Just finished the Nomes yesterday and off they went to Santa Monica. These are some horrible photos but it will do for now so bare with me. These are four autonomous objects that all have identical programs and swing weighted arms from side to side for four minutes and off for one minute.
They are really bad at keeping time. And yeah physics eludes me. Too many variable to keep them synced but that’s not really the point. Instead it’s pretty phenomenal when they do all too briefly sync their motions. Can the difference of one gram of weight contribute to the unique behavior of an object?
I’ve been cranking out some work for an upcoming show I’m really excited about called Machine Muse at the Lois Lambert Gallery in Santa Monica, CA opening November 16. The work has to ship out on Monday so I thought I would take some ugly photos in the studio for now. The image above is the third incarnation of a piece I’ve been working on for awhile called Wirbel. This one has all new electronics, six arms, and is driven from wind speed data taken in LA in 2010.
Certain components are printed on the school’s Form1 SLA printer. The plastic parts are durable and look great. Crappy photos from my iPod in the studio so bare with me.
I’ll also be showing the piece that is now called Radiolaria. I received the massive order from Shapeways and assembled and programmed the five spheres the other day. They all working brilliantly.
The final piece I’m working on for the show is called Nomes. Yes my desk is that embarrassingly messy and no I don’t have any time right now to do anything about it. The prototype uses clear acrylic but I will probably use frosted instead so that you can see some of what’s going on inside but not everything. The grey outer shell is simply printed in PLA.
The plan at the moment is to have four of these guys hanging out on a shelf made of aluminum extrusion and frosted acrylic. Still working on the details.
Since the basic laws of physics seem to illude me, Im using a very monotonous repetitive moving arm that is out of sync with the speed that the forms move resulting in a net random movement. Or something like that. Im hoping to have the four objects occasionally sync up with one another. I do so very much hate the sound of servo motors but the movement can be pretty impressive. This piece uses a new Adafruit Trinket and a Hitec HS45 servo. Final images to come after the 15th.
The first NoCo Mini Maker Faire is happening tomorrow, Saturday October 5th in Loveland, CO. Our Nerdy Derby track, above, will be present with students helping kids build and race cars. We’ve also got a booth setup with some examples of student work from our digital fabrication lab and my wife’s letterpress shop, Green Bird Press, will be present in full force. I’ve also seen some R2D2s, pinball machines, neon art, weavers, Sparkfun, and Epilog plus 150 other booths. Should be a lot of fun!
Either for the home studio or for the University, I get the opportunity to try out a lot of different kinds of 3d printers. Here’s the new Ultimaker I picked up for the University to replace our old Prusa RepRap and hopefully provide a more reliable FDM-style machine for the students to work with. Yeah, good timing with the announcement of the new version…. This is the second machine I’ve purchased assembled–the first the Form1–so I was excited to see if this machine was as easy to get up and running.
The machine shipped with this nice little selection of test prints, which were at the same time impressive in quality but also promised some work ahead. Typical Ultimaker stringiness meant retraction needed some work and the tape on the printbed was both torn and the flat part of the large torture test piece had peeled up meaning the print bed was not at all level. Before I started digging around the interwebs to find proper settings for the new machine I thought I would see whats on the SD card and use the controller to print something right away.
The controller is a joke. I’ve avoided these in the past because I don’t see the value in them and actually using one proved it for me. It’s nice to not have to have your computer attached to the machine the whole time but the user interface is horrible and only barely usable. The new Ultimaker 2 has a much better display and UI. Oh well.
After poking around the SD card I found the Ultimaker robot so I gave that a shot. The first time it failed because of the printbed but the second time around I manually adjusted the z-axis to get the layer height right and here’s what I got. Although the file has some crazy g-code in there to knock the robot off the platform and start printing a new one in succession… scary when you don’t expect it. How about not including weird files on the SD card?! Figuring that the layer resolution was looking sharp I set about leveling the bed and found something that would be a good test for the Ultimaker, if not a little ambitious.
Here’s a Zeppelin Trireme from Arnold Martin sailing a sea of failed prints. Oh how things went wrong. To start with, getting the bed level was rather tedious. I also really don’t like printing on blue tape as it seems there is a small margin of error between squishing the layer too flat and the layer just not sticking at all. Then I had all sorts of problems with the motors overheating and shutting down mid-print.
