Slic3r vs. Cura

I tested out Cura with my new set-up and I'm positively surprised about the results.
Slic3r is working well but there are some parts of the Gcode generation which leave some ugly artefacts. Especially the change from perimeters to infill always leave a mark on the outside.

Slic3r, still having troubles generating clean perimeters

There is still an argument to be made about Cura missing a lot of the fine detail function. No acceleration settings, no manual control and the simplification of giving only a outer wall thickness and not a number of perimeters are a bit to simplistic for my taste.

But the results speak for themselves.
Except for the little blobs, degenerated layer and bigger gaps in the perimeters didn't happen in the Cura printed (golden) parts. I will give Cura more tries in the future.

Cura, overall much more consistent look

Meshmixer support - a very good alternative

I recently tried the Meshmixer support because I'm tired of Slic3r not being able to calculate its support correctly. Apart from looking really cool it actually works really well.
You could even make organic looking sculptures by using this tree and branch technique.
Below you find a tutorial video how you apply and manipulate the support structures a function that comes in very handy if you know your printer better and only need partial support.

The contact point on the model brake of very easy and are surprisingly stable. The only detail to keep and eye on is that the contact point need to be really on the edge of the overhang, otherwise the underside of said overhang will not look good and have a slope of bad layers in them.

The colored trees are the automatically generated support.

Here the model still hot on the bed.

Custom Cookie Cutters

With my 3D printer we can make any shape into a cookie cutter.

The sky or your imagination is the limit. From sigils like game of thrones, the batman sign, a logo or letters, to your littles nieces scribbles on the wall we make cookie cutter from it.
For more inspiration look here: Thingiverse: Cookie cutters

Very popular last year, the Stark wolf from Game of Thrones

Get your own in 3 easy steps

Step 1: You choose a picture, drawing or form.

We can use almost anything to make our shape. The less colors and the simples the shape the easier we can extract the shape.
We can even scan a drawing from you.

The same wolf as finished cookie cutter.

Step 2: Magic and some code.

I have a script ready which allows me to easily transform your picture into information for the 3D printer.

Special order: Rocket from Guardians of the galaxy.

Step 3: Print it.

The printer handles the rest a normal sized form takes around 1 hour to finish, so we can fulfill last minute orders as well. Your welcome to have a look while its made.
Cookies are of course not the only use for these shapes. You can you it as a stencil for spraying or painting or use it as a mold for pouring silicone.

  

Even in tough and chewy croissant dough, the shape comes through.

Practical things:

• In order to get a good cookie long, thin or ragged pieces will make it hard for the cookie to bake evenly.
• The material I'm using is called PLA (poly lactic acid) its used in a lot of food packaging and is non reactive. I've been using it in my kitchen for years. However its not dishwasher safe as its starts getting soft at around 70°C. I can not be held responsible for your safety and/or improper use.
• If you are unsure about safety google the topic and make an informed choice. Here is a short article about the topic PLA and food safety

Vapor smoothing with Acetone

After my failed attempts to smooth PLA with drain cleaner, the acetone vapors work pretty well on ABS. The model is really soft afterwards by that normalizes after some hours

.3mm layerheight, heavenly smoothed

My hub is open

With this this little Marvin I registered my printer on 3D hubs. You can now hitchhike my printer here. I give special discounts for good ideas!

The holy grail of layerheight .1mm

With my new and improved setup I dared testing how .1mm layer height would look like.

Results are mixed, for some reason the model rips itself apart. I also get the feeling that I'm hitting the lower extrusion speed limit if something like it exists.

First I tried normal speeds (30-40mm/s) with otherwise normal settings and furthermore reducing the retraction to 1mm @ 40mm/s.

Then I tried to increase the output by raising the the layer width to 300%, that looked better.

Pushing this even higher to 400% width and 0 retraction made it almost perfect. This is a simple test piece and I haven't managed a complexer piece yet.

Something useful - the stove knob

In one of the earliest videos I saw about 3D printing and Reprap, they mentioned printing knobs (link below).

A rather unexciting idea until you realize how inconvenient it is to fiddle around with metal pins when carefully trying to find the perfect temperature to not burn your pancakes.

