Monthly Archives: May 2011

3D Modelling

Planetary Gear Mechanism on Thingiverse

My first upload to Thingiverse!

Error - could not find Thing 8744.

EDIT: updated with link as the WordPress Thing Plugin seems to have stopped working:


Finding Problems, Resolving Problems

An eventful few days, largely dealing with wear and tear maintenance issues.

X End Idler Split

It seems that whilst using lubricant on the Z axis rods was a reasonable idea it soaked into the X End Idler and caused weakness between layers (I believe). This in turn resulted in a slight split and the head of the hot end ploughing through the last few layers of a print.

Attempting to realign the print (of a gear) seemed to work fine, but closer inspection shows it was pretty skewed and is not usable as it stands.

Luckily the split didn’t grow, or affect subsequent prints, so I could print out it’s replacement.

Perfecting the Pulley

So my circles were coming out oval and I still hadn’t printed the perfect pulley.  Suspecting that a replacement pulley with a dent in the teeth was the cause of the oval problem I brought the original out of retirement with a quick hack.  This allowed me to print a few pulleys, and this time I achieved pulley nirvana.  The secret? Cool in Skeinforge.  It seems obvious now, but my previous attempts to utilise the cool module didn’t work because I was using “orbit”. Changing this to “slow down” meant that the spurs and teeth were laid down much slower, and hence with much less deformities!

Loose X Belt

Having more faith in the X pulley I was expecting perfect(ish) circles, but no joy.  A bit more observation made me realise that the X belt was wobbling quite a bit at the apex of each turn. I had thought the belt was tight enough, but it seems it was not – and tightening it so it “twangs” when tapped results in much better circles.


Hot End

Bamboo as an Alternative to PEEK or PTFE


PEEK and PTFE are too expensive to use in experimental hot ends and quite difficult to come by.


  • Cheap
  • Readily available
  • Easy to work with


  • Irregular diameter
  • May have properties yet untested that make it unsuitable
  • (added 4 May 2011) Friction between filament and bamboo makes a squeaking noise. Doesn’t seem to affect throughput or quality though. Update 2 (added 11 May 2011): The squeak seems to have gone. Perhaps now the bamboo is drier than before and this has caused it to go?

Build Process

  • A piece of bamboo that cut from a plant last year was used, but one could find something suitable and cheap from a garden center. The key is to find bamboo thick enough that will allow a hole wide enough to hold your nozzle to be drilled in.
  • The size of an individual segment seemed to work well, and simply cut at the first knuckle.
  • A 3mm hole was drilled the length of the piece for the feedstock entrance, and then the nozzle end was widened by 5mm part of the way down (however far the nozzle should be screwed in).
  • The feedstock entrance was reamed slightly to provide a wider opening for the feedstock to find its way.
  • Wetting the bamboo makes it slightly flexible, so before screwing in the nozzle the entrance was soaked under a tap. This may or may not really help, but so far the nozzle in the test pieces have remained tight and secure.
  • Several means of attaching the hot end to the extruder were explored. Shown here are several T-bar approaches, and one which utilises the non-symetric nature of the bamboo to create a type of twist-lock mechanism.
  • Holder, NiChrome wire (a heating block would also work) and thermistor were installed.
  • Attachment to the extruder body utilised the DB adaptor by danieldb2. To help guide the filament another small piece of bamboo was whittled into a funnel.

Test Method 1

With the hot end in a vise and connected to the RAMPS, and using a thermistor connected to the bed temperature pins of the RAMPS board as a probe, measuring the temperature at various points. This is very imprecise as can be seen by the fact the nozzle thermistor read 200°C but the external “probe” read between 185°C and 190°C, but gives an indication of how the hot end reacts. It also is usful when compared with a PEEK based hot end.


Test Method 2

Printing objects as part of a standard reprap setup.


So far the hot-end has seemed to work with no problems. The big question is if this continues over the longer term, and whether it handles bigger prints with longer print times.