Makerbot's big problem was that their machine sucks. Low-end 3D printers just aren't very good. The weld bonds between layers of ABS filaments are too weak, and the objects break easily. They often break during building, just due to thermal expansion. 3D printing requires better temperature control of the process than Makerbot ever achieved.
TechShop used to have some MakerBots, but they've been replaced by better machines. Materials costs are too high, though; the Form I and Form II use a working fluid that costs $130/liter. (This is a bit like the inkjet printer ink problem.) They also go through build tanks fairly fast; the working fluid in its solid form builds up at the bottom. But the build quality is excellent.
The "maker revolution", such as it is, is driven by laser cutters. They're accurate, fast, reliable, work on large sheets, and have no consumables.
The Next Big Thing is supposed to be desktop waterjet cutters. The Wazer, though, is apparently very slow, and may use more garnet per cut than the big waterjets. Wazer also glosses over the problem that waterjets generate a sludge composed of water, shattered garnet, and whatever you're cutting. You have to pay to get rid of that stuff. Waterjet cutting is a good industrial process, but not office-ready.
(On a vaguely related note for maker types: does anyone know of a good low-cost surface mount reflow oven? The common low-end T962 has a big hot spot in the middle of the heating area and will scorch boards when used at lead-free solder temperatures. Yes, there are fixes, but I want something that works out of the box.)
The "maker revolution", such as it is, is driven by laser cutters. They're accurate, fast, reliable, work on large sheets, and have no consumables.
We're pushing 14 bar of 99.8% purity nitrogen through our 4kW fibre laser cutter to cut up to 12mm thick stainless steel.
Regular mild steel is reactive cut with 0.4 bar of oxygen and a trickle of nitrogen for lense cooling.
We go through 110 cubic meters of nitrogen on a slow week. We've got a 1400 litre liquid nitrogen bulk tank and 3500 litre liquid oxygen bulk tank.
If you bought a laser cutter earlier you probably have a CO2 laser so there's another consumable.
And we're getting slogged on lenses and protective glass windows. That's how they get you, the long tail of proprietary parts. I need to find a cheaper supplier of parts, but do you really want to put aftermarket lenses in your bosses one million dollar laser cutter?
To add to that, laser cutters are bloody dirty. Our machine is fully enclosed with a large (4 meter high) external dust extractor and self clean filters. The ultra fine dust that collects in the hopper is not something you want to be breathing.
And then theres the tons and tons of oxydized steel waste, stainless steel dross, and scrap that fall through that has to be sent for recycling but no one will take when the scrap prices are low like they are now.
Pretty sure the parent is talking about cutting acrylic, wood, etc. At my last job we had a small epilog laser and it was fairly low maintenance. Metal cutting is another world entirely, for sure.
On a vaguely related note for maker types: does anyone know of a good low-cost surface mount reflow oven? The common low-end T962 has a big hot spot in the middle of the heating area and will scorch boards when used at lead-free solder temperatures. Yes, there are fixes, but I want something that works out of the box.
Out of the box? No, sadly not. But you may find this interesting nonetheless: a friend of mine took a Black and Decker convection oven, replaced the control electronics with his own PID setup, and created a reflow oven that works better than the T962 he had. If you're willing to put in some elbow grease, you may find the DIY route appealing.
We have a Makerbot Replicator 1 in our Makerspace and it actually works well and is very useful. But we modified it quite a bit (mainly with an enclosed and heated print volume and also a more sturdy build platform). A lot of the quality of a 3D-print comes from the software and Makerbot actually did a good job there.
I also don't think the parts are too weak. Certainly, machined aluminium or lasercut MDF is more robust, but for a lot of things I find that 3D printed ABS or PLA is plenty strong enough ;)
I very much agree on the usefulness of laser cutters! Unfortunately they are too big for our space :(
> does anyone know of a good low-cost surface mount reflow oven?
I use a Black and Decker toaster oven I got off Amazon for $25[0] Works great, super cheap, even heating is important in cooking after all. I set the second dial on "Bake" and the third dial on "Stay on" then use a stop watch and the first dial to control the temperature according to what-ever re-flow profile I'm trying to achieve. You can tell when the oven reaches temperature by listening to the relay click noise it makes.
You can buy an add-on controller for this kind of oven called the reflowster but I don't personally think it is worth the money at $140. The stopwatch method hasn't failed me yet.
I feel like a convection oven might be the next level of reliability in this space. You should at least get very even heating since the airflow is fast enough to be turbulent.
Reflow is supposed to be done with convection heating, not radiant heating. The industrial reflow machines are all pure convection. Usually with rather low air speeds, to keep from knocking the parts off the board during soldering. There are good options starting around $1500. Unclear why there's such a huge price jump from the IR rod ovens at $200 and the convection ovens around $1500.
They make convection toaster ovens that are about twice the price of basic toaster ovens. The huge price difference on ovens branded for reflow work is almost certainly because of the lower volume of that market.
> Makerbot's big problem was that their machine sucks. Low-end 3D printers just aren't very good.
This has been my experience with them, as well. They're a low end printer but priced way outside of low end range.
My uni had a lot of them (50+), and the guy that ran the 3D printing lab was constantly repairing/swapping out extruders. So many, in fact, that when he said "per week," Makerbot themselves asked, "you mean per month?" No, it was per week. I forget the exact figures, but it was shamefully high considering their price point.
Pretty much every Makerbot in the school was guaranteed start the month with one extruder and finish the month with a new one, likely with a new one in between as well. And the lab had a pretty big pile of broken extruders that just grew almost daily.
Give it another 10 years. These sort of things often tend to require upgrade in tech to become actually successful. It will be huge one day soon, just not right now. The idea is sound, it's just the tech and the execution is missing.
Well, how do you measure "successful"? I mean, welding machines are pretty damn useful and they're definitely successful tech. Yet you don't exactly see one in every home.
Part of 3D printing's "success problem" IMO is that people started using a hopelessly ambitious definition of success.
This gets the current state of 3D printers half right. I wrote the same thing seven months ago http://hackaday.com/2016/04/28/the-makerbot-obituary/, and yes, the downfall of Makerbot was due to going closed source and shipping faulty 'smart' extruders.
