UP printers, whats their secret?

Why is the print quality of the UP printers overall better than any reprap or derivate?

At first they seem to have the automatic same-material support all figured out. People using it can use it reliably in contrast to skeinforge.
The other thing is the lack of "bumps" along the outside of the print and most surprisingly there are no visible reversal marks. Their surface quality seems to be even better than stratasys if you take that into consideration.

Putting this at the mechanics might be easy but I do not think this is the case. Even prints of expensive millstraps with linear high precision rails fail to live up to their quality.
I latter think this is the software / firmware they use. Since this is all UPs secret sauce we can only guess what's involved but I once found a hint about it.

There was a direct comparison of the UP printer and an Ultimaker of the same tornado model.
The interesting thing I noticed was that the UP print not only was missing any visible layer wobble but also the surface was smoother than the polygon count of the model, so they are apparently using some technique which results in subdivision surfaces.

I doubt this is done directly in a orthodox fashion on the mesh but rather that they are using some predictive modeling, or curve fitting during slicing. One way I can think of it might be to calculate movements in terms of bizier constructions for the shapes involved. Like the movements and their curvature, the shape of the extruded layers (actually an extruded superellipse) and so on.

Could we get some feedback on this from an owner of a UP printer? Possibly some tests?
I highly suspect that a decent reprap should be able to duplicate the print quality with better software..

Fact: It is not better than any reprap/reprap derivative.

A well designed and properly calibrated printer using Skeinforge/RepG/Teacup will out print an UP! any day of the week.

Most repraps do not because the mendel/prusa design is generally lacking in rigidity, extruder design, and often is not perfectly calibrated. Cheap is usually the aim in reprap world, not perfect. I use a much more rigid and mechanically sound printer of my own design, and the results are much better than the UP!

If you don't believe me, look here: [www.thingiverse.com] All of the parts were printed with SF, RepG, and Teacup on this: [www.flickr.com]

The white filament used in the UP! covers up a multitude of printing sins on camera. Look at some of their color prints here and it becomes less impressive : [pp3dp.com]

Yeah. There is no single ingredient that makes good prints possible.

Here are the basics though:

1. Make the frame as rigid as possible. If you have to put stack of books on either side, do it.
2. Get rid of the springs on the print bed, or put some memory foam there instead, but if your bed vibrates in the slightest when it moves, the print quality may suffer.
3. Eliminate all tension on the filament coming from the spool. It may be convenient to have that five (5) pound spool left and forgotten up on the shelf, but it's tough for the extruder to pull along, bearings or not, and affects the layer to layer depth by pulling up slightly (or not) on the X-axis carriage. If the part you are printing only needs 1000mm of filament, then cut that much off, and little more (I know, even I don't follow this advice).
4. Level the bed, print on glass, print with heat, print nice and slow, and use a fine tip.
5. If you are going to use PLA bushing, make sure they are snug. Warm them up, and pinch them if you need to. This is especially true for the bushings around the extruder nozzle. The residual heat there opens the bushings up in no time.

-Wildseyed-

Quote

The interesting thing I noticed was that the UP print not only was missing any visible layer wobble but also the surface was smoother than the polygon count of the model, so they are apparently using some technique which results in subdivision surfaces.

That may just be the difference between ABS and PLA. Corners are sharper with PLA because it is much more fluid when it is extruded. Or they could be smoothing the g-code path like EMC2 does.

I think I would prefer a machine that prints more true to the model than one that smooths out sharp corners.

---

[www.hydraraptor.blogspot.com]

ElectricMucus Wrote:
-------------------------------------------------------
> There was a direct comparison of the UP printer
> and an Ultimaker of the same tornado model.

Link please?

See also: [www.thingiverse.com] for somebody pushing high-quality on an Ultimaker.

