Page MenuHomePhabricator

A Considered Review: 3D Printing
Closed, ResolvedPublic

Description

I formed an impression of 3D printing by watching popular YouTube channels and browsing 3D printer model repositories prior to purchasing a 3D printer. This impression was incorrect in essentially every aspect.

Expectation: Printers are very fiddly. You will spend a lot of time leveling beds, getting parts to stick to the bed with special glues, calibrating and adjusting settings, etc. You may need to build a special spring-dampened concrete frame for your printer to limit resonance effects. Through extensive experimentation, you must uncover the ritual for Good Printing Fortune. Even experts using very expensive printers with many expensive upgrades will frequently have prints fail and turn into spaghetti.
Reality: The cheapest printer you can buy just works out of the box. I've spent about 10 minutes of fiddling, total? It's sitting on a board on top of a bathtub. Essentially every print has come out perfectly.

Expectation: You need to buy an expensive printer and upgrade most of the printer parts to really get anything usable out of it.
Reality: The cheapest printer you can buy is really good. I bought a $170 printer and maybe $35 of quality-of-life upgrades. I haven't needed any replacement parts.

Expectation: Parts are flimsy and it's difficult to print sufficiently strong parts. You need a library of specialty filaments.
Reality: For a very wide range of reasonable applications, it's easy to print strong parts and you can print them all in basic PLA.

Expectation: Interfacing 3D printing with other materials is hard/complicated (special hot-melt threaded inserts, etc).
Reality: Most of the time you can model a 2.75mm hole in the part and just let an M3 bolt cut threads and it works great. It's generally very easy to combine 3D printing with other fabrication techniques. Parts glue together easily.

Expectations: Support material is bad somehow, and maybe 3D printing is limited in what it can do because of how bad support material is? "Prints without supports!"
Reality: It just pops off in a few seconds. It would take longer to open a box if you ordered a part online. Who cares?

Expectation: Printers have poor precision. You get approximate parts with a lot of shrinking, warping, imperfections, etc.
Reality: Parts are consistent and dimensionally accurate below the 1mm level. Nothing shrinks or warps significantly. It's usually fairly easy to print parts that fit snugly. It's trivial to print a reasonably-precise 5mm hole for most practical applications by combining a 4.8mm hole with a 5mm twist drill bit.

Expectation: Printers are mostly useful for printing 3D printer upgrades, small Anime Darth Vader figurines, and things you could buy a better version of on Amazon for a song. The idea that you can, e.g., print replacement parts for things is largely a myth. YouTube is rife with "30 amazing TRULY USEFUL prints" videos that have 15 Anime Darth Vader figurines and 15 fidget spinners, or videos that print a bad version of a $3 part you can get at a hardware store.
Reality: In a few months I've printed replacement/upgrade parts that significantly exceed the total cost of the printer and all the filament I've purchased (if the part is even available), many of which are "upgrades" in the sense that they're stronger or more useful than the original part.

Expectation: Printers are good at "novelty" shapes (print-in-place interlinked Anime Darth Vader Dragon models) but not good at useful shapes because of overhang, support, infill, warping, etc., concerns.
Reality: It's almost always easy to model parts that will print well. Most of the time you only spend 10 seconds figuring out which face to align to the build plate.

Expectation: Prints don't look very good and have a lot of artifacts, layer lines, etc.
Reality: Prints aren't perfect, but they look good and are significantly more consistent than, say, a "handmade" item would be, which no one has any problem with. This concern is probably relevant only if you are 3D printing a Life-Sized Anime Darth Vader Waifu.

Expectation: Printing is slow.
Reality: The actual plastic comes out of the machine slowly, but printing is faster than ordering a part from Amazon. Sometimes it's faster than driving to the hardware store. Almost any custom replacement/upgrade part is much faster to dial in by iterative printing than by other methods. It's not some kind of Magic Wish Fulfillment Parts Faucet that you turn on and parts spill out, but it's almost always way faster than alternative ways you could get the part.

Expectation: You go through piles of expensive filament.
Reality: Prints cost <$1 each and you can't get through that much filament even if you're running a printer 24/7. Anyone on YouTube with a closet full of filament spools is doing some Life-Sized Anime nonsense with a room full of printers.

Expectation: 3D printing might be okay at making panels for projects with panel-mount controls, and maybe a few other things. Perhaps I can find a handful of things a year to print to justify the ~$200 setup cost.
Reality: 3D printing is great at pretty much everything and I've run the printer almost continuously since I bought it.


Here's a replacement Bernina 740 sewing machine hand wheel that's a little easier to get a grip on than the original one was. I couldn't find this part available online anywhere.

Screen Shot 2022-11-03 at 4.23.07 PM.png (671×1 px, 1 MB)

Event Timeline

epriestley triaged this task as Wishlist priority.Nov 3 2022, 11:43 PM
epriestley created this task.
epriestley updated the task description. (Show Details)

They sell "broken" Ender 3 printers on eBay for like $75 shipped. I ordered two. One had a bad board (a $40 part); one had a bad Y limit switch (a $3 part) -- so the actual cost per printer seems to be around ~$100 if you're familiar with how they work and can identify which parts aren't working.

I'm curious what 3d modeling software you use, I've had success making simple boxes and things like that with sketchup but I've been a little intimidated by more advanced modeling software.

I've been using Fusion 360, which I'd say is "not bad". (See also T13697, perhaps.)

I started with SketchUp since it's popular in woodworking, but it always sort of felt like SketchUp wanted me to use it for interior design, not for dimensioning parts. Fusion 360 has plenty of annoyances but it works more or less like I want it to.

(Note that I've only designed functional mechanical parts -- no clue if Fusion 360 is good for modeling Darth Vader Waifus or not. I'd guess it's not a great choice.)

A year or so later:

  • All three printers are still in service, and I've printed hundreds (perhaps many hundreds?) of parts. They're no longer churning out parts 24/7, but this is just because I can't design parts fast enough.
  • I'm planning to build a Voron next, but time/space are hard to figure out. Not sure if I'll build a 0.X or 2.X.
  • I haven't touched any of the software pipeline (really, I've barely programmed a computer in a year or two?). Carrying SD cards back and forth is my biggest annoyance on the software side, but this is minor.
  • Still using Fusion 360, no complaints.
  • Printing in ABS worked great, but the enclosure is hard to fit into my space so I'm still primarily working in PLA.
    • I built a shed in the back yard to try to address the space issue, but you need a surprising number of nailguns to build a shed and the shed is now entirely dedicated to nailgun storage.
      • The obvious next step is to build a second shed.