However, the author gets it completely wrong by not looking at any of the other companies shipping reprap designs. https://www.lulzbot.com/ has seen consistent yearly growth, from tens of thousands a month when the first MakerBots came out to tens of millions per month today. http://prusaprinters.org/ is experiencing the same growth, and that's despite an obscene number of clones of their machine on the market. The growth of 3D printing never stopped. If you look at the well-respected companies in the field, they're seeing consistent, increasing revenue.
The trouble with the industry is Makerbot. They burned their community with the change to closed source, they killed their industrial/educational market with the Smart Extruder problems, but they were the darling of the media. When Makerbot laid off employees, it made headlines. When manufacturing was outsourced to China, tech bloggers stumbled over themselves to get a post out. As Makerbot went south, so went the perception of the industry.
3D printing is still a growing industry, and the tech in low-end printers is getting really, really good. It'll never be a printer on every desk, but if you find a household with a circular saw or a soldering iron, you'll probably also find a 3D printer. That's what it should be, anyway: a tool, and not a fetishized technology.
Well, the issue with 3D printing as I see it is 3D modeling.
Anyone can fire-up a word processor and produce interesting looking documents and flyers.
A lot less people can fire-up a paint program and produce interesting images.
A lot less people than that can fire-up a 2D CAD program and produce 2D mechanical or architectural drawings.
And yet a lot less people than that can run a 3D modeling program and product much beyond trivial trinkets.
Beyond "Look! I printed this tiny Statue of Liberty from a file I got online" and "Check out the letter A I made!" lies real mechanical design. This is where mechanical things start to become useful.
At that level you now need far better understanding of how to design multi-component mechanical assemblies, materials, etc. Anyone can do a little pyramid in Sketchup. Not everyone can design a full robotic hand with differential drive and force compliance.
I imagine most folks buy 3D printers, print a few things they can get online and that's the end of it. A very few are inspired to learn some more and do some basic modeling. I can only see a very, very few go all out, learn advanced modeling and mechanical design and start to extract value out of their 3D printers.
I completely agree. To extend this, even if you can 3d model something there is no reliable way to convert that into something printable. Some of the basics aren't even implemented, such as:
* You have to model explicitly the model walls. That is you can't just say "using X material, print this model with whatever minimum thickness is necessary".
* Even if the model visually looks correct, you may have to remove hidden surfaces or non-continuous (or connected) surfaces.
* Even if an online service like shapeways verifies a model it may not print correctly. And it's not always clear how to fix it.
* Solid models are extremely expensive and so you have to create a shell. Shells have several issues like print orientation, holes for the excess material to drain out, thickness of solid, overhangs, etc edges.
* Typically shapeways, etc either don't support certain formats or don't support them well. This means you have to convert your model into whatever format they support. This can introduce more errors.
* Automatically re-sizing - strangely enough if you can print a say 10cm tall model on a printer you can't necessarily print a 15cm version, even if it fits in the print volume.
Given all these things, even if you have a decent model, there is a large subsequent step in turning that model into something that can print on a specific printer and material. It's likely that printer hardware makers have to create some software that takes popular native model formats like .stl, .blender, etc and convert it reliably into a printable model for that specific printer/material.
My point was to say that 3D printers end-up collecting dust because what's needed to make them useful is beyond the realm of what most people --not all-- are willing to devote time, effort and money to learn. Without learning these things the 3D printers become useless very quickly.
Agreed, I think it's reasonable to ask that ppl are able to 3d model, which is a considerable skill in itself.
However, at this point, there's a huge gap between what a, say, sculptor would create from looking at the model and the garbage you get out of a typical 3d printer/software combo, if you can get it to print at all.
> Anyone can fire-up a word processor and produce interesting looking documents and flyers.
Actually, even this is harder than it seems. Having worked as a teacher, I saw many teacher create their own materials. They created handouts, pamphlets, worksheets, exams, short instructional materials, and some even whole text books. Compared to profession stuff from publishers, these self-made artifacts were often awful. The artifacts worked, of course, but typesetting, layout, illustrations, printing, and editing were almost always sub-par.
The most common example of this was trying to fit a worksheet or exam on one sheet A4, printed on both sides. As a result, my colleagues and I chose to use smaller fonts, smaller margins, decreased image size, less whitespace, and so on. Over-all, we ignored the intended readers (also known as our students), and focused on secondary matters such as ease of distribution.
On the other hand, where these instructional artifacts did shine was with respect to their practical applicability and usefulness because they were custom made (by the teacher) to fit a very specific situation (their classroom). Which is why readily available computers, software, and printers are such a powerful and awesome tool for a teacher.
For random DIY stuff you don't always need advanced modeling.
If I had a 3D printer right now I'd want to make a circular gasket with an outer diameter of 78 mm. That's the simplest modeling imaginable but it would be useful for me.
It's like how a power saw is useful even if you don't know advanced furniture construction...
No it isn't. OpenSCAD is great for mathematically defined objects, like gears and fasteners.
For everything else traditional parametric CAD is a million times better. If you don't mind closed source, give AutoCAD 360 a try (make sure you use it in parametric mode; direct modelling is a stupid): https://knowledge.autodesk.com/support/fusion-360/troublesho...
If you can't stand closed source then there are not many options. Most open source CAD software is worthless. The only one I've found that works as it should is SolveSpace - http://solvespace.com/ - but it does have a slightly 80s interface and... well SolveSpace is to Solidworks what Notepad is to Word. But it does do constraint-based sketching and parametric modelling right.
OpenSCAD is fantastic and, somehow, very "intuitive" for some people (while other people hate it). But it's pretty difficult to export anything from it to continue working on it. The only solution that I have found is to import scad files to FreeCAD and then try to "refine shape" which sometimes works great and sometimes doesn't.
But the import filter in FreeCAD is old and not maintained (for example, text/fonts were added to OpenSCAD after the filter was written, and the filter has not been upgraded to be able to deal with them).
You can certainly print directly from OpenSCAD and it works well, but when you go from 3D print to an industrial process, factories won't accept scad files, and you usually want to do things to your parts that OpenSCAD won't let you do, such as filleting angles. In a plastic 3D-printed prototype fillets don't matter, but the same part in aluminum or steel will have very sharp edges that will cut the users' hands if they're not filleted.
