yellowshark

Once stabilised, maybe 5mins thereabouts after hitting target temp. then variation should be +/- 1 deg. What I have found is that if I need to change temp during a print then it is best to move to the new target in 1 or 2 deg steps. If you jump 5+ degs then it sets off a wilder variation.

Another thing to try is different PLAs. I am not sure of the right words to use, but if you change filament for oil, some oils are really thin and runny and others more thick and blobby. Different filaments are like this (maybe not to the same extent true) but whilst I avoid small like the plague, I have found that some filaments are better than others at the tiny stuff.

Sorry I do not use S3D,I did a quick search on the web but could not find any reference to "mode retraction". There are people on here who use or have used s3D. if you do not get an answer it might be better to start a new thread with S3D in the title and you will probably grab their attention. I think the S3D stuff tends to go to the software\other folder, maybe the answer is there already?

Two weeks ago I went to the offices of Fuel 3D (in UK) to look at their “scanner” and take some test “scans” away for further analysis. Having failed miserably with a Kinect type scan system, I found the F3D very easy to use. I got a perfect “scan” on my 2nd try which no doubt would have been the 1st try if I had put the “scanner” on a tripod. The resolution looks very high, raw resolution is circa 500,000 to 700,000 facets. The software, which includes automatic stitching of multiple meshes, still has some way to go but seemed useable to me. It lets you change the resolution before creating a mesh file, which I did for my tests, reducing the resolution to 80,000 facets. You can also crop the “scan” photograph before creating the file. I use the term “scan” because F3D is not a scanner, it is a stereographic camera backed up by sophisticated software. So for many things it is OK but if you, say, want a scan of a head then you will need to take at least 5 “scans” and then stitch the meshes together, which with their automated stitching function should be very easy. I have tried this in Meshlab but singularly failed. An important test for me was of a foot print made in casting material; in effect instead of taking a scan of the exterior of something, you are scanning the interior. On first look this looks very good, although I do need to do a dimensional check to ascertain the accuracy; but it certainly picked up the surface detail within the cast. I hope to do my detailed analysis next week and will try to post some pics. My guestimate, based on comments, was that right now they are shipping to their Kickstarter funders. They should be shipping to their retail outlets in 4th quarter. So if you wanted one now you might be able to get one via retail if they have not all been pre-ordered; otherwise I guess you are looking at 1st quarter next year. Personally I would be waiting until next year to give the time to get the software debugged. Oh one comment was that though undecided it may be that the automatic mesh stitching would ship with the Pro version.

It really is worth trying a different material. I have just completed printing a chess set; for the "white" I wanted to use Faberdashery Buttercream - never used it before. Whatever settings I used I suffered delamination between the outer walls (never seen this before). As soon as I switched to their Arctic white, perfect.

Lol I think we need a Curley alert thread somewhere.

On curling, my experience this week shows me that the filament itself can be one of the causes. I was using Faberdashery gold and was getting lots of curling - to the point that I stopped and did an atomic pull. Today I spent 6 or 7 hours printing with their Arctic white and did not get a single curl. I was printing the same pieces in a different colour so used the same settings - except for extruder temp which was 5 degrees different

Hi Surfer, I am picking up on the word "jams" in your post. I am assuming that you have to unjam the feeding system to get started again. On that basis I am thinking that with multiple models you may well be doing a lot mode retraction than with a single model. It may be that this is causing your filament to get chewed up, leading to the jam. Next time check the filament, in the Bowden tube, between the drive mechanism and the extruder. If it is really rough with teeth marks then that is probably your problem. On my printer I would take some pressure off the drive mechanism but not totally sure about the UM as this area has historically had its issues; other members better placed than myself to advise.

I would - ensure shell thickness is a multiple of nozzle width, e.g. 1.2mm for a .4mm nozzle; fans on at 100%; slow down, probably 30mm/s or slower

Could be this might help... Taxonomy of Z axis artifacts in extrusion-based 3d printing https://www.evernote.com/shard/s211/sh/701c36c4-ddd5-4669-a482-953d8924c71d/1ef992988295487c98c268dcdd2d687e

Indeed. I always manually feed 30-40mm of filament before starting the job to make sure all is OK. And I watch the1st layer go down before leaving the printer.

