ebooks and download videos
Home / Articles

Created: 10/26/2012
Updated: 01/23/2022

The Cube 3D Printer: an independent review

by Chris Wasshuber

I bought a Cube 3D printer a few weeks ago and have printed with it almost every day. (If you want to see one of the projects I did see my article on Cy Endfield's Chess Set.) Here is my take on the printer, the good and the bad. I am in no way affiliated with 3D Systems the company that produces the Cube. I am an engineer for 30 years and have gone through formal metal workshop training as teenager, later migrated to electronics, and am today mostly active in software. But I have kept my interest of building and making things. During 2006 I have had some exposure to 3D printing at MIT Media Lab where I took Neil Gershenfeld's famous class "How to build almost everything". At that time 3D printers were very expensive, cumbersome to use, and for the most part I avoided using them.

Today in 2012 3D printers can be had for literally a few hundred dollars if you are willing to assemble your own printer. See for example the RepRap or Fab@home movements. I decided to buy the Cube at $1300 simply because I wanted to focus on printing and not on assembling a kit and then having to adjust and tweak my printer. It was a decision of convenience. My primary focus was on learning to design in 3D. The printer was mostly the motivation to do that, because printing out your designs and holding them in your hands, showing them off, is way cooler than just doing it on the computer. It would be like learning to program without a computer. It is possible, but it is much more rewarding to actually execute your program on a real computer, rather than do it on paper alone.

Ok, enough of my background and motivation to buy a Cube. How good is the Cube? Where is the advertisement wrong or misleading? Where is it correct?

In short, the hardware is great, the software sucks. I was up and printing in about 30 minutes after unpacking the Cube. Instructions were clear. The printed objects were marvelous. The ABS plastic is very sturdy. The objects were stronger than I expected. The surface quality astonishingly good.

My first disappointment was that after a few days of printing the print head stopped feeding plastic filament and even after all kinds of attempts to get it going it didn't work. However, 3DS sent a replacement via FedEx and the new print head has been working ever since. I think their customer service has handled this problem well. While I would have preferred a working print head from the start, I do understand that these plastic print heads tend to have issues even in much more expensive machines. This is a sign that the technology is still young and some of the bugs have not been worked out. One has to expect these kinds of problems for the time being.

Let's come to the problems. The biggest issue or limitation the Cube has is in its software. For example, you cannot set printing with 100% fill. The Cube always introduces empty spaces inside objects. While this is great for some models, because it saves on plastic filament, it is utterly useless for structural components, because they are generally too weak for real world use. As long as you only want to print figurines, architectural models, objects that do not need to withstand any significant amount of force you are fine. All others stay away from the Cube. For me this is a big problem, because I was planning to print structural components for all kinds of projects, for example print special parts for the Vex robotics system. I really hope they fix this soon. But the response from customer service does not instill a lot of hope in me.

Another major problem for me is that the Cube changes certain dimensions without good reason. For example, if I design a piece with a thickness of 1mm, the Cube makes it a bit more than 2mm. With a print resolution of 0.25mm there should be no reason to go from 1mm to 2mm. If I design the thickness to be 2mm it ends up about 3mm. If you now think it always adds 1mm you would be wrong. I have not yet fully figured out what it does. Or say we design a plate with a hole of 10mm diameter. The hole ends up with a diameter of about 8.5mm printed. Why? There is no reason to change the dimensions by that much. Particularly mystifying is the fact that the outside dimensions of the plate are pretty much as designed. So the Cube somehow reduces open spaces. To me as software engineer this looks like a bug in the G-code creation. I sent customer support files and descriptions but they have not come back with an explanation or a fix. Well, the work-around is to design the object such that after the changes the Cube introduces it ends up with correct dimensions. The problem of that is that it requires a lot of trial and error, lots of failed prints before it comes out the way you want it. In this case an open platform like the RepRap is a boon, because you have a chance to fix bugs yourself or perhaps find somebody else who is willing to fix it for you. With the Cube you are stuck with what 3DS does. So far they have not done any improvement or revision of the software.

There are other issues with the software which are not that critical but clearly point to the fact that they have not tested, or designed it very well. For example, the tool path is very inefficient. When I printed out a larger set of components in one go the print head was sort of randomly moving left and right and up and down making a lot of empty movements rather than going in some systematic pattern keeping tool paths to a minimum. I think print times could be reduced for more complex prints.

Filament cartridges are very expensive. Yes, the cartridges are convenient. It is fairly easy to change from one color filament to another. It also keeps all the filament nice and neat. But you are paying about $150/kg of plastic filament. The usual going rate is ~$50/kg filament, and the raw material cost is about $5/kg. Maybe some clever chap figures out how to refill these cartridges. With a markup of 30x there is a lot of room for clever entrepreneurs to make a killing by selling at lower cost.

And finally here is a one point where the ad copy is at least misleading if not outright wrong. It states that it uses the same material for support structures to allow printing of overhangs and that these supports are easy to break off. Well, they are NOT easy to break off. And once broken off the result looks really ugly. You would then have to clean it up with a knife or sand paper which can be difficult depending on the object. In my opinion this defies the purpose of a 3D printer if one has to do all that post-processing. The reality is that you cannot print out objects with overhangs of less than 45 degrees. Anything that builds out into the open at 45 degrees or steeper is fine. At the most you can bridge short distances like an arch with overhangs less than 45 degrees, but the printer typically distorts these and makes other errors.

With everything I now know I am a bit on the fence recommending the Cube. I think it is a fine first 3D printer if you do not have any specific requirements or anything specific you want to print out, which may or may not work well on the Cube. But if you simply want to learn more about 3D printing, print out some toys and play around with the tool, it is an ok entry printer. If you can wait it is probably better to wait and see what else comes on the market, or simply wait a year or two until some of the bigger problems have been worked out.


Christoph Wasshuber (07/02/2014)

On my 1st generation cube one can achieve the same without this device. Simply leave an original cartridge in the machine but feed a filament from a different spool into the print head. This way the machine thinks it has an original cartridge, which it does, but filament is consumed from a different spool that can be any manufacturer.


CUBE3DFREE device alllow to use alternative materials on printers CUBE 2 and CUBE X from Cubify/3D Systems ... 10x lower price for material ....