Yeah that's why I take advantage of the student discount that taking a class at the local technical college offers and get full solid works for $170/year.
But then of course, that license is only valid while you are a student. So after the class, you might as well have just pirated it since either way, you are breaking the license.
I still frequently end up crawling back to OpenScad. It is often quicker than the rather click business of setting up constraints. Just don't ask me to make a rounded corner in it.
This is one area where I think VR will produce a solid win. IMO the steepest parts of the modelling learning curve are the result of the complexity required to manipulate a 3D world using 2D interfaces. With VR, both the input and output interfaces have the same dimensionality as what you're building - I think that'll have a big impact on how easy it is to get into, and consequently how much content is created.
It's like confusing learning Python with producing solid software products.
OK, let me illustrate the point with a question: You have two parts that attach to each other. The parts are joined using a set of six 5 mm bolts with washers and locking nuts. The hole pattern is circular and has a radius of 3 inches.
What are the minimum diameter and dimensional tolerances you should specify in order to guarantee mating across multiple parts and across thermal variations?
OK, same two pars. Now they mate using no screws. They use a one-time-only locking tab and catch system along the inner periphery. What's the geometry and what are the tolerances?
Interestingly enough, none of the above talks about how to use the UI of some 3D program to solve the problems.
To ellaborate, when you're using cad you have to give thought to how the part will be produced as well. There are a lot of shapes that 3d can't build, there are minimum angles, all sorts of restrictions. And there's the nitty gritty of actually fitting real shapes together in the real world, having the right flange sizes for holding screws, nuts, etc. And then you get into wear, balance, etc.
You're right. When I wrote the above, what I had in mind was sculpture (what I've mostly see consumer 3D printers making these days - representational things like D&D set pieces, or cosplay props). The interface of programs like Maya, for making very complex 3D shapes, I wouldn't call trivial.
But in hindsight the context of the discussion is more about industrial design of functional consumer goods, which I completely agree is a highly skilled trade, and while learning the CAD program is still a lot of work, that's one of the least of the requirements.
I don't think so. I do a bit of AutoCAD at work, mainly for pressurized vessels (i.e. pretty simple stuff, just has to follow the regulations). 95% of the time, you're working with a 2D sketch in either the x, y or z plane and extruding from that. The other 4.9% of the time, you make some other plane based on those extrusions and do 2D sketches in that plane. 0.1% is reserved for 3D splines.
That's pretty funny. I've been using Solidworks for somewhere around 12 to 14 years. I also run Siemens NX.
Clarification: Not saying it's impossible to learn. All I am saying is that a very, very, very, very small percentage of 3D printer buyers develop the chops.
Don't confuse an intuitive UI with mechanical design.
It's the difference between the intuitive UI of a word processor and writing a novel.
The same could be said of, say, Python. Easy to learn from a few videos on YouTube. There's a vast difference between that and actually producing good, efficient, fast, sophisticated and accurate software.
I've been programming, designing hardware and electronics for over twenty years. There is no way anyone can obtain the equivalent chops and experience from a bunch of videos on YouTube. Maybe they can learn to use Altium Designer, Solidworks, a compiler or two and Xilinx's FPGA toolset. That does not mean they can actually design products at all. That takes years or learning and dedication and is hard.
Learning to drive Solidworks from a few YouTube videos does not make that person a mechanical engineer or a mechanical designer beyond the basics.
Example, unrelated to 3D printing: Design a differential, cable-driven wrist actuator with 3 degrees of freedom. Use a combination of 3D printing, machined parts and off-the-shelf hardware. Use SW to run load analysis on the parts and assemblies.
Calculate allowable dimensional tolerances. Produce mechanical drawings with annotations for class of fit of various components as required.
Now make the entire assembly (all components) parametrically driven by a reference sketch. Automatically produce a new set of prints and revisions if any dimension is changed. In other words, if you change one gear all other components in the actuator redefine themselves based on the new gear and so does the assembly. If the new gear is large enough to accommodate four mounting holes rather than two, it should implement those changes automatically.
I still hold that creating anything truly useful for 3D printing requires background that most people don't have and are not interested obtaining through hard, self-directed work.
What you are saying is true, but you are also setting a ridiculously high standards with examples like "differential, cable-driven wrist actuator with 3 degrees of freedom." There is a lot of room between that and what a beginner in solidworks or tinkercad or blender can do. Just some time browsing Tinkercad's gallery of things. I promise you most aren't products of mechanical engineers (nor will they threaten the jobs of mechanical engineers.
Exactly. I don't need a 3d printer to build my own cable driven wrist actuator. I just need to print a beefed up version of that weak hinge you included because you spent so much time building a fancy 3d model in solidworks that you forgot to test your final creation in real world conditions.
> spent so much time building a fancy 3d model in solidworks that you forgot to test your final creation in real world conditions
And that, almost precisely, betrays a lack of understanding of mechanical (or electrical, or architectural, etc.) design.
CAD is a tool for documentation and, to some extent, verification. A mechanical engineer (or self-taught designer) isn't made by being able to use the tool. Mechanical design starts with a fundamental understanding of principles, techniques, structures, familiarity with historical solutions to similar problems and a deep understanding of DFM as well as DFMEA. CAD isn't about creating fancy 3D models at all.
You can use 3D CAD to design things that are impossible to make. I think you are equating 3D CAD with mechanical design, which isn't quite right.
Do you have any recommendations for embarking on that path of hard, self-directed work?
My co-founder at http://8-food.com/ and I are at the earliest section of that path right now and - while we are attempting to optimize for manufacturing by assembling primarily from pre-existing commercially available components - do recognize the need to improve our presently nearly nonexistent mechanical engineering skills.
In your case you are either trying to deliver a product or educate yourself. It is very, very hard to do both. Particularly if you are, as you said, starting with "nonexistent mechanical engineering skills". A lot goes into a product that is reliable, durable and manufacturable. This would take a lot of time and errors to learn. A better formula is to hire someone or a company well versed in that part of things. Focus on your core competency.
