This topic contains 9 replies, has 7 voices, and was last updated by 
Anonymous 3 months, 3 weeks ago.
-
Posts
-
$ 2.799
java (EE), JavaFX, HTML, GIS) programmer, database wizard, framework inventor, looking for a job ! http://roboxing.com/wizardshttp://3dprintingindustry.com/2014/10/17/3d-systems-cube-pro-review/
Check the comments by BillD …
My first 3D printer was the 3D Touch 3D printer, which I understand is based on the BDF3000 and the Cubify printers are the continuation of this line. I know there is a lot of room for improvement from the 3D Touch I had, which required a lot of calibration (almost after each print or each day).
mmm I have been using one of the CubePro Duo’s for a couple of months now. I’ve had zero feed problems with cartridges, if BillD had read his instructions booklet he would have seen a few do’s and dont’s with the cartridges which most likely explain his failure to get them to feed consistently.
Other than everyone’s gripe of locked down software and high prices, I’m relatively impressed with the CubePRO . . .
The Good:
- min layer height appears to be 0.03mm - according to the software and my verniers, not sure why they spec it as 0.07mm - - its practically impossible to see the laminations . . . .but is slooooow at this resolution.
- 100% reliability so far . . over about 20 models.
- does what it says on the tin
- huge build volume
- NO stringing or blobs (they’ve done this with retract settings only, no nozzle valves)
- excellent auto generation of support structures (but not as advanced as MeshMixer)
- Bed moves away from the nozzles between paths, avoids blobbing and creates strong pillar supports.
The Bad:
- My biggest gripe is that you can’t print PLA and ABS together, with a heated build chamber (verified by 3DS Support staff). So if you want to print dissolvable supports (with a U\S bath and caustic soda) you’re left with a poorly laminated ABS main structure . . . yet they still try to flog you a $1200 ultrasonic bath . . . grrrr
- Wall’s won’t fully fill on certain thicknesses, 1.2mm fully filled, 1.8mm fully filled . . . 1.5mm and its a hollow wall with 2 perimeters and little strength . . . 3DS Support inform me that this will be rectified in the future but couldn’t tell me when.
- accuracy to CAD model isn’t exact, you still have to adjust for shrinkage.
- use of a glue stick rather than a heated bed is ok for most prints but I’ve noticed a few curl up off the bed. . . . this could be fixed if they let the build chamber heat longer before printing but there’s no manual way to do that.
- price of cartridges is extortionate . .. but having said that, cartridges for a Stratasys Dimension at my old work place cost £250 each!!!
- a complicated calibration procedure that’s a faff to get right.
Would I buy one knowing what I do now? Yes I would . . but that’s because I use it for my work with clients and I know I can get a reliable print from it when I need one . . .without any hassle or fuss getting settings or hardware set-up correctly (apart from some thin walls!). . . . it just prints what its supposed to, when its supposed to. When you’re charging your time at £x per hour, that’s worth a lot of money over time. I wouldn’t buy one for hobby work . . but not many would for that price!
AnonymousVery interesting Rob - must admit on your point no. 5 of the good, I’m not convinced by needle valves. I don’t know what the cause of so many nozzle blockages is, but the needle valves do not seem to solve stringing or blobbing and they also make it impossible for you to switch slicers, as it seems no one else uses that approach. So any slicer you use has to have the proprietary Robox gubbins added to the gcode for it to work.
Basically retraction is good enough, it’s tried and tested and works fine for many printers. I do wonder if it’s possible to just ignore them (other than opening the nozzle), and try to print just using retraction? Be interesting to see if it would work.
Yes I agree @biscuitlad, but hindsight is a wonderful thing! Cel have done a fantastic job with Robox hardware but the amount of time that must have gone into the valve feature must surely have distracted the small team from perfecting the rest of its development. Everything must be so well tuned and controlled for it to work successfully, it seems amazing to me that they’ve got it to this stage. If they’d continued to develop a proprietary slicing engine with integral nozzle valve control it would have made more sense, but given the pressures to get it to market and decision to use an off-the-shelf slicer, additional complexity vs benefit seems minimal . . . . but then again I’m sure perfecting a retract system with enough power and speed to avoid stringing and blobs creates its own challenges. Hopefully their search to identify a better slicing engine will prove fruitful, especially if it has in-built nozzle valve control.
Anonymous@bespokeproductdesign - yes I imagine retraction took a very long time to get right. But it’s now done, and although printers like Ultimaker 2 went through a long and sometimes painful birth process, they are now producing some incredible prints. Shame is that CEL can’t build on Ultimaker’s experience and use Cura or other slicers off the shelf, without removing the retraction parts and adding their own needle valve commands. Or at least, not without a lot of work, not sure what else causes slic3r’s gcode output to nearly triple in size to become Robox gcode.
Still where are where we are. We have to keep in mind that they have produced some pretty good prints, albeit with handtuned gcode. So in theory the hardware can do it.
TBH I think I’m just letting my frustration get the better of me, I so wish I was printing stuff rather than moaning on a forum!
@biscuitlad, Regarding the code size increase… Apart from the necessary start/finnish, nozzle switching, valve opening/closing and wiping, the AM post processor also ads a LOT of comments. These are probably used during post processing. But does nothing for the printer, and could have been stripped out when post processing was done.
We need an option somewhere in system settings!
(Do you want faster transfer of GCode to Printer: Yes - no comments, No - debug comments left in.
AnonymousCode without comments? No thanks!
Well, I guess the needle valve debate is set to run and run. It might take them a year and several iterations to get them to work properly. Caveat emptor and all that…
You must be logged in to reply to this topic.