To get things working I had to dive under the machine on more than one occasion to adjust the tiny little trim pots that control the current going to the motors. Oh and thanks Ultimaker for making those trimmers turn in the opposite direction to just about every other driver out there. Now, why would a printer shipped fully assembled and calibrated ship with motors that could raise a blister on your finger when you touch them? Oh and I also had to re-lube all the drive shafts with some PTFE grease to get things to move smoothly enough to work with the adjusted driver current.
After the miserable zeppelin attempt I went ahead and started on some upgrades like this new extruder drive gear from Felix to replace the pathetic lasercut drive gear that comes standard. Hopefully this will help the retraction and make some better prints.
Many who have used the Ultimaker call it the Lexus of desktop 3d printers but so far I am not at all impressed. For nearly $2,500 delivered, I think I rightfully expected plug and play performance but that is nowhere near what I received. Instead I spent my entire weekend working on a machine that I intend to demo to the students Monday night. I had considered purchasing one for myself, drooling more than a little bit over the new pretty version of this very machine but at another $200 that just is not going to happen. This summer I built an ORDBot for about $700 which, even though there are some things I don’t like about it, has produced much greater print quality in a much shorter amount of time. In the end, if this is normal for this kind of high-end machine then I seriously doubt that a completely plug-and-play 3d printer even exists in this current market.
Final test of the new piece now titled Radiolaria, for the upcoming show at Lois Lambert Gallery in Santa Monica in November. I think there will be a small group of five of these guys, each with a slightly different shell and behavior. The shell and internal arm is laser sintered nylon 3D printed by Shapeways. I really like the crispness of the white nylon and the movement is that much better with the lighter weight shell. Here’s a sample video:
Here’s the presentation for this morning’s Inside 3D Printing conference in San Jose.
Just picked up some of Form Labs new Simple Grey resin for the Form 1 at the university. This stuff is awesome! My first print didn’t entirely work because I tried to be smarter than the software and use some of the settings that I had used with the clear resin before. After going back to default print settings this is what I got. I also like to pack things tightly together so that all of the support bases connect – seems to improve the odds of success. So all of that becomes this:
Sean Huxter’s Treadbot v01 and a dodecahedron from Wouter Glorieux’s Platonic solids set printed at 0.05mm layer heights in about 2 hours and 25 minutes. The grey resin is much easier to cleanup and properly snaps off the supports. It doesn’t have the gooey uncured resin feel like the clear did. This is a much better resin and I am looking forward to working with it more in the future.
It started one day when the Chair of the Art Department sent me an email with the announcement of the new MakerBot Replicator 2. To which I replied something like, “You think that’s cool, then check this out…” with a link to the new Formlabs Kickstarter, also launched right at the same time. After a few conversations passing in the hallway, the Department quickly put some money down for what I believe would become the only SLA 3D printer on campus. While a few months past due, a package finally arrived in my office this summer so I naturally had to take the opportunity to drag it home for a week or two to see what it could do.
This machine is simply gorgeous. Clean lines and good formal design means this thing looks less like an Erector set and should find a place right at home on any designer’s desk. Right out of the box, connect power and USB; pour some resin in the tank; and then start printing things.
The software is reasonably easy to use and is fortunately now available in a Mac OSX flavor. A little clicky button and OLED display on the front of the box provides some nice feedback at the start and during a print.
The first job out of the machine was a nice little wireframe dodecahedron as I figured there was absolutely no way something like this and at this small of a scale would ever come out of one of my other FDM machines. The Formlabs laser and photopolymer based SLA system means much smaller layer heights – right out of the box at 0.05mm – and smaller feature size when compared to more conventional extrusion based desktop printers. The software generates a base and support structures that need to be cleaned up during post-processing.
When done the Form 1 is capable of some amazing things. This dodecahedron is straight out of the alcohol bath and snapped off the supports. No sanding, sandblasting, or other surface treatment. Blown away by this initial success, I started work on a new project that could put this printer through some full scale tests. Initial results can be seen in this post, and from these prototypes, further testing and comparisons needed to be made.
This is a section of the previous sphere printed on the Form 1 at 0.05mm layer height in clear resin compared with the same file printed at Shapeways in polished, laser-sintered nylon (SLS). While the white nylon looks, and is flawless, the Form 1 output is surprisingly good. The resin is a little more rigid than the nylon (a good thing), but you can still (barely) see the layers and the little nubins left over from the supports are annoying. Still, its beyond words what the machine can do in a very unfair comparison. For another odd comparison, we could look at a previous print made on one of my FDM machines.
