3D Printers

I've recently become fascinated by the burgeoning 3D printer market. It's a brand new industry that's rapidly maturing much like the PC computer market in the 80's.  I decided to purchase a decent quality system to begin learning about the process and help develop an understanding of the issues and limitations in the current systems.  There are dozens of possible choices, with new machines & companies literally springing up every week.  

My initial search criteria were:

  1. Good mechanical design 
  2. Open hardware design if possible 
  3. Good extruder design (difficult to judge)
  4. Large build volume
  5. Dual extruders or the option to upgrade

Unfortunately it is very difficult to research printers in this market at the moment.  The on-line reviews are mostly just regurgitation of the marketing spiels done by people that do not have any practical experience with the printers.  And the few real test comparisons are very limited in which printers they review, and do not follow a methodical quantitative comparison of the test results.  

Given all that, after doing extensive searches, I narrowed it down to the following options:

Makerbot Replicator 2:

Pros:  Established Company. Decent mechanical design. Enclosed Build Area. Dual Extruder Support.
Cons: Closed Design.  Mediocre Support. Smallish build area. 

Cubify Cubex:

Pros:  Established Company. Enclosed Build Area. Dual Extruder Support.
Cons: Closed Design.  Proprietary filament.  Very Poor Support. Smallish build area.

Lulzbot:

Pros:  Good reputation. Open Design.  Decent mechanical design.
Cons: Hard to enclose.  Not convinced of X-axis motion of bed. Long lead time.

Isis3D:

Pros:  Open Design.  Good mechanical design. 
Cons: Hard to enclose.  Not convinced of X-axis motion of bed. Long lead time.

3dBotz:

Pros:  Open Design. Huge build area.  Easy to enclose.  Good mechanical design.
Cons: A bit rough around the edges.

In the end I ordered an Isis3D, but delivery has been delayed do to a design change to a dual extruder head.  In the mean time I went ahead and ordered the 3dBotz.  I was interested in the mechanical design of this system, and the build area was huge.  Moreover, right after I ordered it they offered an upgrade to a 14" bed with a massive 390mm^3 build area.  This gives me a great platform to start experimenting with:

It is a testament to the overall design that my very first print came out successfully:

I've had good results on a number of prints so far:

I'm just beginning to experiment with it, but I've already come to some conclusions:

First and foremost, Z axis homing and bed leveling are critical to repeatability.  One current shortcoming of this (and most) printers is that the Z axis relative home must be adjustable to within sub-millimeter accuracy, but there is no precise and repeatable way to do this.  So you end up spending a long time "fiddling" with the bolts to try lining everything up correctly.  Moreover, precise bed leveling is crucial to get good layer consistency and adhesion.  Current leveling techniques are crude at best.  Automatic height adjustment + level measurement is crucial to make the system easy to use for general (i.e. non-technical) users.  I have some ideas on this that I'm going to try out.

Second, the design of the model you're trying to print is critical.  Most of the models you see printed as examples happen to be ones that fit well with the behavior of the medium.  Once you start running arbitrary models, especially ones that require support structures and more complex shapes, all kinds of issues arise.  Bed adhesion becomes more difficult when there are lots of short paths on the bed.  And overhangs require more careful tuning of temperature, speed and airflow over the build space to produce clean results.  Moreover, the way the software generates supports is very important.  I started by using Slic3r, but ended up switching to Simplify3D for slicing the models because it is much better at support generation.  By fiddling with all of these parameters I've been able to dramatically improve the quality of the overhang areas.

Before:

After:

Third, the bed surface and temperature is critical to adhesion.  I'm currently using a glass sheet on top of the aluminum bed plate coated with a layer of Elmers glue diluted 1:10.  This works OK for PLA, but must be constantly reapplied.  And it doesn't seem to work as well for ABS.  Also, the current heater on the bed has a hard time maintaining temperatures above 70C.  I've ordered an upgraded (i.e. more powerful) heater that should resolve this.  I'm going to try covering the glass with Kapton tape and then sanding that.  I'd also like to try a sandblasted glass surface as well as a couple of other variants.  

If I were going back to purchase a system now, here's what I would be looking for:

  1. Auto bed homing & leveling
  2. Heated bed capable of at least 110C
  3. Dual extruder heads
  4. Large build area (250mm^3 at least)
  5. Enclosed (or enclosable) build area

FYI I'm also researching the concept of print hubs.  I've added my printer to the 3D Hubs site:

http://www.3dhubs.com/houston/hubs/larry

I haven't run any prints through it yet.

 

 

Posted on January 17, 2014 .