AreDigg

My first attempt: 270C, 150% flow, 0,8mm nozzle, 0,1mm layer height, no fan. What I realized here is that the 100% infill in Simplify3D don't work, the pattern makes dots and not entirely clear. The flow work great on the lower part, but it ruins the upper side completely. But it is really clear through the object, except where there is infill. Another interesting observation is because of the increase in flow of material and very good bonding, it becomes incredible strong! This was another attempt with lesser flow and with no infill, because of the need to overhang the top it became "snowy", the walls are however somewhat clear, but you can't see it from the picture. Here what you see as not so clear is actually the bottom. The "coin" is 5 mm thick. As I printed these on buildtak they did not become fully transparent. The red markings are also on the bottom, it is just written the amount of flow as it was different for each "coin" (from 125 to 145%). To solve the problem with the infill I adjusted the Top/Bottom solid layers so they span the entire object. This is printed on a glass plate, which must be the recommended surface, but I think because I printed it alone and with a high flow of 150 you get too much material, and causes blobs, and as well the nozzle makes a deeper groove on the top layer. It seem to be working as colorfabb claims, the idea is basically to squeeze enough material into the previous layer so that they bond completely. However the flow setting is sensitive. I think it can be tricky to use this method for complex objects. Also the pictures don't do entirely justice, it is cool to actually fdm print transparent objects where there is no refraction from the individual layers! I am trying a bigger object now, and really curious to see how it will lay out.

I have lots of HT so I think I will try to replicate this.

I like bridge nylon, it is very practical and strong, however for thin parts it becomes too flexible. And looks boring. If you want something transparent also Bluprint is great and have the high temp like platec. HT is cool except it lacks the colors like other colorfabb filaments. Most people seem to like their prototypes either white or black.

You can read the raw values with a serial connection over the usb.

After tried different options I have some suggestions. What I have used are the original UM2 glass plate, borosilicate glass from aliexpress/ebay, zebraplate, as well as coating like PVA-glue and BuildTak. I see no usable difference from the original glass to the chinese glass, both are uniformly flat. The UM2 glass have however started to chip off more and more, so I will see how much more it will last. Zebra is great for removal and stickiness, however they are not completely flat and when clamped down by the clips they bend upwards in the middle. When I printed with 0.2 mm layer height in the first layer, the extrusion almost stopped in the center. I was able to do some mitigation by bending the plate downwards before clamping, but I believe that the difference still will remain around 0.05 to 0.10 mm. If that is not a major issue in the print, its stickiness is great I usually use it for smaller and medium sized parts where the full flatness is less critical and I love how just bending them after print pops off the parts. When using BuildTak on the glassplate I recommend that you cut and remove off where the clips are, because otherwise the clips are becoming loose as the BuildTak adds some height and I had to get replacement, because when switching back to non coated glass it was starting to slide from side to side. BuildTak can really make objects stick so be careful with layer height. I use pure glass and zebra for PLA, ColorFabb NGen and NinjaFlex, and BuildTak for ABS. For Taulman Nylon and XT I think glass with PVA glue works the best. It also depends on the finishing I want for the underside of the parts, BuildTak makes the best matte finish, while glassplate obviously the best glass finish.

I think in kisslicer the brim become smaller on the upper layers as well, i.e. fillet.

The new fans are much wider. The firmware surely should have a limit, but maybe that is not updated if you upgrade the firmware from old non-plus models? The S3D priming deposit is hardcoded in the startup script, as well as the wipe.

Print Studio is part of the "Spark" framework. It is a standalone program, and are able to slice to g-code as well. You can read more about it here https://spark.autodesk.com/developers/reference/desktop-applications/print-studio https://spark.autodesk.com/developers/showcase/d0716b0d-5cb4-4959-9f10-032be8591c2a

Autodesk Print Studio can make same kinds of support as Meshmixer and is way more stable.

So custom firmware upgrade is definitely buried now?

I believe this is what causes glas chipping.

Probably spelling mistake.