Dave Durant Wrote:
-------------------------------------------------------
> ElectricMucus Wrote:
> --------------------------------------------------
> -----
> > There was a direct comparison of the UP printer
> > and an Ultimaker of the same tornado model.
>
> Link please?
>
> See also:
> [www.thingiverse.com] for
> somebody pushing high-quality on an Ultimaker.
this was it, wasn't a tornado though.
[picasaweb.google.com]


But I think that is shows the quality differences pretty nice, and we probably can assume that the ultimaker was very well calibrated since done by the creator himself and it is a promotional picture.

@Andrew
The pulley on there looks really nice, the best I have seen so far, but when I look at the UP printed ones mendel-parts sent me with my kit they still seem better. Maybe that really is only cosmetically, the difference between abs and pla.
But what's definitely different is the smoothness outer shell at the reversal point. The isn't any bump and I can't even notice it when I slide my finger over it, I only feel it when I know its there... that is a difference.

ElectricMucus Wrote:
-------------------------------------------------------
> this was it, wasn't a tornado though.
> [picasaweb.google.com]

Er.. That picture's dated November 2010. Ultimaker started shipping machines in May 2011..

It's in Eriks gallery, too.. ERIK!!! REPRINT THIS WITH MARLIN!!!

That pulley was the very first print my machine ever did with PLA. I'll redo it in ABS. What color are your pulleys, so it can be a fair comparison?

They are in plain natural abs, here is a photo, sorry for the flash.

So is it worth it to purchase an UP! and start selling parts off of it? Good business no?

owism8 Wrote:
-------------------------------------------------------
> So is it worth it to purchase an UP! and start
> selling parts off of it? Good business no?


Depends on the value of your time. I think a RepRap could produce parts about as nice with enough calibration (or not, mine produced remarkable prints right out of the gate) but would take longer to get up and running. In the end, once you have one printer, you're not going to want to stop there, especially if you plan to make a business out of selling printed parts. It's an add.iction, this hobby. I started out with a nice Makerbot but it wasn't long before I had my eye on some Prusa parts… It'll never end!

I think that it used to be (way back in the dark ages, last year) that getting the software right was a bigger battle than getting the hardware right, assuming you started with a decent kit. Now, not so much. With skeinforge 40+, screwing around with calibration prints and all that is virtually gone. And if you don't like the full SF interface, there's SFACT which can make things even easier.

I wouldn't say the UP! isn't worth the money but their big advantage of being a very plug-n-play machine isn't nearly as big any more.

If you're going to try making money printing parts, you can get a MakerGear Prusa for 30% the price of an UP!. More work to build it and not quite as plug-n-play but.. well.. nearly $2000 less. If you're mechanically inclined and don't mind having to learn some stuff, that's a pretty good deal. Or you can go the Ultimaker route..

Research only costs time.. If you're not yet sure what to get, read more.

Thanks for your input guys. Ibuilt a huxley from Techzone and I know the pains of construction, calibration in SF and all that. It is addicting indeed and an awesome experience which actually helped me land the job I have right now (I showed how technically inclined I was by showing the work I did, its a good thing to include in your resume if you can).

Anyways, yes techzone was the cheapest of all the kits out there, so thats where I really lost because their kit is really bad. especially the main part: the extruder and hot tip (bowden extruder, brass in PTFE hot tip- just oozed out). So i purchased a adrian extruder from a reprapper and its working good enough, but I have to constantly watch the machine because my Z axis stalls or something if it goes too fast.

Anyways, I am selling lens cap holders ($15) right now bc it does not have a big Z so they print nicely if anyone wants one

Thanks for your input.

www.omemon.com

owism8, have you managed to achieve some decent prints on Huxley with bowden?
Can you help me here, because I stuck with SF settings: [forums.reprap.org]

Quote
owism8
So i purchased a adrian extruder from a reprapper and its working good enough, but I have to constantly watch the machine because my Z axis stalls or something if it goes too fast.

Do you use springs? [www.youtube.com]

Am i missing something or what have Z axis speed being too high ... have to do with springs... nvm.

Edited 3 time(s). Last edit at 10/01/2011 10:17PM by NoobMan.