Yup me too. Dedicated laptop lets me control the printer during production with the Repetier-host software. Like elegant design, easy is best even if more expensive

PLA only. You need to be careful with a Dremel as it will generate heat. I have only used a Dremel once, with the paper based abrading insert on a surface that had support scarring. It certainly smoothed the surface (required) but the end result was not pretty (not a problem). I use one technique which gives excellent results. Typical reason would be to smooth off artefacts from corners. I use very high grade files, from the jewellery industry, and lightly, and I do mean lightly, file them off. This will depend on the filament but using black PLA as an example - the filing will remove the sheen from the plastic. I then apply with a rag some Baby oil over the area which makes it nice and shiny again.

+1

... I should add that there could be other reasons why it is happening so posting your settings would be helpful

In my experience with cylinders this is a fairly common occurrence with Cura. You need to do one of two things; print real slow - no faster than 30, 20 is preferable. Or do it with Slic3r. If you watch the print real close you will see why it is happening. For whatever reason it does not happen with Slic3r, at least to anything like the same extent.

... actually make that 15-20mm. If the job came to normal eoj then retraction will have happened so 10mm may not be enough

Looks to me that your filament has got stuck in the feed system, reasonably unusual for Colorfabb but not impossible; my Colorfabb black stuck today - first occurrence in months. Warm up your extruder to 200, give or take, and manually try to extrude 10mm. Probably nothing will extrude - existing filament in the head may leak a bit. Remove the filament, use the atomic method to remove, which will clean the head. Cut off the filament where it joins the feed system, then reload and off you go - hopefully!

I have no direct experience on this but I would say that if the printer is fully enclosed and you wait for the internal environment, not just the print bed, to warm up then it should be OK. If dimensional accuracy is important then I have found recently that leaving the object in the printer case for longer, to cool down, reduces the shrinkage; possibly even more important in your scenario.

Hmn, I have just been able to access your pics. On the first ones the green gherkin has a negative indent which starts say half way up. Personally I never experienced a z scar like this, the line has always been positive. I am wondering whether this this is a fault in the mesh rather than a z scar? The white/clear round thing looks a pretty good print to me!!

Hi Erwin Select the print settings tab and then the Layers and Perimeters menu option. Scroll down and you will see Seam Position, under Advanced. Set this to random to avoid the Z scar. Please note thatSlic3r was incredibly slow at generating the gcode compared with Cura which is almost instantaneous. IE for a complex model Slicr 3r could take 10minutes. It may be that vs 2.n variants are much faster.

George's last point I think is very important. When I first started I used Slic3r because it has loads of settings options and different speeds for everything and naively I thought that was cool. I then moved to Cura because I found the Slic3r extraction process to be poor compared with Cura and I certainly found a better quality. Over time I have now moved to having no changes in the settings at all - i.e. level 1 bed temp and extruder temp are the same as the rest of the model; no changes in speed for anything. Two caveats to the above, if the geometry tests the bed adhesion, eg a large flat bottom covering most of the build plate then I will start with a higher bed and extruder temp. If the geometry ends with a point, as well as probably printing two up I will slow the print speed as I get near to the end. Funnily enough I am using Slic3r right now; I have some smallish pieces running at 100 microns and 20mm/s print speed and I was getting a z scar on one of the vertical edges, so hoping that redoing it with Slic3r will remove this. If it works I will post the pics. Also and this may well differ with geometry but with the pieces I am working on at the moment I am finding that 100% infill gives a faster time than using 40% or 20% infill.

A couple of points. Your shell thickness should be a multiple of your nozzle width, which I assume is 0.4mm. Now if you cannot resolve this and it is important for you - you will probably find that the slicer Slic3r will give you a better result on this point as it lets you start the new layers at random points which stops the pronounced line being formed. I prefer Cura these days but Slic3r is not bad and I do normally use it for cylindrical objects to avoid artefacts on the external surface

If you are struggling with glue (which I have never used so cannot comment) and do try hairspray on glass, it really is effective and once the glass has cooled to 40-50c it just pops off. I was advised to buy extra strength hairspray and that is what I use. I apply a light covering over the bed and wait 20-30 secs and apply a second light covering. No idea what impact it has if any but I do wait for the bed to warm up before applying the hairspray. Sometimes I just do the centre of the bed if I have nothing large scheduled. I will normally go a week minimum before applying more spray. And I will often but not always clean the glass under running warm water before applying another coat

Surprisingly Brasso is good at cleaning brass- it is what horologists use