People like me are able to do what we do because, at some level, we are a product of a different generation. I was disassembling and rebuilding engines and modifying cars in my driveway when I was 16 years old. I designed and built my first computer from raw chips when I was 18 and wrote my first OS out of hand-keyed hex machine codes afterwards.
Today I've worked with young engineers who openly admit not being good at soldering and know PhD's who have trouble assembling Ikea furniture. Not sure what it is, but it sure seems engineering is being done differently these days. Back then, if you went into any of the engineering fields it was very likely you tinkered ad nauseum way before hitting college. Not so these days.
What I am saying is that as an electrical engineer by schooling I had already done tons of mechanical, software, optical and other work and picking up some aspects of these other fields over the years was an organic process. Starting from "nonexistent mechanical engineering skills", well, that's a bad place to start. Driving CAD isn't going to make you a mechanical product designer.
I agree there's a wide gap between modeling and designing workable parts, but parameterized modeling mechanism is kind of out of scope for home printers. And while I guass many won't know about tolerances as intimately as professional most of the time guesswork and trials will be good enough for home appliances.
It's not like everyone will build robotic arms with articulate hands as phrotesis on repraps.
What would be some resources one could use to step up from Solidworks warrior to "decent enough at mechanical design that you wouldn't make a pro cringe"?
Experience. Make lots of mistakes. Build things. Read. A lot.
I would also add, ironically: No 3D printing.
It's very interesting to see young engineers do such things as specify, for example, an array of 0.250 in holes spaced 0.750 in apart and require a tolerance of 0.001 in.
You then take that same engineer into the shop and ask them to drill just two of those holes.
The surprise in their expressions when they can't produce a single good pair of holes is very telling.
3D printing doesn't teach you about mechanical design and manufacturing any more than a word processor teaches you about good typesetting and layout.
So, I'd say, design things and make them by hand using hand tools, manual milling machines, manual lathes, etc.
Just drilling a single precisely positioned and accurately dimension-ed round hole can require an hour of work, if not more. There's a reason professionals call it "hole manufacturing" rather than "drilling". You need at least three tools to manufacture precise round holes.
Anyone who thinks that drill bits are for making round holes needs to go do some research.
Impossible to boil it down to one book. So many disciplines.
I would suggest you get a copy of Machinery's Handbook and study it.
There are also books that are collections of mechanisms. I probably have ten of them. There are some online resources today with animated versions of mechanisms. Google is your friend here.
Learn about materials. What's the difference between mild steel, hot rolled steel, stainless steel, 6061, 7075 and MIC-6 Aluminum. What do you use, when and why?
Research topics such as "hole manufacturing" and "high speed machining".
Read online forums such as "Practical Machinist", ask questions.
Learn about Geometric Dimensioning and Tolerancing.
If you don't have a background in Physics, learn some. Mechanics and Heat would be the basics. Don't need Calculus for the fundamentals.
Plastics is a topic in and of itself. Study injection molding. Visit sites like Protolabs, read through all of their materials. Understand their process. Learn about conventional injection molding.
Learn about analyzing structures/designs for strength, deflection, etc. Do this by hand first with simple shapes.
Learn FEA (Finite Element Analysis) and apply it to both mechanical and thermal analysis of designs.
Understand how to use fasteners. You'd be surprised how many people don't understand such basics ideas as "you never use a screw for positioning, only for clamping".
Read about finishing: paint, powder-coating, ceramic coating, electroplating, anodizing, porcelain enamel coating, chromate, passivation, texturing, sand blasting, brushing, chemical etching, etc.
Learn about laser and water-jet manufacturing processes and when to use them.
If you have access (or the money) get yourself in front of a milling machine or lathe (or both) and learn to take parts from CAD to finished product. Not trivial. Don't need a big sophisticated machine to do this. A mid-size bench-top mill and lathe will do. No CNC. Manual first.
Design and fabricate all the parts needed to convert your mill and lathe to CNC. Now learn CNC machining.
There's more, lots more. For example, adhesives is a huge topic. Lubricants is another. Composites. Casting. Extruding. Sheet Metal. Die casting. Etc.
Yet another huge topic is understanding all of the above in the context of costs and manufacturing efficiency. What's the consequence of designing a rectangular hole with 0.005 in radius corners on an aluminum part to be machined? When do you switch between extruding and machining and why? When do you completely redesign a part in order to better fit the best manufacturing process? What are the implications and manufacturing realities of requiring extreme accuracy?
You are not going to pick all of this up in a week or a year. It takes time, years. I've put my hands on most of the above during my career, but I am talking about 30 years designing and manufacturing all kinds of products. Don't expect instant results.
That, BTW, is why becoming a Solidworks "driver" does not make someone a mechanical designer. Solidworks is simply a design documentation tool. Becoming good at driving the software will not provide anyone with any of the knowledge listed above.
Better done in the context of actually working rather than tinkering. You don't have to become an expert at all of these either. Almost no such thing any more. Yet, it really pays to have as wide an understanding of manufacturing as possible. 3D printing is an almost insignificant segment of the totality of manufacturing.
I'd look online for mechanical engineering courses. If anything you can put together a curriculum and slowly follow it.
It's hard to make a recommendation. I did not learn mechanical design formally. Also an EE. However, mechanical has always been a part of my life since I was a teenager. The only way I can summarize my path is: doing thousands of projects and constantly learning.
I think this is the key point. I have years of experience modelling for 3d animation and have been trying to print a complex object for the last few weeks. Modelling an object is one thing, but modelling a visually interesting object that doesn't require support structures everywhere is a real skill that takes some mastering
Because if your object is tricky enough, you'll have to spend a lot of time removing them, risking damage to the object. Dissolvable supports are a thing, but I haven't seen many people with double-headed extruders in their printers.
I think the main issue is that there is no killer app. Anything that would make me want to buy one. Once that is figured out, everything else will fall into place.
I envision Crayola being the first to put out a consumer grade 3d printer aimed at kids. Their creativity product line is pushing the envelope in refreshing new ways every time I pass that aisle in the stores.
Haha, as a mathematician, the notion that math sculptures are the killer app to bring 3d printing to the masses is pretty suspect.
