Time-lapse video link of Mr.Yoda being printed.

Time-Lapse video of Mr. Yoda being printed at 50 micron layer height with speeds of 30mm for all except perimeter 40mm, infill 20mm and 200mm rapids. MakerBot black ABS filament 1.75 with .35 nozzle. 245c at the hot end and 50c at the bed using CA.
[youtu.be]
Have a great day!
Werner

Edited 1 time(s). Last edit at 04/01/2013 02:49PM by BerryBot3D.

Werner,
After watching this video I realized that there is something very important to be learned from your machine. I've been reading the forums for a couple of months now and looking at all the different designs people have. However in almost every case my brain screams "ITS NOT STIFF! ITS NOT STIFF!" I've watched the so called progress from threaded rod construction to aluminum extrusions and just when it looks like they are going in the right direction you see spaghetti noodles in the design that completely destroys any possibility of real stiffness. But how do we demonstrate that stiffness is an extremely important factor in good 3D printer design?

I am also amused that most of the demonstration prints every does are almost always of rather organic shaped objects with lots of curves, which conveniently hides the inaccuracies and problems of most machines. Your Yoda figure could be no where even close to the 3D model in shape and size, but no one could possibly tell. I would guess that if you asked your machine to simply draw a straight line it might not do an exceptionally straight line over it's entire stroke. The kinematics are such that you actual motion versus step size varies from one end of the stroke to the other and it may be difficult to actually position to an exact spot.

Your machine though has zero backlash mechanical joints using the magnets. Your slides have adjustable preload too, so no backlash there. Finally, timing belts on precision pulleys and well tensioned, also exhibit very low backlash. Your compression struts are extremely stiff carbon fiber tubes. So you machine is very, very repeatable. Repeatability in motion systems is usually way more important than accuracy. Your frame is also quite stiff, although I would prefer to see the upright members rotated 90 degrees, as they are probably about four times stiffer in that orientation. So how do we know that all this stiffness is important?

Looking at your Yoda we can see that the layer to layer alignment is extremely good. You even commented on how smooth the surface is. The pictures clearly show that the layers are nearly imperceptible. So it should be clear to everyone, that the one major difference between your printer and all the others is simply that it is very stiff. Any cartesian style machine will easily produce greater accuracy, but it is much more difficult to make them stiff.

I can't wait to get my own Ifactory machine completed and do some printing with it. My machine is both extremely accurate and stiff, (about $4K of precision slides purchased surplus is why) and for a homebuilt CNC milling machine should be very fast (big steppers with high voltage power supplies). However compared to a standalone printer it will be very slow (axis weights of 110lb, 65 lb and 30 lb). My expectation though is that the print quality and and accuracy will both be exceptional. I hope my printer will reinforce how important stiffness is to getting real quality 3d prints.

Gary H. Lucas

I think his machine is quite accurate. The goal of the delta machine is more for models anyway.
No on ever claims that precision is the #1 goal of 3d printing. Fast prototyping and custom modeling are where the technology strengths are at.
Injection moldings are not precise either. If precision is needed, then ask NASA for a vendor that can help. I'm sure the price will be rediculous.

James,
You've taken my comments the wrong way. I was pointing out that stiffness and repeatability are far more important than accuracy. I even noted that my own machine is very accurate and also very expensive.

Now as to the accuracy of the Berrybot. That is going to be the sum of the location accuracy of the 3 uprights + the accuracy of the bent ball mounts + the accuracy of the ball diameter + the length accuracy of the push rods + the accuracy of the bent extruder plate + the extruder ball diameter accuracy! Now put a dial indicator on the extrusion nozzle and sweep it along a straight edge and lets see how accurate that path is. However, touch the dial indicator against a fixed point, move some place else, and come back to the same point and the Berrybot will be dead on, repeatable not accurate.

Gary H. Lucas

very nice print!

What does cartesian style machine mean btw?

What's that clear stuff over the blue tape? Is it heated? I've never been able to print ABS on an unheated bed. The only exception is when I printed on an ABS sheet and the piece never came off.

Quote
brnrd
What's that clear stuff over the blue tape? Is it heated? I've never been able to print ABS on an unheated bed. The only exception is when I printed on an ABS sheet and the piece never came off.

Werner Berry's, "Easy to apply glass juice for 3D printing"

Thanks. I missed the part about the bed being at 50C. That's quite low. What's CA?

Quote
brnrd
Thanks. I missed the part about the bed being at 50C. That's quite low. What's CA?

Cyanoacrylate or contact adhesive