Avisos
Vaciar todo

My experience with first assembly  

  RSS
lawrence.g2
(@lawrence-g2)
Active Member
My experience with first assembly

TL;DR: MK3 assembly 6 hours, my Ender3 assembly 50 minutes, both from the printed instructions away from my computer. Time to reliable prints after assembly, MK3 a couple of hours, Ender3… still counting.

I have been interested in--and reading about--3D printing for a couple of years. I watched most of the videos I could find on YouTube. Thomas Sanladerer, Angus Deveson (Maker's Muse), Michael ??? (Teaching Tech), Joel Telling (3D Printing Nerd), Simon Sörensen, and others were all very informative. My nephew, scores of years younger than I but with a year more experience than I, also provided some useful information.

Late last year, I thought the time was right. Open-source 3D hardware and software development had matured to the point I felt I could jump in and start experimenting with it without requiring huge amounts of my time.

My research had narrowed my choices to two relatively inexpensive but highly regarded and recommended extruder-based systems: a Prusa i3 MK3 and a Creality Ender 3. With very similar build volumes, supported materials, and mechanical geometry, their prices were surprisingly 4:1. I am fortunate my budget could just handle the MK3 plus a decent supply of filament. I was trying to figure out what features really justified that large price difference of the Prusa over the Ender and which were just fluff. Then I realized that I could order the Prusa as a kit and use the money I saved for the Ender. Thus begins an unintended comparison by a 3D printing noob.

I live in the USA, and I ordered the Ender about a week after ordering the Prusa. DHL and Amazon Prime conspired to have the Ender arrive one day earlier than the MK3.

I started building the Prusa on a Saturday at 6pm and didn’t finish until midnight. My lack of dexterity and aged eyes coupled with identifying unfamiliar plastic shapes and tiny hardware parts combined for an unpleasant experience. Doing the assembly on a tiny table over carpet was a mistake that made inevitable dropped parts hard to find. I think I may have dropped the hex wrenches hundreds of times during assembly, often several times per fastener because they needed to be removed and rotated. If I had purchased a power screwdriver with the right size hex bit, I think I might have shaved a full hour or more off the assembly time.

The Prusa printed instructions were extraordinary in quality and clarity. Had I created them, I would have begun each assembly chapter with an exploded view drawing of the components that were going to be assembled, particularly the extruder. I was amused by frequent instructions to tighten/do/insert something “firmly, but not enough to break the part”. Without first breaking a part, it is really hard to know where that breaking threshold might be. Maybe Prusa needs to include a tiny sacrificial part, so that the assembler can break it intentionally to master that “feel”.

Having all the Prusa hardware parts kitted into small bags ordered by assembly was very much like a Lego kit, normally a welcome organization. However, I would prefer the hardware to be sorted by kind, labeled by their size and type. Sometimes I opened a kit bag and had to figure out which of the very similar bolts within were which by length; I did not have a ruler handy. The terminology printed on the bags did not always match the printed instructions.

I might also have re-sequenced the Prusa assembly so as to assemble the extruder first, when the person assembling the kit is still mentally fresh. I would definitely add one page with an illustration of each hardware part at 100% scale, cross-referenced to the abbreviated name (M3nS) and the full name (square 3 mm nut). All those minor criticisms aside, the Prusa assembly manual remains a stunning example of proper user documentation and a great reference when it comes time to service any part of the machine.

Towards the end of the Prusa assembly, I could not find one small 3D printed part, the upper strain relief. I had checked it off the list when I unpacked the bag, so I knew it had to be somewhere. Fortunately it did not prevent the printer mechanicals from operating. This missing part was what I would attempt to print first.

I made two assembly mistakes building my Prusa MK3. My first error in Prusa assembly was the positioning of zip ties used to bundle the wires coming from the extruder assembly. I followed the instructions to position them slightly to the left, as seen from the rear. As the X carriage went to the extreme right during axis calibration, those zip ties contacted the power supply prematurely halting the carriage, and the firmware reported an error with the X axis length. It took me a while to figure out what it was complaining about because it looked like it had gone as far as it needed to. Reinstalling new ties to allow that last two millimeters of travel allowed the self-test to complete. Axes were considered to be perpendicular, and skew reported to be 0.00.

My second error in Prusa assembly was that--after the printer was complete and passed self-tests--I added some zip ties to secure the ribbon cable to the extrusion in the base. These zip ties would slowly relax over time and later interfere with the U bolts holding the Y bearings in place, causing the bed to hop 0.2mm each time it ran over them. Small parts where the table didn't hit a ziptie printed perfectly. It took me about a week of printing before I even noticed it, and community members of this very support forum here helped me narrow it down in a matter of hours.

