I’ve started to get spoiled by being able to cut things with my Glowforge that come out the size they’re supposed to be. Yes, there’s kerf, but for one thing it’s fairly predictable, and more importantly for most of the stuff I do it’s well within tolerance even without compensating.
Yesterday I 3D printed something for 8 hours. Not even that big of a something. My Lulzbot Mini is pretty reliable but even then the first attempt failed because the object was so close to the maximum size of the print bed that it happily crashed into an obstruction while trying to reach the far back edge. I kind of prefer the Glowforge saying “no, I won’t go there”, even if it’s currently limiting the print area. But more to the point, after waiting all that time, I now have a lovely augur and a lovely track. And they don’t even pretend to come close to fitting together. It’s not a matter of filing down the edges, it ain’t gonna happen. So I get to play around with the settings and try again (sometimes deliberate underextrusion helps), or hope I can scale one piece up and the other piece down without causing any other problems, or my go-to-last-resort step of importing the model and trying to tweak it in CAD (not so easy with STLs). I’m going to wait until my order of ball bearings arrives, because there’s a very good chance that the holes are too small for the marbles and I might as well not burn another 8 hours to find that out.
By contrast, I wanted to mount some control knobs in a sheet of acrylic. I measured them with my calipers, typed in the number, hit print, and they fit perfectly.
As long as I am printing something I designed myself I have no problem with dimensions, everything fits perfectly. This is because I set up my printers and slicer to print real dimensions instead of fudging my models. On the other hand most things I have tried from Thingiverse don’t print the right size because most people don’t set up their printers correctly.
This is an important point and why I get frustrated when the suggested solution to these problems is to make the hole bigger in the model. It’s important to me that CAD models reflect the actual dimensions, and any fudging required to make a particular machine comply with those dimensions should be part of the CAM/slicer/postprocessing or whatever workflow understands how plastic melts or laser beams have a nonzero width.
To “set up my printer correctly” in that sense, however, is easier said than done as it conflicts with my other priorities. In an ideal world I want something that is, a-la Glowforge, calibrated at the factory and ready to go so that I can take it out of the box, press print, and have things come out with real dimensions. In my experience, Lulzbot nails the open-box-to-first-print time and overall reliability I want, but their factory profiles overextrude and their outdated version of Cura lacks a few compensation features that could help.
The way to get dimensions correct is to apply simple maths: http://hydraraptor.blogspot.co.uk/2011/03/spot-on-flow-rate.html. Since working that out I never had a problem with linear dimensions. However I only print with a single outline. If you use multiple outlines then go from the outside inwards otherwise it depends on plastic viscosity. See http://hydraraptor.blogspot.co.uk/2014/06/why-slicers-get-dimensions-wrong.html
To get round holes the correct size I use this method: http://hydraraptor.blogspot.co.uk/2011/02/polyholes.html.
Wow, I’m in love with these three links! I’m going to have to read more of your blogs. I love the details! I just got an Anet A8 (I know, not the best printer, but it’s what I got)…does Cura have the ability to make the changes you made in Skeinforge?
Sorry I don’t know because I haven’t used it for years. I still use Skeinforge!
I have an old and tired metaphor I pull out when people ask me about 3D printers and similar things. I don’t know anything about [here I usually say “motorcycles”, but there are many valid substitutes] personally, but from what I observe, there are two styles of motorcycle ownership. Some people buy a motorcycle because they’re really interested in riding motorcycles and they want to spend all their time driving it around. Other people buy a motorcycle in order to spend their weekends taking it apart and putting it back together. Perhaps some of the hard-core ones even have blogs where they share the journey of building one from scratch. These are all perfectly valid.
As for me, when it comes to my making-things hobby, I am interested in the math behind 3D printing profiles, but only to the same degree that, as a curious person, I am interested in how most things work. I have a basic comprehension of what’s going on inside my refrigerator, but I have no desire for opening it up and adjusting the low pressure circuit cut-out setting to be a normal part of refrigerator ownership or something I have to do to get the best results.
So I’ll be very happy when the Glowforge software does even more for me behind the scenes, but right now it’s pretty darn impressive.
Form an old school maker perspective, the consistent precision of computer numeric control is amazing and coupled with a tool footprint of a laser beam this machine thrills me.
I have one 3D printer that I play with, and one for just printing things. With the GF I have become confident enough that sometimes I don’t even bother with a prototype for fit.
I second this comment. I’m still trying to work on some extrusion settings to get Gear Down For What’s robotic arm working. The tolerances are super tight and while I have been very happy with what I design by myself and having it fit together, someone else’s design with minimal tolerance is a test for my setup at the moment. going to work on extrusion speed and see.
I just enjoy making stuff that just pop together and stay there but still can come apart, but then go back together.
I am a 3D printer novice. But I have been designing a marble machine that will use this auger. It needs to slip over the shaft of a stepper motor and a skate bearing slips over the collar of the auger.
I have been chasing down the right dimension. Tweaking and printing just the shaft over and over it is either too big or too small. Obviously I need to start reading @palmercr’s blog to understand my slicer. But I did start fantasizing about laser accuracy.