/ engineering

Lessons in Engineering - Assembling the Prusa i3 Mk2s and Multi Material upgrade (Part 1)

Introduction: This was originally going to be a blog about the multi material upgrade build for my Prusa i3 Mk2s, but it got a bit lengthy, so I split it into 2 parts. This first part is about the original Mk2s build, and I won't be going into too much detail as it has been incredibly well documented already. But it will give you an idea of where I was coming from when I approached the MMU build.

Part the One.

It might seem a bit odd, but when we started Enstaved I (Design Monkey) didn't own a 3D printer. That was the Tech Monkey's department, I provided the design skills and a fair bit of pop culture and geek knowledge.

Obviously that didn't last long, and I soon became the proud owner of a Prusa i3 Mk2s kit. Top of the line, loved by makers and allegedly the best printer available outside of commercial industry.

Prior to my career as a world reknowned industrial designer I used to be a mechanic. A grease monkey, if we're going to continue with the monkey theme. Which means I am familiar with working on other designer's products, especially on the diagnostic and repair side. For those of you who have never raised a spanner in anger, you can learn a lot about a designer from trying to fix their mistakes. Some classics from my days on the tools were discovering 3 out of 6 spark plugs hidden under the intake manifold, which turned a ½ hour job into a ½ day, removing engine mounts to replace fan belts, rather important bolts hidden behind sub assemblies, and delicate sensors failing because they were located next to the hot exhaust. Not to mention trying to fit my 6' frame into a 2' engine bay...

What this tells me about the engineers and designers that came up with these solutions is usually that they have never worked on a car outside of a factory, and this shows in their assumption that everyone has access to a hoist, an engine crane, and special tool (Part No. 186439520SQ-666). The problems usually stem from seperate design teams working on components and assembling them outside of the car. Sure, that starter motor fits beautifully, but once it's dropped into the engine bay you can't get to it because there's a honking great chassis rail in the way!

No, I don't miss being a mechanic.

What exactly does this have to do with the Prusa? When I started working in the design field, I carried over the lessons I learned as a mechanic. With regards to assembly of parts and models, I always tried to make them as user friendly as possible, because I know the pain of staring at a recalcitrant part and chanting the mechanic's mantra - “What muppet designed this P.O.S.?”

Here's the meat of the tale. The i3 Mk2s was touted as a game changer for personal 3D printing, and it provided all manner of support in terms of troubleshooting and customer service. To set a benchmark for future Enstaved expansion, we decided to grab 2 kits and build up from there. The kits arrive, and I'm thrown in the deep end building a complicated piece of technology I'd only ever played with a couple of times before.

Now, 2 rules of design I've followed over the years. More parts = more problems, and Tolerances Are Important! The Mk2s broke both of these rules straight off the bat. I do understand the Mk2 was built off an existing design, and there are legacy parts carried on from this. I do understand that this is a highly sophisticated pece of equipment that can literally make something out of nothing more than a spool of plastic, and it needs precise calibration to do that. I don't understand why they carried over the Y frame bed assembly made up of 18 adjustable parts (not including fasteners) that all need to be perfectly aligned for the thing to work.

The Y axis in question

01_y-axis

18 parts give you 18 points of failure for this step. It may have looked perfect on the computer screen during the design phase, but for people buying the kit with no prior experience assembling something as complex as this (and even those that have!), this is very over engineered. Where the Z frame bolts into the bed is a perfect example – a critically important fit that is compromised by too many outside factors (This will be covered in more depth in part 2).

Over engineering is not uncommon among tech oriented people who assume a similar level of expertise from everyone else, but for a company trying to make this technology accessible to everyone it is a bit of an oversight.

Z axis assembly

02_z-axis

That hurdle overcome (2 days later), the rest of the kit went together reasonably well. Sort of. Remember rule 2? Tolerances are Important! Not tolerance as in putting up with a co-workers' excess perfume or love of sharing stupid youtube videos, but clearance between parts to counteract errors introduced in the manufacturing process. Shrinkage of injection molded parts is a big thing, once the plastic cools inside the mold it shrinks a tiny amount, sometimes by only a fraction of a millimetre. Good designers can use this to their advantage or correct for it, modelling a part that fraction bigger to compensate. Ideally, every part would turn out exactly the same as the original model, and I think Mr. Prusa was relying on this with parts of his design.

A 3mm hole on a computer screen should translate into a 3mm hole in the real world, and a 3mm thread should fit down it neatly without touching the sides. This is the theory. In practice, a 3mm hole can become a 2.85mm hole after manufacture due to warping, shrinkage, bad tooling or any other number of reasons. This means your 3mm thread will not fit nicely, in fact it will try and cut its own way down, putting stress on the material around the hole and the thread itself. With 3D printed parts, this shouldn't be an issue, as they should come out exactly as modelled. Unless the machine is slightly out of calibration, the filament under or over extruded, or the slicing program introduced a scaling issue.

Internal housings for nuts to be captured and screwed into isn't a bad idea, having a friction fit to stop them falling out can work well too. But when the friction fit stops the nut from sitting where the thread of the screw can reach it, it causes no end of frustration. The highly detailed (and much appreciated!) instructions are quite specific about screw lengths, but the physical actuality means that trying to start a thread on a 10mm screw either a) doesn't work, or b) puts too much stress on the part. The comment section on the online manual is peppered with stories of people breaking printed parts trying to fit nuts into housings, or clamping them together to start a thread.

There are a few sections of my build where I replaced the specified screw with a slightly longer one after spending way too long trying to start a thread (power supply base, looking at you...). Thankfully, Mr. Prusa added a heap of extra screws and nuts to the kit in anticipation of a few vanishing into the depths of the carpet during the build.

The wiring and cable management were incredibly well documented, with the high resolution pics on the online manual being a life saver when the pixels in the physical manual just didn't cut it (needs moar jpeg!). After a hunt through some of my boxes of accumulated stuff to find a power cord adaptor to suit Australian power points, and few false starts at the calibration stage I managed to get it up and running, and printing beautifully!

Literally can't even

04_powerpoint

Success!

05_printing

Now as stated above, this is the first 3D printer I have ever built. And my history as a mechanic and designer might put my assembly skills a wee bit above a first time hobbyist, but probably not by that much. So when I say that I struggled at times to get this kit together, I'd call it a 50/50 mix of my inexperience and the design issues I've just gone through. I was highly impressed with the detail of the online manual - comments from other builders helped me through a few spots that weren't 100% clear, and the development team was quick to respond to complaints. Workarounds suggested by users were occasionally implemented or improved upon, and the team has updated the guide quite often with new assembly techniques.

And for anyone who has trawled through this looking for hints on the build? Get a set of Vernier callipers, a 5mm socket, a 3mm allen head drive and some snips to go with the tools provided. The callipers were essential for the Y frame assembly - double check all measurements on the threaded rods!, the socket and allen head will make life a bit easier on a few of the longer threads, and the snips are for when you have to replace any of the 2 dozen cable ties you've just put in the wrong place. Also, see part 2 for a few more tips on the Y frame.

So overall, aside from the problems outlined above, the i3 Mk2s is a good kit, and worth the hassle of putting it together. I understand the Mk3 has addressed my issues with the Y and Z frame join, and, as with any new technology, there is always a learning curve for designers – it is a never-ending cycle of trial and error towards improving the design. Now for the multi material upgrade...

Part 2 is up now!

Ben the Design Monkey

Ben the Design Monkey

World reknowned* industrial designer and self confessed geek, Ben is the design guy for Enstaved. Using his love of comics, gaming and pop culture to create the designs we sell. *(in two countries)

Read More