Over the summer, while testing a new ORDbot printer and preparing for the Inside3DPrinting conference in Chicago, I made this copy of Ola Sundberg’s Zombie Hunter using 0.2mm layer height with Matter Hackers 1.75mm Grey PLA. It’s an awesomely detailed model and came out very well on this new and unproven machine. After testing the Form 1, I had the wild idea to see if I could take this, the largest single print I’ve ever completed, and print the smallest print I’ve ever made.
Here is the same model printed 9mm tall, at 0.05mm layers on the Form 1. I slightly modified the original model in MeshMixer to remove the base and round out his neck, making the head suitable of 25-32mm scale models. And while I am no model painter, I put my geek on and gave the finished print a little coat of some model paint just to see how it would look. I am absolutely floored that the cigar and texture in his beard came through such severe reduction in scale. With a little planning in the design stage, there is no reason the Form 1 could not easily be used for small scale model design. And again, good luck getting something like this out of even an industrial FDM machine.
So, this is the future of desktop 3D printing, right? Eh, no. What this machine is capable is truly impressive and I simply can’t wait to find out what my Jewelry students will do with it this semester. (No pressure guys!) However, there are still some problems with really fine layer heights (0.025 is notoriously problematic), leaving you with sometimes noticeable layering, and some of my early tests saw de-lamination between the first layers of an object. While the per layer print speed is ridiculously fast, the double whammy of fine layer heights and an excruciatingly slow peel process, to separate the cured resin from the bottom of the tank, performed after each layer combines to create some mammoth sized print times – in my testing ranging from 6 to 12 hours or more. Also, working with resin just sucks. I don’t care what you do with it – small scale casting, patching up an old boat, or printing with it – its nasty stuff, gets everywhere, smells noxious, and is a pain in the ass to clean up after. When compared to the syrupy goodness of printing with PLA on a cheap Printrbot, I have a hard time seeing the Form 1 in an elementary classroom.
And then there is the cost to consider. When you factor in $3,300 for the machine and $149 for a liter of resin this thing adds up. For the same print cost or maybe a little more, I can print my rolling spheres at 100mm in diameter through Shapeways in polished white nylon. Paying a little extra and waiting a couple of weeks means that I don’t have to deal with the resin cleanup or the odd artifacts in the print. Basically, I couldn’t justify one of these machines here at home. This printer, more so than any other desktop FDM machine that I’ve worked with, makes for a really good prototyping machine. Which means to say that chances are, the output of this printer is most likely not for a finished product. If for example you were a designer that needed to know if a design would work by the next morning before sending it off to a service bureau for a final print, then it would make sense. For my students, they don’t have much time in a 16 week semester to wait for Shapeways to get around to making their thing either. For the rest of us though, I’m not sure that I see much room for the role of desktop SLA printing in our day-to-day making for the time being.
Working on a prototype for a new project, or at least a remake of an older project anyway. This 10cm diameter sphere will be one of maybe seven if I’m lucky for an upcoming exhibition in Santa Monica in November. It is printed in photopolymer resin on a FormLabs Form1 desktop SLA machine. The quality is good but not great… layers are still visible, lines are not that crisp, theres a slight yellow hue to the plastic, where the supports attached there are little bumps, theres a little de-lamination in the print on the seam, and so on.
The two hemispheres are held together with three very strong neodymium magnets. (One of the magnet holders on the top in the back was filled with resin and had to be drilled out – breaking in the process.) Because “complexity is free”, each of the seven spheres will be customized with a different pattern of ley lines on each one. The board is a custom Arduino setup with an H-bridge for controlling the micro metal gearmotor and a charging circuit for the lithium battery. The arm attached to the motor swings around a lead weight acquired from the wife’s lead type collection to create the movement of the sphere.
The ball is overall pretty cool, and definitely not possible to manufacture on my other more traditional FDM machines, but Im not terribly happy with it either. The print quality issues, combined with the print time (16 hours total) and the cleanup involved after the print is done with the cost involved ($260 for 2 liters of resin at discount) on top of everything just has me rethinking it a little. If I’m going to spend some cash on it I might as well do something really cool like this example of polished alumide from Shapeways and not have to do the labor myself.
Option B is definitely not a cheap road to go down but it would be really nice…