Luckily someone like minded already provided a template (Oven knob fully customizable).
This works great, so great indeed that I can put in a custom marker for the prefect pancake setting.

Stability mod part 1

I finally build Masos stability mod (http://www.thingiverse.com/thing:171418).

I replaced the printed arm with 40x20mm rectangular aluminum profile and used a simple L bracket to connect it to the carriage (grey). The wobble is gone now, but the parts were not printed very precise and therefore the carriage doesn't slide very well on the straight rods. I'm waiting for normal linear bearings to arrive.

Same old, same old.

I had hoped once I bought a high-end hotend I'd get good prints, but the Eventorbot frame is still holding me back.
Although much has gotten better such as the solid infill and surfaces

Left .3 and right.2 mm layer height

However I'm still puzzled by the gaps in the walls.

It looks like parts of the wall just curl upwards. However if I see it during printing I cant really see a wobble it just curls up like it doesn't want to stick on the layer beneath. It also is connected to the layer width, which I had down to 120% for different reasons but it appears that 150% reduces the number of gaps visually.

One glimmer of hope is that the gaps appear to be only in places when it wobbles towards the inside, all outward wobble have not gaps.

I can't really imagine that the bed is jumping up and down when the general wobble is only in the x-y direction, but nothing is impossible.

I'm implementing a stability mod at the moment, so we'll see where this leaves us.

Mending Filament

While I was thinking about a temp controller circuit I had the idea to build a little mechanism where you can put two filament ends in and melt them together.

It could basically be a little metal tube cut in half and drilled out to either 1.75 or 3mm hold in block of heat resistant stone or plaster with a hotend 40watt cartridge and a thermistor attached.

Once heated up you put the end in so some seconds, open it up and have continuous filament.
The same heater circuit could also used in a acetone chamber where the cartridge get replace by a heated bed.

Parts:

Solid metal tube like the 8mm straight rods in our printers.

Temp controller like this come already with a thermocouple
http://www.ebay.de/itm/AC-220-V-Digital-PID-Temperatur-Kontrolle-Thermoelement-Controller-0-bis-400-/281291174288?pt=DE_Möbel_Wohnen_Sonstige&var=&hash=item417e42b190

A 12V 40watt hot end cartridge
http://www.ebay.de/itm/Cartridge-Heater-12V-40W-1-5m-Heater-Cartridge-Heating-Element-for-3D-Printer-/361098962779?pt=DE_Computer_Sonstige&hash=item54132ce75b

E3d hotend under control?!

It took some time but I think I managed to crack the code on how to work with the E3D all metal V6 hotend.
The problem had been that the extruded lines were to far spaced based on my old extruder settings which let me to believe that there was too much resistance in the hotend.
I still have the feeling that the manual pressure to feed filament is way higher than in my old hotend but the main point was reducing the extrusion width. The width is now between 90% 150 of the layer height.

Wolfstock vs CherryPi 3

The CherryPi 3 just got released and from what I gather it looks very similar to the Wolfstock I had a mind. The flexibility of the whole Delta platform allows to exchange elements freely.

http://www.thingiverse.com/thing:344408

• The brackets are the same. 
• For my size I will need the double Vslots, so i use the one from the Wolfstock. 

• The carriages could be exchanged from one model to another.
• The one from the CherryPi looks a tad more solid.

• Whether the belt tensioner is on the frame or of the carriage doesn't make much difference. 
• The Wolfstock has a neat mechanism which will tension the whole upper roller.

• The hotend mount will be adjusted by everyone anyway for the preferred method of cooling and hotend. 
• If I come around to bed leveling ill print me the CherryPi one.

Homemade filament spools

The Filastruder works fine now, so I was thinking about how to get some cheap spools for all the future filament. There are some printed version, which would work if I'm the only user but if I want to sell a bit filament to my local 3D printing community I will need a cheap alternative.

I did not find a lot of offers which would suit my needs for a spool so I glued together two toilet paper roll to make the core diameter bigger and put some cardboard on the sides, drilled holes in the sides and voilà a nice filament spool basically no cost.