During the search for the not-really-missing strain relief, I managed to unintentionally hide Prusa’s beautiful 3D Printing Handbook under a stack of unrelated paperwork, so it was almost a week before I uncovered it. In the meantime, I tried to accomplish the first layer calibration based on what I recalled from having seen it online.

Prusa very generously supplies a whole reel (1kg) of new filament with the printer. I was very worried about crashing the nozzle into the bed, so I brought it down a tiny bit at a time, and kept getting a stringy mess. The desired value for the “live Z” offset appears to be highly dependent on how high the PINDA probe was tightened, and I needed to go down more than -0.1mm before it started sticking well. Going down by 0.080 each iteration took a while.

Once the Z offset was set, I was ready for my first print. Slicing and printing Prusa’s strain relief in PLA was straightforward. The PLA print was flawless. Trying to install the newly printed spare, I finally found the original wedged into a crevice in the door for the controller board. I ended up using the original part as shipped and not using my printed part.

I started assembling the Ender 3 the next morning. It was sold as a kit, but it only had about twenty subassemblies that were each fully assembled. The printed instructions were a single tiny wordless sheet mostly indicating order of assembly. However, the large subassemblies made the instructions somewhat redundant.

It took me about fifty minutes to assemble the Ender 3. Then it took maybe another half hour to adjust the X and Y axis rollers and belt tensions to my satisfaction, as those rollers have the tiniest difference between very loose and binding tight. Bed-leveling on the Ender took another hour to dial in, as every adjustment to one of the corners moved the others a tiny bit, and then the whole thing was too low and I had to restart at the top end of the spring range.

At the low price point Creality is forced to be a bit stingy and only supplies a meager starting quantity of somewhat dirty filament. It was a loose coil and fouls itself into a snarl almost immediately, resulting in one of my first projects on the MK3: a reel to keep such short bits of filament under control.

I have yet to get a really decent print from the Ender 3, mostly at the start of the prints. The flexible magnetic sheet is an ineffective facsimile of the MK3 spring steel PEI plate. I’m suspecting it never quite lays the same each time it is removed and replaced, and the bed level itself seems to need fairly constant attention to keep it dialed in. I have no idea what material the surface might be, but I’m pretty sure it is not PEI.

The CWOTI is to swap in a fixed glass bed option. I feel sorry for all those 3D printing pioneers who forged the path before us, suffering through scraping prints off a fixed bed. I do not want a fixed bed or an inflexible bed after I've been spoiled by the MK3.

Unlike this Prusa forum, I have not found a definitive English-language company forum for the Ender 3. Various subreddits are pretty active and have been a useful source of information, and of course the YouTube community is vibrant.

I will probably tinker with the Ender 3 to try to get it to work better over time, but such tinkering is not a high priority for me as the Prusa MK3 is working so well. Alternatively, it might end up as a “donor”, with selected parts being reused if I decide to try to homebuild something with a little more build volume. I don't leave the Ender 3 unattended because I haven't yet reflashed the firmware to include thermal protection; the MK3 includes that out of the box.

I have generally only printed parts of my own design, the exceptions being the temporarily missing strain relief, and three or four test structures I downloaded from Thingiverse. PLA is a rather forgiving material; I’ve now switched to PETG for mechanical reasons, and I’ve found PETG makes one work harder to get everything tuned on the MK3. I don’t even want to think about PETG on the Ender3.

Although the experience of building both machines yielded me this comparison, if I had it to do over again, I think I would just have gone with the assembled MK3 to save those hours.

For those who have endured through this whole saga, I salute you.

I think the key takeaway is that although the MK3 took more time to assemble, from that point forward prints have been almost zero drama. One printing issue was my error, and the other a defective part for which Prusa promptly shipped a replacement. The Ender3 took much less time to assemble, but the time to first successful print was overall longer. I’ve spent easily more than 20 hours trying to coerce it to repeatedly do the one thing it is supposed to do, and it has a long list of “upgrades” in front of it just to bring it close to what the Prusa does stock. I can appreciate that not everybody can afford a Prusa, but for those that can afford it, the Prusa MK3 is a comparatively polished, reliable platform to start with.

All of the above just my opinion, and I have no relationship with either company or their suppliers other than having purchased their products with my own money.

Respondido : 25/01/2019 3:00 am
Compartir: