My K8400 story

I purchased a K8400 3D printer in order to realise some projects using this new 3D printing technology.

As a newby in this area I got trapped in the different specialties and problems which are partially coming from the software, partially coming from the hardware and finally coming from the misunderstanding of several approaches established in the 3D printing community..

To understand the solutions implemented in hard and software I find it essential to have a good impression on the theory of the 3D printing. I did a comprehensive wrap up here.

In the very beginning I was very busy to find out the weak points in the hard and software to get good, not only good looking printouts. My intention was to get parts which can be mechanically stressed, which was the more critical approach.

I had to elaborate for myself a way to determine the proper extrusion temperature for the selected filament as well as the optimum printing speed.

During this course did several modifications on my K8400 printer.

One essential is the modification of the heating block design to allow a proper thermal control of the hotend. Together with the modified firmware an improved thermal control could be realized.

The basic modification is the implementation of an radial buildt thermistor like an EPCOS NTC. The modified heater block is described here.

I replaced the extruder housing by self printed parts coming from the community. (http://www.thingiverse.com/thing:1133248)

I added a z-axis holder to reduce the sound during printing (http://www.thingiverse.com/thing:635751).


However, my aim is to have a printer which is capable for the most variety of materials available on the market.
I did several successful trials with the following materials:

 - Vellemann PLA white
 - 3DK PLA  Metallic
 - Broncefill
 - Brassfill
 - PET T-Glace red
 - Flex-1000-TPE50 yellow
 - laywood
 - Carbonfill (with ceramic nozzle and polyimide coated bed)
  - PVA (with Polyimide coated bed)

Material prints still pending:
 - PC-750 clear

Initially I updated the firmware to be usable with newer development kits (tested with 1.6.6). I switched to arduino-eclipse IDE because the arduino IDE, especially the old revision, is a nightmare for programming. I removed some errors in the setup (like steps/extrusion) and implemented some of the newer features from Marlin 1.1 together with some additional extensions coming from my analysis. An pressure advance mechanism is implemented as well as some menu entries to modify the new parameters from the display. In the meantime this activity has expanded to a separate firmware flavor: MarlinHOH

Beside the modification on the printer side I became aware of some weaknesses of the slicer. I selected very early the Slic3r as my standard slicer and had to implement some modifications too. The modified extrusion model is described here.

So, what comes next? I'm still keen in printing Carbonfill filament and PVA. This are in fact two different approaches. The first is related to the nozzle, the second one to the coating of the extrusion base.

So let's start with the first issue. According to some statements in the net the carbonfill material is capable to degrade the the brass nozzle in short time. Therefore I tried to use some steel nozzles. Unfortunately there are limited suppliers for these kind of nozzles and they doesn't fit in the Vellemann design. So I designed a new heating block which should be able to fit the steel nozzle. I did some experiments with this design before I withdrawed it. I learned a lot of small details on the way how the thermal design can be selected. For the steel nozzle I found an internal drilling diameter of nearly 2mm for a 1.7mm filament. This is generating a large reservoir of filament, but you have to take care that the molten filament will have sufficent thermal conductivity to liquefy the hard part of the filament in the middle of the reservoir. I could get a short, fast extrusion, but then I had to wait for the next molten material.

As as consequence and a step forward in the material selection I started to design my own nozzle: a ceramics type of nozzle.

I did some experiments with these new ceramic nozzles and was able to print the carbonfill material and other high temperature filaments.But with the carbonfill material I faces some bad experience: I couldn't be removed from the BuildTak foil. I could do some prints with poor first layer, which I could remove from the buildak surface, but with proper z-alignment and a googd, homogenous first layer if becomes nearly impossible to remove the parts. I damaged my BuildTak foil very hard by the trials to remove the prints.

This led to the second issue, the proper selection of the print surface. When printing most of the parts (small test pieces, ), the BuildTak foil is a very good surface for printing. After my initial trials, where I damaged my original part step by step, I tried several alternatives too. Glue layers, hairspray, some kind of blue tape, which didn't fit my expectations. So the Buildtak foil remained the first choice except for two material: Carbonfill and PVA. The first one adheres too much, the last too less.
The best solution I found is a Polyimide foil (e.g. Kapton or similar trademarks). Cleaning this foils before printing using a OH-free solution (like Acetone) gives an excellent surface for Carbonfill material and PVA too. So with this approach I could continue my trials to print with two nozzles at the same time: one for base material, one for support material (PVA).

So finally I could manage to print carbonfill material and PVA together. There was still the problem of printing ABS or other materials demanding a heated bed. After some "silly" presentations in the Velleman K8400 forum I build my latest modification: a heated bed. Due to missing other solutions a light heated one.

Since this summary is a work in progress I would appreciate some feedback and comments how to improve.