I'm gonna test its functionality tonight but it looks promising.

Filastruder working!

Seeing how much hassle I still have with my printer I´m positively surprised how well the Filastruder and Filawinder work together.

Everything comes together nicely and the documentation is good, but the "hobbyists charm" is not lost.

This is our configuration. The Filastruder is angled 45° in order for the plastic to exit in a nice curve. In the final configuration the machine will stand completely upright so that the hot and soft plastic can exit smoothly.

This is the laser meter. I find it the most genius solution of all. The laser project a line onto photocells. Depending on how high or low he shadow of the filament is it tells the winder to speed up or slow down, so that constant tension is applied.

The last step is spooling. On the left side of the blue box is a little servo that controls the arm in a way that distributes the filament evenly on the spool.

There are some minor details still to work on like finding a good equilibrium between temperature, speed and filament diameter, but I´m already printing with self produced filament. Ill report on that next time.

Recycling test

Since out Filastruder is doing so well I was looking into reusing my print scrap.

First I just put the parts in my kitchen blender. Everything bigger than 2-3mm did not get shredded

Then I melted down some pieces in the oven and got a niece sheet of plastic.

 After freezing the sheet for 15 min I put it in the blender again. This worked better, but nearly destroyed the blender. You can see the clear plastic pieces broken of from the blender it self.

Maybe with a glass blender, more time in the oven to get the sheet thinner, and more freezing this will work better, but a blender with a glass case is not cheap to come by.

Fora TVReal-World Uses for 3D Printing

More like that please!

I still have trouble thinking about real world applications that go further than just printer improvements, so a little clip from someone like ForaTV hopefully starts this discussion further.

New ways - the Wolfstock

I´ve decided to abandon the Eventorbot. After I calculated that I needed at least 100€ more to make the machine work properly I checked what different frames would cost.

There aren´t many frame kits but once I looked up the Delta design and also found out how cheap aluminum extrusions are I decided on the Wolfstock http://wolfmanjm.github.io/wolfstock/

There isn´t much published about this machine so I will leave my opinions along the way.

Spaghetti Filament

Coming soon, filament homeproduction...

The fruits of my research on filament making machine will hopefully bear a nice little webshop where we will be selling filament.

Find all update on this endevour und the label Spaghetti
http://printnewworlds.blogspot.nl/search/label/Spaghetti

New pins for Emmets heart gears #Smart Design

Emmet is on it again.
I always had some trouble printing the heart gears with PLA. They just would not turn. He changed the gear ratios and the pins got a mokeover as well. I´m exited to try them out.
You´re not a full fletched printer commander anyway until you printed a heart gear!

My custom direct extruder

I´ve seen some great bowden extruders which even worked with very good retraction settings, the one on the Eventorbot is not one of them.
The first problem I had with it was that the geared motor did only the backward part of the retraction but failed to push the filament forward again.
A new NEMA 17 motor solved that problem and even when directly driven without any gear reduction seems to work fine. I still am not convinced about the use of the huge gear reductions the Wade extrudes bring with them but that a topic for another conversation.

The clothpin held down the motor

I decided rather quickly to go for a direct driven non-bowden design and started with a faceplate which would serve a connector for motor and filament guide. In the first version my goal was just to mount everything but didn´t pay much attention to the forces that are at play here. The way the motor was connected to the housing would push the motor up rather then push filament down into the nozzle.

In the 2nd version I went for a spring loaded filament guide which directs all force down where it belongs. One issue still even in the rather balaced design is that the clamping pressure is between the housing/filament guide and the motor screws. If that is not well fitted the plastic will give in and filament will just slip.



For the 3rd version I had the idea of putting a cogwheel on the motor shaft and adding a 2nd axis with a 2nd drive gear to hopefully give the filament grip from both sides. The would apply clamping pressure only there where it is directly needed.

Wing nuts, simple, cheap and useful

Here is something I like for multiple reasons. Thingiverse user Mooncactus provides us with a customizable option to make wings for normal hex nuts. In my local DIY store they cost ~4€/6pcs, making them my self is much more fun.

Fine thread Z axis - succes

I am carefully optimistic to have managed one of the main Z axis problems.

With the original 5/16" threaded rod I had one deteriorated layer for every rotation the rod did. Which resulted in a gap every 1.4mm. I still dont exactly how this happened but there must have been something in the thread which would cast the extruder to not lay down material well.

Before with 5/16" threaded rod

Here you can see how I not only get a gap and an upwards curled filament line but it also looks like the extruder stuttered.

After with M8 fine thread

Now this looks much better. Since the filament is half transparent they layers are much more visible and the main bed wobble is still visible but the whole wall look much more uninterrupted.
With better quality filament (I really don´t recommend buying "Jet PLA") and a solid color I might be able to actually print decent quality.


Coming up: Speed testing. I want to reach an accelerated 80 mm/s and the print parts to change the whole Z axis assemble, but the fine thread will definitely be a part of it.

Final judgment for filament producers

Comparing all machines brought some interesting insight, but judge for yourself.

Here is a judged comparison (for raw data see below).

	Filabot	Filastruder	Extrusionbot	Filafab
Price in €	475	205	460	850
Crusher	In planning	No	No	Yes
ABS	Yes	Yes	Yes	Yes
PLA	Yes	limited	Yes	Yes
Speed cm/min	12-50/1kg in 6.6h	15-60/1kg in 12h	60-91	120
Spooling	No	Yes +120€	Yes	No
Kit/Assembled	Kit	Kit	Assembled	Assembled
Max 21	13	13	17	17

Red = 1 points; Yellow = 2 points; Green 3= points

Filabot:  Relatively high priced for its capabilities. For the same price the Extrusionbot looks much better.

Filastruder: The big plus here is the price however the specs show that this machine only covers the very basics and the PLA issue troubles me.

Extrusionbot: Has the best cost benefit ratio, combines a medium price with high extrusion speed and crusher and spooling are built in.

Filafab: Leads with some impressive speed, but is also double as expensive as all other models.

These are the raw data.

	Filabot	Filastruder	Extrusionbot	Filafab
Price in €	475	205	460	850
Crusher	In planning	No	No	Yes
ABS	Yes	Yes	Yes	Yes
PLA	Yes	limited	Yes	Yes
Speed cm/min	12-50/1kg in 6.6h	15-60/1kg in 12h	60-91/n.a.	120/n.a.
Spooling	No	Yes +120€	Yes	No
Kit/Assembled	Kit	Kit	Assembled	Assembled

Fine thread Z-axis

I noticed some gaps between layers in all my prints which reoccurred every 1.5mm I quickly assumed that this had something to do with the lead screw/ z-axis because 1.41mm is the pitch of the 5/16" threaded rod.
(See here for details)

I went out and tried to find threaded rod with a pitch of 1mm not only because M8 fits better in our coupler but also because there is much less play between nut and rod. That turned out to be surprisingly hard because even though fine thread rod was easy to find, the nuts weren´t. Everything except a standard nut is somehow not produced.

With half hex rivet nuts and flanged hex nuts out of the game I ordered normal M8 nuts and planned on upgraded my z axis with kkronyaks Floating Motor Z-Axis. As a first step I wanted only to change the part which connects to the bed carriage, ripped my machine apart and noticed that the rod mount does´t have the same width as the original metal mounts.

But then I had a stroke (of genius). Since I have now 100 M8 fine thread nuts I have some space for trial and error experimentation. So I dremeld me some nuts to fit in the holes of the original mount.

At least I managed now to get the fine thread into my machine and when I put it back together I print all parts of the Floating Mount, because the 4 screws in the diamond shape promise much more stability for the bed as the original pattern with only 2. 

Filament producer options - Extrusionbot

The Extrusionbot

A mid priced machine (625$) from a successful kickstarter campaign looks like it was designed for desktop use with a small foot print and a convenient auto spooling feature, claims to be very fast and doesn't seem to discriminate against PLA.

This seems to be it for available filament producers. There some building templates and how to´s for DIY machines and 2-3 more in the development phase I didn´t consider so far.

Next time: A comparison of all machines