Ultimately, these printers are one tool in an arsenal of tools for making things; they don't stand alone. There article gets it right that these are more likely to be tools for hobbyists and professionals who need lots of physical prototypes and (maybe) schools. These aren't trivial markets at all, but also aren't all of the people in the world, either.
Home 3D printing is not worth it for most people in the developed world, but in developing countries it can be.
When I lived in the UK, I would default to trying to 3D print anything I needed around the house. But manufacturing is cheap enough, and availability of products so ubiquitous in the UK, that it was almost always cheaper to just buy the thing on Amazon instead of print it. The only thing I printed that I couldn't have bought cheaply was custom plumbing parts.
Now I've moved to Kenya where Amazon doesn't deliver, import duty is extremely high, and availability of low throughput items is non-existent. I 3D print everything from spare parts for my truck to basic DIY supplies like rawl plugs and cable tidies - things I just cant get cheaply. Of course the fact that 3D printing is so useful in this environment is offset by the fact that hardly anyone here can afford it. I'm planning to set up a community makespace where people can print cheaply, but I don't expect it to cause a revolution.
My wife and I were both coming to the end of our PhDs and knew we didn't want to continue in academia. We have a successful software consultancy on the side, and have been teaching bioinformatics around Africa for a couple of years. We came to Kenya for the summer, then realised we had a much higher quality of life here, our work had much greater potential impact, and there was nothing stopping us from staying. So we stayed :D
The author is wrong- but it's not their fault. All the interesting things (new technologies) in 3D printing today are happening at an enterprise price scale right now so if you're not looking too hard, it can be hard to see.
Today 3D printing is eating small-batch manufacturing. Tomorrow any production run of less than 5000 units will happen on 3D printers.
Bre Pettis and the maker (and investment) community had the right ideas about the future, but I believe Makerbot focused too much on how to sell a hot glue gun (with extra steps) to the public and far too little on how to make a product that was truly useful to their customers.
I am looking at your machines now. I can't seem to understannd why your smaller machine is $3500 and the slightly larger one with manual laser scan is $69,000. Its not like the laser scan is controlling the machine, and adjusting printing in real time.
20x price change. If I just want to print in Onyx material, there seems to be no point in the more expensive machine. The carbon filament printing is cool, and I may step up to the $13,000 machine. I just wish someone could justify the big machine - or in reality I wish it was actually significantly bigger. Double each dimension, so 6x the build volume of the small machine. It is using linear guides, it is easy to purchase them in any size needed.
The pricing seems to be arbitrarily aimed at maximum cash extraction from the customer, not reality. I will probably end up with the cheaper machine. I am looking for a way to print parts I would normally machine from aluminum.
It is a well built machine, however. It makes very very nice out of the box prints too. Retail quality, probably the first machine that I have seen using filaments that can do that (in the pro-sumer price range, excluding big dollar machines). It will be machines like this one with good parts inside (slides, bearings, servos) that will make 3D printing really useful.
At the tradeshow I saw the machine at (SEMASHOW) , the side by side comparison to the machined aluminum parts was quite impressive. The part felt like half the weight, and had really nice surface finish.
If what you need is a regular FDM machine our $3500 printer is a really, really fantastic one. It's the least hassle and has the best out of the box prints out of any desktop 3D printer I've used.
If you're looking to print aluminum replacement parts, you should look at the higher end machines. They have the ability to print continuous composites (our homegrown, weird, and very powerful 3D printing technology) which will let you print parts which are strength/stiffness competitive with Aluminum.
While our $69,000 printer definitely has an enterprise price-tag, it's still the cheapest (on an amortized cost basis) way of getting strong parts with a 24hr or less turnaround. The purpose of the laser scanner is to close the loop on what is typically an open-loop system in essentially all other 3D printers. I can't say too much more about this but we're adding a lot of cool features all the time.
(Obvious Disclaimer: I'm trying to sell you on the printers we make because I believe they are the best)
So why doesn't the big machine actually print something larger? I have a Haas Mini Mill, which has a 16x12x10 inch work volume, and that is small a lot of times.
I would buy the $69000 version in a heartbeat (we are a manufacturing company) if it would print a 12x12x12 inch size or a little larger
On the Mark X (69k version) our build volume is almost exactly 13x10x8 inches (we round down slightly when we list dimensions on our website).
If you want I can put you in touch with someone who can help you figure out if your use cases will fit our printer. Send me a note if you're interested: abe@markforged.com
I visited a 3d printing center a few weeks ago. The guy was extremely busy printing all sorts of products for hospitals and medical labs. The occasional comicon type stuff came in, but a huge demand for b2b 3d printing.
I meant were they doing the more advanced printing methods like a stratasys or the light based ones or the simple extruder type ones? Or TI sintering or what?
They were all extruders. He was talking about getting some more advanced, but that was it. It also meant he had to walk through designs with people so they could do things like have proper scaffolding and other things. A lot of people, including me, thought you could just plug a design in and just hit print, but it was more complex than that.
This is one thing that Stratasys totally missed out on. They are heavily reliant on support material and other consumables, which is great for making arbitrary shapes but not for cheap production.
3D printing is very much like coding, promising a revolution for those with the mindset to work with it. My wife, a coder by profession, is extremely active in the 3D printing community, and her ability to research and stick to problems is why she is successful at it. This is not an accessible hobby, as is evidenced by so many of my friends who own 3D printers that started gathering dust the moment they learned it's not simply plug-and-play. You have to figure out your infill settings, watch your inclines, your print-speeds at various points in your prints, know what temperatures work well with what materials, know how to replace the parts that go bad every few hundred hours of printing, etc, etc, etc. It doesn't help that you will sometimes be hours into a print before you realize you screwed something up.
The 3D printing community, like some coding communities, are very supportive and incredibly welcoming to beginners, but the skills we use in coding, the ability to research, experiment, and stick through extremely frustrating problems, are skills most people don't ever really learn. Our public education system is actively trying to address this with the much-misunderstood Common Core standards, but until such a mindset becomes more prevalent in our culture, things like coding and 3D printing will only continue to grow slowly in popularity.
In the meantime, nerds should see 3D printing as a still-uncharted territory for innovation and opportunity.
It sounds a bit like we are at the stage when computers were using punch cards. Maybe in a decade or two, things will get a bit more interesting.
