Need Help with Design Choices - Vertical or Horizontal Printing?

Quote
dc42
If the part will definitely be satisfactory if printed vertically, I suggest you use a delta printer. They are especially good at printing tall parts and the mechanical construction is simpler than for other types.

Quote
Don McLean
@dc42 have you seen a duet board being used on a DLP printer?

Quote
Don McLean
....................... @ deckingman, it looks like you don't have any problems with peeling from your large bed? Any comments on the chances of prints lifting from the bed based on size? Thanks for sharing the vids BTW.
..............................

Quote
Don McLean
Let's say for argument's sake 2" x 2" x 30" as the worst case scenario.

Quote
o_lampe
Wouldn't it be easier to ask a professional maker shop to print these things for you?

Quote
o_lampe
Or make the parts in two pieces and glue/screw them together?

Quote
o_lampe
edit: I confused cm with inches. A 30" long part would be taylor made for the "blackbelt"-style printers. Their layer orientation is around 45°.

Quote
DjDemonD
I like your go for it attitude Don but in Olampe's defence 30" long parts are on the extreme upper size for reprap style printers.

Quote
DjDemonD
I spent £1000+ and 50 hours of my time to make a very accurate delta that can print 270mm circular xy, and 460mm in z. But I can't print your parts in one piece,

Quote
DjDemonD
The idea you might want to be that guy is a great one, but don't expect to be making 30" parts on a $250 printer unless you've got some new ideas to bring to 3d printing to help you do so.

Quote
o_lampe
...
edit: I confused cm with inches. A 30" long part would be taylor made for the "blackbelt"-style printers. Their layer orientation is around 45°.

Quote
Don McLean


The big question in my mind is are there any other effects when you print larger objects, particularly narrow ones, horizontally, as in mechanics specific to 3d printing, like once you get above a certain size it's hard to stop lifting at the edges, etc? If so, that would be a true upper size limiting concern for a printer, and that's something I need to understand when designing one.

Quote
Don McLean

digital dentist, I've looked through your blog, thank you for sharing! I agree with you that a belt isn't a bad choice for a Z axis. I looked at your stretch calculations, and I think that the flex couplers that attach the leadscrews on many 3d printers have the potential to stretch far more than your belt system does. It's interesting that you only needed two of the good quality rail blocks for your Z axis. I looked at your videos and your printer performs very nicely.

*someone recently asked me about print quality toward the top of tall prints, fearing that the print may wobble a bit far away from the bed plate. Any play in the Z axis bearing blocks will be amplified by the long lever of a tall print. If one were really concerned about the top of a tall print wobbling because of the Z axis bearings, adding a second bearing block to one or both rails would probably stiffen things up. The large white lamp is the tallest thing I've printed at 500 mm, and is a poor model to use to check for that sort of problem but does illustrate a different type. If you watch this video carefully, at about 24 sec in you can see the top of the print being pushed a bit by the extruder. This is due to filament curling a bit combined with the height of the print making it less than solid. A print with infill would probably not have the same issues, though curling filament can always cause the nozzle to drag a bit. The next tallest thing I've done is a filament spool holder that's a bit over 300 mm tall. I posted some macro photos here of corners of the print to check the layer registration, because it should be obvious if there are problems there first. image #9 is near the center of the spool holder, adjacent to the small screw you can see in the other photos, about 150 mm above the bed surface. The other macro images are right up at the top of the print. You can see quite a bit of ringing- I don't recall the exact speed and acceleration settings in this print, but IRIC, after this print I may have reduced acceleration and junction deviation a bit to reduce ringing.

Getting into FDM printing, especially with some specific things in mind where quality is important is a bid daunting, because I see some YouTube vids where FDM printers are making parts that almost look similar in quality to some DLP or SLA parts while other videos (not yours) show people merrily printing things that look pretty awful. It's hard to know what you will end up with at the end quality wise.

* I think you can get quality prints, even from a crappy machine, if you're willing to put in the effort and willing to print slowly and wait for it to finish. A lot of people don't seem to mind tweaking the printer for a half hour every time they print, and some will even watch the print and tweak settings on the fly, but very few seem willing to print slowly enough to get a good result. Still others are obsessed with print speed over everything else, resulting in youtube videos of machines printing bird's nests at 600 mm/sec. I don't like tweaking and I don't mind waiting for prints to finish, and I like 40 hour long prints to make it all the way to 40 hours, so I build and configure accordingly.

The big difference obviously between a belt and a leadscrew is that the belt can move things quickly but the leadscrew is better at lifting loads. You have cleverly fixed this problem by adding a worm gear, so well done.

In your firmware, I assume that you can set a lower acceleration and speed value for your Z axis? As the Z only moves in tiny little steps, it seems to me that you'd definitely want the acceleration settings dialed down on the Z axis relative to the other two axis.

* I use 700 mm/sec^2 for XY (pretty heavy X axis) and I think Z is set to 1000 mm/sec^2. I don't think the acceleration in Z is nearly as critical as in X and Y, as long as you don't set it so high that it starts skipping steps. The load on the Z motor is pretty low, so it can easily follow the acceleration. Junction deviation in Z is set to 0 and I run all the Z axis motion at 20 mm/sec. The 1.8A motor gives up at 30 mm/sec (with the 1A that I use) but works reliably at 20 mm/sec.

I can see you've used some small pillow blocks in your design, are those the inexpensive pillow blocks from EBay? How are they working for you? I am also thinking of using EBay pillow blocks to do something similar for my gantry movement run from a single dual shaft stepper. With the dual Idex heads I think I will stay away from the Corexy but also I think I will use just one motor to move the gantry, perhaps a Nema 23 running on 2A.

* the pillow blocks are cheapies from ebay and they work fine. The bearings can pivot inside them to allow for minor misalignment. IRIC they were $6 each.