Thing's I've learned about owning a 3D printer

Hello everyone, and Merry Christmas!

Just thought I'd take a few minutes and share some wisdom for those that are thinking of building/buying a 3D printer.

I have owned a FolgerTech Prusa i3 for about a month, having purchased it on eBay for around $325. It has been a good printer, considering the price, and has taught me some valuable lessons.

• If you are married, consider having your significant other help you build the printer. Otherwise, forget getting to spend time using it.
• Do not assume that the laser-cut plexiglass holes are big enough for the M3 nuts. Otherwise, you WILL crack the frame.
• Build the printer's structure (frame and axes) on a flat surface.
• Measure everything to be sure the printer is assembled evenly before tightening any of the fasteners.
• Before printing anything, take apart the extruder and hot end and familiarize yourself with the parts. You WILL be taking it apart several times to clear jams.
• Before attempting to print anything, exercise the printer to be sure all of the directions and limits are operating properly.
• Make sure that the software is set to match your printer. (filament size, nozzle size, etc.)
• When you're finally ready to print, make something nice for your spouse (I printed a little heart) to thank them for their help.
• DO NOT touch the hot end (it does leave a painful reminder)!
• Print extra parts for your printer, in case you break something.
• Learn to use a CAD program! Printing your own designs is awesome! (I use SketchUp and FreeCAD)
• Never be afraid to ask questions! Others may have already solved your problem...

There are many other things I could list here, but that would take all your learning away.

-David

Edited 1 time(s). Last edit at 12/22/2014 02:01PM by MindRealm.

Wise and true m8 !

---

Collective intelligence emerges when a group of people work together effectively. Prusa i3 Folger (A lot of the parts are wrong, boring !)

Hey man, I feel your pain. Here is an STL of a "Heart" picture frame for your wife. It's sure to buy you some alone time with your machine. 2 parts, a round base with a round insert of 3 little hearts ( wife and you in big space and kids in 3 little hearts ). This is drawn in "INCHES" so reset your units of import specs. Enjoy.

One very important bullet you forgot to mention.
Building a printer is not plug and play. You will have to learn a basic understanding of programming just to get the dang thing to move.

So true. I just learned that it is very important to remove the filament after you finished printing, otherwise you have a challenge cleaning the nozzle
Good idea about the print for the partner.

Quote
Wiggie
So true. I just learned that it is very important to remove the filament after you finished printing, otherwise you have a challenge cleaning the nozzle
Good idea about the print for the partner.

I never remove the filament unless I am changing colors. You just need to change colors while things are hot.

And a warning, before buying a Prusa or any 3D printer at current state, BE PREPARED TO HAVE ENOUGH TIME to play with it, and a High Tolerance Temper!

---

www.3dfilamenta.com - International Sales Platform for 3D Printing Products

For the 3d printer hobby as any other hobby goes know when to turn it off and put a cover on it for awhile.

aunque no leí antes sabias que la boquilla quemaba e igual me queme el debo calibrando la boquilla.

So true stories about reprap...

---

Collective intelligence emerges when a group of people work together effectively. Prusa i3 Folger (A lot of the parts are wrong, boring !)

AquaticsLive: I totally agree! It is easy for a hobby to become an obsession.

Edited 1 time(s). Last edit at 01/12/2015 03:08PM by MindRealm.

Fen2006: Muy cierto. Está muy caliente!

Edited 1 time(s). Last edit at 01/12/2015 03:07PM by MindRealm.

After all of this caution, though, I just want to add that for me, at least, building my Reprap was one of the most fun and rewarding experiences I've had. I've learned a ton about mechanics and precision, as well as CNC machines in general, with some soldering and low-level programming as well.

Plus, I printed Emmet's gear heart for my SO and she loved it. So there's that.

Good advice!

---

-David

Find me online at:
Thingiverse
Instructables.com
LinkedIn
Facebook

Check out my FolgerTech Prusa i3 (plexi) at MindRealm.net

I would just add this:

Circles aren't always circles.

This was a major pain, since so many of my parts were crap, and I had a lot of my i3 jury rigged with zip-ties holding de-laminated pieces together, and things that were really difficult to keep square, endstops that won't stay still, and belts that don't want to stay tight, I tried to use the printer to make parts that didn't suck so bad. My first set was a complete and utter failure. Nothing fit in them, despite my using measurements taken from STL files used to create the original parts. My 8mm diameter holes wouldn't fit my 8mm threaded rods. Why? Well because the CAD program I was using decided that it wasn't really a circle that I wanted, but a 16 sided polygon. Since the diameter is measured vertex to vertex, instead of flat side to flat side, the result is that the circles have an open diameter about 2.5% too small. Add in a bit of squeeze and Z wobble, and you get something about 0.5mm smaller than you asked for.