Agreed. MakerBot in 2012 was like trying to start Apple in 1962. What kind of personal computer could one build from the era's transistors and magnetic core memory? That primitive thing could have sparked someone's imagination, but it wouldn't have had any mass-market applications.
3D printing's Apple moment is still waiting for its equivalent of the integrated circuit (a technological enabler) followed by a VisiCalc (a killer application).
I'm guilty as charged. I predicted in 2014 that within 3 years majority of upper middle class households will have 3D printer just like they have paper printer. Boy, I was wrong and it feels quite laughable claim now. After initial hype of continuous reduction of prices, things have just came to halt. Most people don't want to spend few hundred dollars just to build lifeless models of dragons and there has been no killer application that can justify the price.
I predicted in 2014 that within 3 years majority of upper middle class
households will have 3D printer just like they have paper printer.
I don't have a paper printer anymore, and I wonder how many households still do. I'd wager that more households got rid of their paper printers in the past 3 years than bought 3D printers.
I also don't use paper printer more than may be 5 times a year. But those 5 times are important times and these things are cheap enough for the convenience of not have to travel to some shop with USB.
Having been buying quite a lot of toys for our young children recently, it did sort of strike me that a moderately affordable 3D printer that really did Just Work, where I could go to, say, lego.com, click what I wanted a model printing of, and then it all just happened, as if by magic... would be something I'd buy. I think the combination of convenience and "Wow, that's cool" would be quite compelling, even if it didn't really work out all that much cheaper.
On the whole, though, I'm really struggling to see how these things will ever take off. (Like others, I have a paper printer which I use a handful of times a year, and it's never an experience I relish. At least printers are very cheap now, even if ink isn't.)
Well, toner is, and most laser printers have sensible prices nowadays. At least with laser the ink in head won't dry out, causing cleaning procedure or ink replacement.
On the topic of 3D, I don't see a market for home use, but in enterprise - well, I could use one right now if they were cheap enough, easy to use and produced quality products (or maybe I just don't know that they are? Last researched a year ago...)
I own a 2D printer and have found my use of it declined dramatically over the last five years.
Given I just became a Google Shopping Express member and I am already addicted to Amazon Prime, I anticipate my desire to make my own 3D items at home to also decline over time.
I do agree with @JDDunn9. There will be businesses that will provide 3D printing services using state of the art printers, high quality materials, and their expertise. Bay Photo is a 2D example of this kind of service. They will deliver the results quickly, cheaply, and reliably using distribution services like Prime and Express.
I'll buy a 3D printer when I can print all the pieces to build a Nerf N-Strike Elite Hail Fire for under forty dollars worth of materials and it can be printed and assembled in under three hours.
I picked this item because it was featured in the Wired magazine article right after their "3D printers are the next PC" article.
~
3D printing is inherently limited by what you can do with plastics (until you start getting into laser sintering or something else exotic). The niche application ends up being Yoda figurines and etc that require weird shapes and high detail.
For my money, cheap & modular mills like those at http://carbide3d.com/ are much more promising because you can actually make structural components. A $500 CNC'd Grizzly mill equivalent (hell, probably $1000 if it was user friendly enough) would be extremely useful and would sell like hotcakes, but of course that's empirically extremely difficult to design since mills are a mature technology that requires things like calibration, knowledge of tools and materials to use, etc. Even their $1100 machine would probably be fun as hell, but it's a price point that actually requires a use in mind.
Incidentally, as far as I can tell the most successful home-manufacturing product of the past 10 years or so has been https://ghostgunner.net/ .
I think the economic model of 3d printing in home is off.
The goal should be to have advanced 3D printers, latches and CNC milling machine shops that provide easy to use services for walk-in customers or mail-orders. It should work just as easily as copying and printing services work for PDF files. You don't know if they have Xerox or Canon and you don't care. You want to know if they can print A1 and things like that.
As 3D printing advances, having affordable access to several $20,000 - $200,000 specialized printers in your city beats the $2,000 printer in your home for a serious hobbyist or small business. Maybe its worth paying $200-$300 for printer you can use for prototypes before ordering more expensive work.
The problem with 3D printing / lathing / milling as a service is that particular parts can have orders of magnitude differences in machining costs, and typically you need humans in the loop to figure out if you're asking for something ridiculous (or dangerous, or harmful to the machine...)
These guys are close to that model: https://www.plethora.com/ . Their supposed killer advantage is that they have spent a lot of time deriving a cost model for arbitrary designs which allows relatively unsupervised submissions, but because they're very dependent on that cost model, the number of operations and customizations they support is fairly limited.
That sounds like the PCB prototyping/short run production house model. Which makes more sense. As long as the cost isn't too high and the turn around time is short.
I disagree. Yeah, it's plastics only, but that's not really that big of a limitation in my opinion. Now, I've printed a yoda or two (who hasn't?), but where it really shines for me is small things that need a little bit of precision to be good.
And yes, I could use a CNC mill an create a lot of my parts and they'd be just as good, probably stronger. The problem is that I live in a studio apartment. Since I don't have a workshop and I don't want to drive to a hackerspace when I want to make something, I'm going to choose not to subject my girlfriend (and preferably not myself) to to the mess and noise of a mill. Plus a printer is so much easier in my experience, I don't need to worry about setting feed rates or replacing bits, I can just click print and go watch TV while it prints.
I'll admit there are things that I can't do that I could do with a mill (I would really like to be able to mill circuit boards), but at least to me a printer makes much more sense for the vast majority of lazy stupid people like me.
Machining plastic on a small tabletop machine isn't very noisy. And it's a lot less messy too since it doesn't involve any liquid chemicals or fumes, you just have the chips.
> Writing on Brokelyn.com, former employee Isaac Anderson placed the blame for those three machines’ problems squarely on MakerBot’s decision to go closed source. They could no longer rely on their old customer base of “capable hobbyists who provided tech-savvy feedback and suggestions for improvement.” The new class of buyers, he wrote, “were largely incapable non-hobbyists with no useful feedback, only unrealistic expectations.”
3D FDM plastic printers are an useful component of the future the hype described, but they are just a component, they are not energy to matter Star Trek replicators.