@ devijvers - I do have springs on my bed. but Im thinking of removing them. The thing which is wearing out is the plastic gear, i see it is starting to warp, i think they sent me a bad print. Techzone already sent me a replacement wood gear for the x-axis, so I might ask them to send me a replacement for this y axis. I will remove the springs after I get an alternative bed material (right now its the cheap wood they sent me).

@ DeltaFlight - I did get it to print decently, but it could be better if I used the latest SF software but I cant because the Techzone board is different and techzone did not create something compatible with the latest SF software yet so I am using SF38 or 39 (sorry i forget). You can see my website for prints: www.omemon.com . I use detraction (or retraction) of 24, I know its a lot but thats necessary with the bowden, fw and SF taht I have. Just make sure your bowden doesnt push against the direction of the X axis or else it might make the gear slip on the belt.

Enjoy... Or get a Makerbot or Makergear kit. or even an Emaker kit. Techzone is cheapest and I did learn a lot from that build experience, but it was tough to get all the correct parts and yea so thats DIY .

@ovism8, it looks I made last skeinforge to generate G code suitable for techzone firmware, see [forums.reprap.org]

Edited 1 time(s). Last edit at 10/02/2011 05:18AM by DeltaFlight.

Also the chat site has some experienced users who are always willing to help : [webchat.freenode.net]

@ Delta flight, I just copy past that into the gcode_small.py file? or do I create a new .py file with only that code in it?

Thanks

owism8 Wrote:
-------------------------------------------------------
> Enjoy... Or get a Makerbot or Makergear kit. or
> even an Emaker kit. Techzone is cheapest and I did
> learn a lot from that build experience, but it was
> tough to get all the correct parts and yea so
> thats DIY .


Don't get a Makerbot. I have one, and while there's nothing really wrong with it, it's been at least as finicky as my Prusa and makes a hell of a racket. And a Makergear mosaic will have similar or better capabilities in a smaller package with a lower price.

Yeah the mosaic looks really nice in terms of print, I think all wooden repstraps have that noise factor though. Anyone can share?

My Makerbot's noise is entirely caused by the motors. They make music! But it gets old really fast. MBI's drivers still don't do 1/16 microstepping, and the whole shebang is based on old gen 4 stuff. If I had to guess, I'd say that was the problem more than the wooden frame.

yeah I saw that they use cork plates to reduce the vibrative noise of the motors. I read that that reduces the noise a lot.

Anyways. back to the subject. there needs to be more business out there like the UP! printer business. A device that is cheaper than the UP, but same quality. Time will tell.

> My Makerbot's noise is entirely caused by the motors.

90% of my cupcake noise went away with one of these: [www.thingiverse.com] .

owism8 Wrote:

> @ Delta flight, I just copy past that into the
> gcode_small.py file? or do I create a new .py file
> with only that code in it?

No, it was a patch, it shall be applied with patch tool.
Here is a modified file, just replace the old one (from SF 43) with it.

I printed out (2) T5-8 pulleys in natural ABS this morning on my Fablicator. Stock settings. Took 27 min (for two) with 80% infill, .25mm layer height.

There really is no significant bump where the threads start/end.

Excellent!

---

Bob Morrison
Wörth am Rhein, Germany
"Luke, use the source!"
BLOG - PHOTOS - Thingiverse

I have an up. (Actually I have the first one they sold) Linear bearings and rails everywhere (all axii are belt driven). The aluminum build platform doesn't hurt. The big thing is software and plastic though. They don't use regular abs it is some kind of polycarbonate abs blend. It extrudes at 270C. Also I think the filament feeder is a big help too. The software definitely deserves a good looking over as well. As soon as I get my makerbot dual extruder build done I am going to feed it some up filament and see how much of an impact their proprietary plastic has. If anyone want me to print some comparison parts let me know what settings you want them at...

I'm up for some friendly competition

The natural ABS plastic I have is some of the Chinese stuff from esunpla.