial 3D printer companies, like most 2D printer companies, operate by selling you the “razor for cheap then making it up on the blades”. The profit isn’t in the printer, it’s in the supplies the printer uses and the support contract to keep it running. Next time you see a really inexpensive inkjet printer for sale, research the cost of a set of replacement ink cartridges. Compare the volume of ink in the cartridges and their price and compare that with the price of refill ink, or look at the effort some manufacturers put into forcing you to only buy new cartridges by using DRM. (There’s an excellent eBay scam that takes advantage of the pricing disparities: buy a printer, pull the ink cartridges, then sell the printer “like new” for near what you paid for it to someone who doesn’t know how much the replacement cost of the cartridges.)
Two things you usually have to buy from the manufacturer if you own a commercial, closed-source 3D printer are the material to print with and the base that you print on. The printing material is probably a spool of ABS plastic in a vendor-specific housing and the printing base is also ABS and also vendor specific. There’s a nice article over at Time Compression that goes into cost details to be considered when buying a commercial 3D printer, but we’ll skip to the chase and say we’re talking about US$ 1-2 per cubic inch on the proprietary systems vs. USD $10 per pound of raw ABS from MakerBot. Oh, and instead of those $5 one-use print surfaces only available from the vendor, the Cupcake prints on a variety of surfaces available at any art supply store, some of them reusable for dozens of prints. (I’ve used a small piece of acrylic for ~20 prints on the Cupcake with no signs of wear and tear.)
This is opposite to how 2D printing has worked going back to the earliest days of printing. Once someone had the idea to cut blocks of wood or cast lead as type, the printer could control costs by simply buying raw materials for the best price they could negotiate and recycling them when possible. Cast some metal into type, then melt it down when you no longer need it. Screw up a print run? No problem, we can recycle that paper. Wore out your wooden printing block? Have someone carve another and get back to printing.
When I learned to type (“yes, grandpa, on a typewrier, we know”) it was on an IBM Selectric that used a ribbon and “typewriter” paper. The ribbon was sold by IBM, but replacements were available from third parties. Likewise, I didn’t have to buy my paper from IBM, I could buy it from any office supply store. I could even buy paper that IBM didn’t approve of (as if such a thing existed). If my typewriter needed repair, I didn’t have to call the IBM tech, I could go to any typewriter repair shop I choose.
This is pretty much exactly opposite to how 3D printers work now. If you own a FooCorp X1000 you are pretty much locked into buying everything from FooCorp. Having problems with your X1000? Is your support contract paid up? Are you allowed to even open it and try and fix it yourself without violating your contract?
While the Cupcake is opensource, and one is not locked into buying ABS from MakerBot, it isn’t a completely self-sustaining ecology just yet. The first problem is that there’s no way to convert ABS models and scrap back into spools of ABS for printing . The technology to melt and extrude ABS plastic is there, it’s just a matter of someone building a melter/extruder that’s safe for home use. Safety might end up being the real problem as ABS fumes aren’t something you want to breath on a regular basis. Instead of recycling ABS on the individual level, perhaps the local door-to-door ABS recycling firm comes by and trades your scrap for fresh rolls of ABS, similar to the newspapers for toilet paper biz in Japan. One step further would be the ability to take broken ABS items and recycle them into replacement parts. If a knob or some other small part breaks, bring it over to my place, I’ll print you a new one then give the old one to the recycler in trade for more plastic.
So there you have the costs — under a grand and a dozen or two hours of your time to assemble it, adjust it, and get it running. Some of the money you’re “saving” by buying a DIY printer is going to be translated into hours of your time assembling, adjusting, and generally tweaking your Cupcake to get a decent print.
Next we’ll look at who the real customer is and whether you should buy a Cupcake or just ship your STL to RedEye.
Excellant article
Thank You,
~cw
Comment by C Wright — 2010/02/18 @ 03:30
Two points to consider besides cost are quality and strength. All of the DIY printers prints are poor quality with the best surface quality leaving rough stepping artifacts of at least .010 or greater. If your looking for a quality surface this is not acceptable. The powder based systems are simply too fragile. Another consideration is size. Most DIY system prints are tiny.
Bob
Comment by Robert Cunningham — 2010/11/27 @ 00:33
I think that “poor quality” is both subjective and temporary. The question is if the print quality is good enough for the task at hand not if it is the best print possible. If I need a tighter finish I can prototype something on my Cupcake then send it out to a firm like Shapeways for a high-end print. Meanwhile, I’ve printed a number of objects that were perfectly functional at the low resolution of the Cupcake.
It’s really no different than printers. In the 80s most of us could only afford dot-matrix printers and couldn’t print anything nearly as nice as a professional print shop. 20 years later, color inkjets were cheaper than computers and the average person could make photo-quality prints at the push of a button.
Comment by jet — 2010/11/27 @ 07:20