I have also installed the extrusion upgrade, and after a week with various prints, I have come to some experiences. First of all the tinkergnome 16.03 var really buggy for this, which is a pity as it was much easier to see whats going on. To get better prints I had to reinstall the 2+ extended firmware from Ultimaker, and do a full factory reset. A lot of minor annoyances disappeared, however the prints were still under extruded. This is the 3d print test: Looks horrible. However, I decided to try the ultiarjans temp steps test, and the direct g-codes. Interestingly enough it looked much better: This lead me to believe there are something else going on here. I am now going yo try to slice the 3d print test in Cura instead og Simplify3D, which I have used until now. And I will slice the ultiarjan test in S3D and see if there are differences. I have a feeling it has something to do with either slicing settings or firmware. At least I hope so. So far Cura version seem to be much less under-extruded than S3D. What confuses me is that I have tried to increase the extrusion multiplier almost to double the old value and it still is not good. So maybe I have just been lucky with the last prints.

I have the new extruder, and know the problem. I let it push the filament to the second clip on the bowden before I ask the printer to forward to the head. If I just wait till the first, the printer will timeout after forwarding and then waiting for material to reach the printhead. First I thought the new bowden was much longer, but it has the same length. The same happens when retracting the filament out of the printer, it should not need to stop way into the bowden tube.

Finally received the gold-priced cardboard box and installed it. It took me two hours and certainly not a click and snap procedure. The cables and mainboard are delicate, and I even had to disconnect several of the stopper because whoever assembled my original machine had them intertwined in such a way I couldn't get the hotend fan cable out of the Ultimaker. But after wrestling and fighting with the "ikea"-wrench, now the old print head lays quietly on the table like a dead alien. Even took the opportunity to clean up and freshen the printer, so it looks almost like a new one. Now it is time for first test print. Updated to Tinkergnome 16.03. And the new sound from the new extruder certainly is different. So far the print looks very good. So everything seem to be working properly. After it finishes I think it is back to find the optimal settings for Simplify3D and all my materials. Oh and one thing, the "+" stickers should have different sizes so they match both the front and backplate.

Will the newer versions of Cura be able to load custom firmware?

I have used both XT and nGen in their translucent form, and the trick is basically to print with the highest layer setting you are able to and still have they layers stay together. Usually around 70-80% of the nozzle size. And the bigger nozzle size the better. You will never have glass-like translucent, there will always be refraction from the layers and as such is gives a more ice-like feel than clear glass. Also infill will of course be highly visible. And shells should really have only one layer for as low refraction as possible. I use it more to create a unique artistic effect, rather than an object you can see through. Another cool thing is to create thin and flat objects on the glass plate, and these will usually be completely see-through. Also if you need some strength and only semi-translucent, you can use Taulman Bridge Nylon. It is surprisingly translucent, however with a sheen of whiteness (which makes it a little bit more plasticlike than the Colorfabb, a bit like the bowden tube). And when printed very thin becomes rather flexible.

It is not interesting to have semiconductors. By them selves they are just poor conductors. What you want are like p-type and n-type and be able to combine them to form diodes and transistors. Going by trying to rust old metal particles in PLA feels to me like to study black powder to invent a nuclear bomb. To be honest I think modifying your laser printer will be closer to creating IC's easier than your FDM printer. However when that said, I could imagine a two-step process by using a plastic extrusion print head which could extrude some semiconducting base-material and in a next process use another "print head" to do something similar to photolithography to turn parts of that layer into a certain semiconducting type. I know that researchers have created proto-type printers to create electronics, however they are highly specialized and with several processes, and I think it is not viable today to move those into a basic FDM printer.

I don't know if all oxidized metal becomes semiconductors, but of course the MO of CMOS is metal oxide. Modern semiconductors are now of Si, Al, Ge or Ga and doped to make certain types. As there are already various types of conducting filament, surely it should be possible to create something with semiconducting stuff, however it should also be printable. You have to know the "flow" of the electrons.

You can also use Autodesk Print Studio. It has some good functions for manual supports.

From the newsletter it seem like they are only in Black, White, Grays and Clear? I hope that more info will be available, I really want something else than ABS.

Most photogrammetry software improves a lot with better GPU as they use OpenCL. Also lots of photos need a lot of memory. I would start with 32GB memory.

I gave up on 123d, but are now using Agisoft Photo Scan. You get very good results if you take some precautions with lighting and shine. It would be the most cost-effective solution for scanning objects like a car. These videos are excellent resource , , . I use an iPhone 6S Plus, and I believe that the metadata that the smartphone adds provides hints that the software uses to improve the result. But using a DSLR should get you more details. Sharp and "flat-lighted" images are important, and downsizing the resolution helps the software more (to avoid the minute graininess in high-res images).