So, go and re-draw the parts making the 8mm holes 8.5mm in diameter, and give about the same leeway for parts that are supposed to fit together, and I was able to PERSUADE the 8mm rods to go into the holes, with the persuasion being aided by a dead-blow mallet. So even though my exterior dimensions are very accurate, interior dimensions seem to be not so much.

It would work out so much better if the polygon circles used the dimensions flat side to flat side as measurements, instead of corner to corner. I'd even be happier about the ramifications to exterior diameters, though it's still a pain to work out the sized needed if I want parts that fit together.

I read somewhere that that's actually intentional. Some effort was undertaken to do compensation for the various factors that make internal holes undersize but it was determined that there's always going to be variability and the holes could be fettled if small but if they were big you were out of luck. If you have something that must be net off the printer you'll need to figure out what size to compensate the model.

Edited 1 time(s). Last edit at 01/13/2015 02:38PM by IMBoring25.

If so, that's stupid.

If I wanted to machine things, and drill them out, then why would I even use a 3D printer? It would be then far better to just get a CNC mill and blocks of plastic. If you have to adjust your models to get the dimensions that you want, then the models aren't portable, since they can't then produce the same thing on a different printer, or if they're used for different machining methods, such as the CNC mill. On top of that, if I just know that I need an 8.55mm diameter hole if I want a real 8mm hole, then there's still exactly the chance that I'll end up with something too big as if I modeled it as an 8mm hole in the first place. It makes far more sense to have the printer produce as close to the object that you model as possible, and as accurate as possible.

I remembered where it was...The Slic3r manual.

[manual.slic3r.org]

Many of the six reasons they give are machine dependent, filament dependent, mesh density dependent, or will change with wear and calibration of the machine.

I printed a high-poly-count hole test on my machine as I was closing out my first spool of PLA. My nominal 7.62mm hole (closest to your example 8mm) was 7.56mm. That you had a .55mm error on a hole of similar size to one where I had a .06mm error illustrates the difficulty of trying to code a one-size-fits-all solution.

I've read that. I did miss the bit about people complaining about holes too large though. My point is still that it's stupid to say that too small is okay, the slicer needs to attempt to make things as accurately as possible. Having an OPTION for the compensation method would be better, or better yet, have it as a variable amount, so that it can be tuned to your printer/filament.

My verticle and horizontal overall dimensions come out right, or at least as close as I have measured. <.5mm over 150mm is an acceptable margin of error for the parts that I'm making. Verticle dimensions are as close to exact as I care to measure. <0.1mm over 60mm. therefore, I'd say that my printer is well calibrated.

I'm also measuring after the plastic cools, so shrinkage shouldn't be the issue.

"More material on the inside" could be a part of the problem, but that will always be a very small error, and probably within tolerances that I can accept.

Z wobble and inconsistend filament widths I'll also rule out, as they would also affect the exteriors, and that does not appear to be the case. Of course there is SOME Z wobble, as the i3 frame isn't particularly rigid, but very little when the exterior extrusions are done, since that's at a much lower speed than other printing. (I'm also working on improving that overall, since I'd like to be able to print faster with less wobble.)

For me though it's the circle=polygon thing, as I can do the math for the amount that the polygons which Tinkercad (Which I've used so far as the web-based interface allows me to work on things from different machines, different locations, and not have to carry portable storage creates. I've written a script which takes a layer of gcode and converts it to PostScript, so I can send that to a laser printer. This would have no Z wobble, no filament thickness variations, no corner rounding. The line width is derived from the calculated extrusion width. There is slightly less error than I get in actual printed parts, in that an 8.50mm "hole" on paper would fit the 8mm circle, so about 0.05mm overall difference, similar to your margin of error. What comes out VERY clearly ont he PostScript output is the polygonal nature of the "circles" which was more or less my first point. Circles aren't always circles.

What I'd most want is instead of the corners being neasured vertex to vertex, being measured from side to side. So if I were to ask for a circle 10mm diameter, say that the CAD produces instead a square. Instead of a square 7.071mm by 7.071mm (Giving a diagonal of 10.00mm) which cannot possibly fit a 10mm peg, I'd much rather have a 10mm by 10mm square, even though that produces a diagonal opening of 14.14mm. that would produce results that I could deal with, though I suppose that the opposite problem would happen if I wanted to produce a PEG of 10mm, which would then need a 14.14mm hole to fit inside. I can see that it's a difficult problem, with no clear-cut solution. I guess what I'd like is to be able to choose which circle I want in the CAD application. For the time being, I'm going to start using hexagons for holes, as those are (Or can be) measured flat side to flat side. Better CAD software won't hurt, either, but I'm not shelling out for a copy of Solidworks!