If one designs for it, it is possible to build complex products. FarmBot is a good example. It is made with a combination of 3D FDM printed parts, flat metal parts that could be made with a plasma cutter or CNC, open source software and compute provided by small ARM computers, and commodity produced parts like wires, motors and extruded aluminium rails.
Thats how RepRaps are made, and that is how you have to build in this new way. Very few things are solid plastic shapes.
3D FDM printers are great for designing and producing the case that your etched PCB, switches and LED's attach to. It is an ingredient in the recipe.
FDM can be used to make custom metal components through green sandcasting metals.
If more projects were designed like RepRap, or FarmBot, this new industrial revolution could still take shape. Robotic PCB production is another key ingredient that is needed for reality to match up to the hype.
You can't even print composites. Seriously, most 3d printing fanboys I've met couldn't even begin to define the term "tensile strength".
I had one guy insist that his 3d printer could 3d print wood. I said there's no way, so the next day he brought in an object made out of ABS with wood flower suspended in it. He literally couldn't comprehend how that differs from a real wood object, structurally, or even better how it'd differ from a cold-molded composite wood object.
>3d printing sucks for any kind of production structural part.
Good thing AuroraLabs is working on that problem. Won't be long until their 3D printers are standard accessories on oil-platforms and deep mines and the ISS, oh my...
So at Tinkermill.org in Longmont CO, where I am a member, we have 3d printing machines.
We also have a good machine shop, and some of the members got together and refurbished an old injection molding machine.
So, we can 3d print a test, then can transform the finished design into instructions for the CNC machine, to produce a mold which goes into the injection molding machine.
So far no one has used this to prototype a part that started a business, but that is a great way to use a 3d printer, as a first run prototype which can then be used for other manufacturing.
A guy who is designing a game used both the 3d printer and laser cutter to make his prototype board and pieces.
We also have a few whizzes who have whipped up replacement parts and e.g. hose adapters for the dust collector in the woodshop.
Overall 3d printing is like the early days of the commercial Internet: people intuitively knew it was changing things, but had only scattered anecdotes as proof.
I think that we hackers tend to focus on tech-oriented looks at why things succeed or fail. I propose another idea: that we need education of 3D modeling in schools, at a basic level, in order to encourage more people to understand the possibilities.
In middle school we'd learn various things about art, art history, art techniques, and while they weren't comprehensive, they provided a basic level of appreciation for art. As an example, if I hadn't learned about techniques like shading or perspective drawing, I would've thought that they were impossible for a layman like me to attempt.
For most people (and by "most people" I mean anecdotally most people I know), 3d modeling and 3d printing seems to be at this unapproachable level of complexity. If we can show kids the simpler side of 3d modeling, there would be a much greater consumer interest in 3d printing.
I'd rather have logic taught in schools before we waste time with 3D printing. We need to help people develop their natural bullshit filters so that fake news, advertising, and rhetoric have less impact.
Unfortunately, I don't think teaching logic will get you what you want here. Fake news may be false, bit it is often (mostly) logically consistent - this is part of the power.
For anyone who stops half way thinking this is some comeback of the marketing of makerbot, keep reading. Overall it's a fair summary of what happened in the world of makerbot.
As someone who was listening when makerbot was just starting up, I think that Makerbot as a company both did 3D printing a great service and a great disservice. Without makerbot it's entirely possible that without makerbot making as big of a splash as they did and without the press that they were able to create 3d printing wouldn't have taken off like it did. Unfortunately the disservices that they then did to the 3d printing community had so much more of an impact because of all that press that was primarily focused on them and equating '3d printer' to mean 'makerbot'.
All that said, I'll never give up my 3D printers. For me it was still a revolution, I think I just had a lower bar for what it needed to be to be a revolution.
> The new class of buyers, he wrote, “were largely incapable non-hobbyists with no useful feedback, only unrealistic expectations.”
I find this to be a strange statement when the previous paragraph talked about a class-action lawsuit due to a faulty extruder. Is it an unrealistic expectation now to expect your thousand-dollar 3D printer to Just Work? You can't just promise someone the future and then backtrack when people hold you to it.
It's just too early. I think we will see a 3D printing revolution but you need to be able to do useful stuff with the printers.
A good use case would be anything that's usually mass produced but actually benefits from high customization. Like clothes. If you could 3D print clothes, you might have a market. Or cables (electrical, Ethernet, etc.) that are an exact length. Or perhaps things like toothbrushes, or contact lenses.
The technology isn't very close to that yet. But if it keeps improving, there's a good chance that we really will end up with a 3D printer in every home.
Cabling is actually serious business; from an electrical standpoint, there is impedance matching, delay matching, cable shielding for EMI/C, etc. Then the connector itself has a slew of mechanical requirements: mate/demate cycles, tolerances, thermal. Then it all needs to be made for a reasonable price - there are entire groups of engineers that band together to produce specifications so that things like Ethernet cables can exist and give the performance that people need, while being interoperable.
Ethernet cables and (quality) clothes have to be assembled, not printed. And you need very specific materials to start with. Not realistic examples imho.
Well, there would be a revolution if you could buy a robot that would take fabric and a CAD design as input, and gave you ready-to-wear dress as an output. Without that, it's "meh, faster to go to a store" situation.
The problem with 3D Printer is, it is no match to the existing manufacturing technologies. Imagine, the sub micron accuracy of iPhone, which in no way achievable with current 3D printing technology. The technology in to which 3D printer can really revolutionize the world right now is Electric Motor. ( Here I have said it, please don't patent it. :P ) I am actually doing it. If interested, drop me a mail. :)
The real killer app for 3D printing is rocket engines. The engines for the Space-X Dragon spacecraft are 3D printed. Rocket engines are mostly plumbing, much of which is internal - the combustion chamber usuually has cooling channels in the walls. The traditional way of making engines has far too many parts and welded joints, just to make an object that's one rigid piece. This is the best case for 3D printing of metal.
I'm curious how you are using 3-D printing for an electric motor. Most of the key components, steel laminations, copper windings, magnets, commutator and brushes (if you are going old school :-) ) don't lend themselves to 3-D printing. I just see some mounting hardware being usefully printed.
Induction motor as currently built are actually pretty good. Nothing to improve there, as they are usually fixed speed motors. When you think about automative or very variable speed motors, which have much higher power density and energy density, in those situations rotor cooling problems, eddy current losses, hysterisis losses are very fun problem to solve in context of 3D printing. Im not sure how much improvement can be done, but lets see..
They are meant as prototyping tools, and some do quite well beyond that.
The trouble is, we can't 3D print PCBs .. not safely anyway. There are quite a few more toxic chemicals to deal with, not to mention the complexities you'd run into trying to make multi-layer boards.
Also I think people forgot plastic still cost money.
That's not how they were being hyped a few years ago. Everyone was talking about machines building themselves and "everything factories" sitting in people's garages.
I just sat there and thought "they finally managed to invent an automated manufacturing process that's even more slow and expensive than machining."
3d printing might not have had the huge impact expected for it in the consumer world, but in the professional world printing is hugely important. Getting a physical model of your cad file overnight from a black box at your desk or in your model shop is a workflow game changer. Make an ergonomic change to your grip? Test it out almost immediately. Manufacturer needs you to change the draft on your shell? Print it first before sending it out and make sure everything still fits. When previously you had to make engineering accuracy models all day in the shop or get a part made by a small run manufacturer, now prototyping and testing variations from cad can all be done in house. Consumers might not be directly benefiting from the advent of democratized 3d printers, but they are benefiting in the gains of efficiency at the firms designing their goods
I think it is similar to what happened to AI in the so called "AI winter", where professionals, consumers, companies and funding in general walked away from AI for quite some time.
We do know that for sure 3D printing will make a comeback for good one day, probably one decade in the future.
I'm not sure that is comparable. 3D printing isn't dead in professional use, research is still happening, has an active hobbyist community, ... it's just the "3D printer in every home" idea that Makerbot and others tried to peddle that's not happening.
The propaganda echo chamber apparatus killed 3D printing in the womb, when they started harping on bullshit about 3D printed guns.
Almost zero nerds gave half a shit about printing guns. Like, at all.
But news, news, news sites had a story to sell, right around the time of the Silk Road takedown, and poof, nerds received a message, which, in the context of other dubious criminal prosecutions, communicated that an alphabet agency might just destroy you and ruin your life because some abiguous technical detail made for an easy target.
Pirating music and movies, becomes a child pornography felony. Trading in bit coins becomes a Silk Road drug bust. 3D printing becomes weapons possessions with mandatory minimums.
Message received, propaganda apparatus.
Risk total ruin, by buying a 3D printer.
Got it. Nothing to hide, nothing to fear. Move right along.
What's the point of 3D printing a bad gun in a country where in some states you can go to Walmart and buy a nice gun there, for a fraction of a cost? And it's not like you could 3D print ammo.
It's a useful tool for engineers, but not everybody is an engineer. It was hyped like it was a replicator from star trek. The novelty of printing little plastic figurines wears off quickly. At some point you realize that you can buy a bag of a much higher quality injection molded version of the same thing at the dollar store.
They are useful for making things that don't yet exist, but that means it must be something you come up with and model yourself. And this is where they are no longer the tool of users, only engineers.
My wife and I have a Type A Machines Series 1 Pro and use it often. It prints like every kind of filament, including wood and metal filled. We've done several cool projects and are constantly on the lookout for the next creation. Next up we're looking at fabric printers.
Makerbot was a Model A: clunky, slow, power-hungry and limited to all but geeks; it is certainly a lot cheaper for a hobbyist than buying a Stratasys the size of a large refrigerator and costing as much as a small home.
To have distributed manufacturing, a much faster, higher capacity, more energy efficient, more flexible (more materials, deposition/removal methods), easy-to-use/-monetize/-protect IP machine/front-end would need to happen to be practical to do more than prototyping.
Amazon could pull this off and save on 'round-the-world shipping by becoming more of an IP to end-user fulfillment marketplace instead of a warehouse logistics shop. (for products which can be built JIT from standard parts/materials)
Oh, that was a painful read. Repeatedly they mentioned that all development to this point wasn't able to provide a cheap enough solution for the common user. And if you ever had a 3d printed piece in your hand you also know that the desktop printers still are not detailed and flexible enough.
So in end effect the current status doesn't proof at all that there is not a market, just that the technology can't reach market yet.
I just got a MonoPrice Select Mini a few weeks ago for $200, and spent about $30 on upgrades for it (an e3d hotend, the rest of the upgrades I printed myself), and about $80 on 4 1kg spools of filament. It's really great, I'm loving it so far. Been making my own things with OpenSCAD too, and some of my friends have been making things with SketchUp for me to print.
Reading the comments, it seems like the main thing missing here is multi-material 3d printing. Yes, there's no killer application for just printing plastic, but checkout, e.g., http://www.voxel8.com/
I'm glad someone finally said it. I don't have a particular habit of predicting things correctly, but in this case I really couldn't see what the point of a 3D printer was. Even when I was given free access to one I had no idea what I would ever want to build.
"Pretending you know what you’re doing is almost the same as knowing what you are doing, so just accept that you know what you’re doing even if you don’t and do it."
Can anybody explain how that should work out? Usually people laugh about someone who only pretends.
TechShop used to have some MakerBots, but they've been replaced by better machines. Materials costs are too high, though; the Form I and Form II use a working fluid that costs $130/liter. (This is a bit like the inkjet printer ink problem.) They also go through build tanks fairly fast; the working fluid in its solid form builds up at the bottom. But the build quality is excellent.
The "maker revolution", such as it is, is driven by laser cutters. They're accurate, fast, reliable, work on large sheets, and have no consumables.
The Next Big Thing is supposed to be desktop waterjet cutters. The Wazer, though, is apparently very slow, and may use more garnet per cut than the big waterjets. Wazer also glosses over the problem that waterjets generate a sludge composed of water, shattered garnet, and whatever you're cutting. You have to pay to get rid of that stuff. Waterjet cutting is a good industrial process, but not office-ready.
(On a vaguely related note for maker types: does anyone know of a good low-cost surface mount reflow oven? The common low-end T962 has a big hot spot in the middle of the heating area and will scorch boards when used at lead-free solder temperatures. Yes, there are fixes, but I want something that works out of the box.)