Wednesday, 2 November 2016

Converting 3MF Files to STL

When I create new items in Google Sketchup, they sometimes have holes in them when I export them as STereoLitography or STL files, as such I usually make use of the Microsoft repair service on Azure.

This allows you to upload an STL file, the site will repair the holes and then allow you to download a repaired STL file of your model.

Unfortunately, Microsoft being Microsoft, along with a bunch of other big companies decided they wanted to use a file format that had Digital Rights Management DRM encoded into them and so the 3MF file format was born.

At some time in the recent past, Microsoft updated their repair service so it now only exports 3MF files - great - let's stuff the free community!

Repetier Host, Slic3r, etc do not support the 3MF format, although Cura does, but only because the people that write it - Ultimaker are part of the 3MF Consortium.

Windows 8 and 10 users can use the Microsoft App 3D Builder to convert their 3MF files back to STL format, however I don't have any Windows machines and only use a Mac, so I need to look a bit further afield.

So I go looking for a potential solution, and find one on released on GitHub by Charles Shapiro called 3mf2stl.

This is intended as a Linux solution to the problem, but I see no reason why it should not also work on a Mac as this is also Unix based.

I download the source code, extract it and type 'make' in the top level directory (OK, so it only has a top level).

First stumbling block is that it requires libzip - no problem, that is available via the App store and is a pretty simple command line install.

Second stumbling block, libzip won't actually install because my version of Xcode is out of date and I need to update to 8.1.

Third stumbling block, after waiting the best part of an hour for Xcode to upgrade itself, libzip will now install, however the make file is still not finding zip.h from /usr/local/include.

This requires re-installing the Xcode command line utilities by running 'xcode-select --install' at the command line.

Fourth stumbling block, now make can find zip.h, however it now barfs about zipconf.h, which is not actually held in /usr/local/include.

This is fixed by creating a link to its real location - 'cd /usr/local/include; ln -s ../Cellar/libzip/1.1.2/lib/libzip/include/zipconf.h'

This time make runs through cleanly and produces an executable - 3mf2stl, which I then copy to /usr/local/bin.

Now I can convert my 3MF files to STL by running: 3mf2stl -i <input file> -o <output file>

Halloween 2016

Following on from last year, I wanted to print a few more spooky items.

I bought a 1kg reel of 1.75mm glow-in-the-dark filament from ebay seller etang_uk for £11.99 with free shipping.
I then used this to print a couple of Cute hug me Ghosts from Thingiverse:
First I printed one at 100%, then I printed a second one at 155%, so it could wear the Microsoft Witches Hat from Thingiverse, that I printed in black.
Here are my ghosts:
I then decided to print a couple of the Skeletonz from Thingiverse:
Unfortunately, my heated bed was not as level as it might have been and the one on the right ended up with fused leg joints.

These weren't the easiest of items to remove from the bed either - I ended up using a plastic ruler to swipe sideways against them to detach them form the bed.

Friday, 28 October 2016

TCT Show 2016

One of the main items I was wanting to see this year was the new HP printer that uses inkjet technology to print, unfortunately, it seemed that the machine was in pieces and also that I actually knew more about the machine than the HP representatives on the stand!
 Here is an interesting Prusa I3 on Josef Prusa's Prusa3D stand with 4 extruders.
 This delta printer had an interesting carriage design
 and also used silicone tubing to suspend the extruder motor above the extruder.
The big brother of this printer was huge - you can actually see peoples feet next to it to give you an idea of the size of this thing, no it is not a perspective thing, this printer is actually standing on the floor and it was printing directly from pellets:
The Zortrax superhero and his new bike (in the background), both printed on Zortrax printers.
A prototype STL printer on the RepRapWorld stand
A very interesting printer made from a pair of Ikea tables on the Printtable stand (unfortunately I did not win this printer in their raffle)

Thursday, 29 October 2015

Halloween 2015

I saw this in our local Home Sense and thought it looked like a great project for a laser cutter or 3D printer:
First off I drew it up in Sketchup
I made a plate that could be cut on my laser cutter (when it is finished)
but ultimately scaled it down and printed the pieces individually.

Here is the finished article, with an orange LED tea light inside it, which I believe was from a pack of 8 bought at Wilko for £2.00.
I have also uploaded the Sketchup source as well as the STL files to Thingiverse.

I also decided to print Makies Jack-O-Lantern from Thingiverse.
Here is my one with an LED tea light inside it:

Saturday, 3 October 2015

TCT Show 2015

The TCT show was held at the NEC in Birmingham over 30th September and 1st October 2015, I attended on the last of the two days, arriving at about midday and expecting to be on my way home by about 14:00 - how wrong I was!

I new in advance that most of the exhibitors were going to be big business looking to sell their products or services, but was surprised at how many smaller firms, individuals and other new start-up people were there, along with elements of the reprap community.

I was also surprised at how many of the larger company representatives were more than happy to talk to someone who was highly unlikely to be in the market for one of their machines.

As expected Stratasys had one of the largest displays and many of the other print bureaus were also using their machines to promote their own business.

Autodesk were promoting their Ember resin based 3D printer - a snip at $5995.00 (£3870.00).
This is most likely competing with the likes of Formlabs and their new Form 2 printer which currently retails for €3299.00 (£3000.00).
Another highly impressive piece of equipment on display was Renishaw's Equator, their high speed delta comparative gauge for parts inspection.
I am always interested in what Renishaw are doing as my brother-in-law works for them, and I went to school with the son of the owner of the company.

I had a long discussion with the guys from ReprapWorld about the sliders I am using on my Kossel Mini printer, this is an area that they would like to explore further for use on one of their larger designs.

I would like to thank them once again for the spools of copper and bronze PLA they gave me at the end of the show as part of their raffle.

Another area that caught my eye was the kideville product from the start-up company kidesign, this is a collaborative city design project for schools, that incorporates design, CAD, 3D printing and teamwork, each student has a design spec for one element of the city that is eventually printed and then included in the city on the grid board.
Towards the end of the show I spotted the Roland iModela, which for £499.00 ($750.00) can mill, engrave and 3D print in it's 86x25x26mm build envelope and pretty much every part is available as a spare in their parts store.

I would also like to thank Sam Deng of 3D Elements for kindly providing me with a half a dozen 50g (about 12m) samples of their PLA filament to try out after some very interesting discussions regarding their various filaments, some of which reacted to light and others to temperature.

Tuesday, 16 June 2015

Mini Kossel Build Part 4 - Spool Holder and End Stops

My new E3D hot end clone turned up the other day for my Mini Kossel, this cost £7.49 ($11.60) with free P&P from ebay seller digital_store_2014.
I will replace the thermistor with a ATC Semitec 104GT-2 so that it is a known quantity, rather than the complete unknown that came with the hot end.

Based on the issues I had in the early days with my Prusa I3 printer and having to manually turn a spool of filament, I decided to get ahead of the game this time and print a spool holder.

I particularly like the look of this minimalist spool mount and guide on Thingiverse

As there are really only 2 different items to print, I need 3 of one of them and they are all pretty small, I decided to print them all at once.
In total this used 5767mm of filament for a cost of £0.216.

To make them up required 6 x 625 bearings, 5 x M5x20mm screws, 3 x M5x10mm screws, 1 x M5x30mm screw, 6 x M5 nuts, 3 x M5 T Nuts and 11 x M5 washers.
The 625 bearings were £1.61 ($2.50) for 10 from ebay seller warm_tech, the screws, nuts and washers were all from Orbital fasteners, and the T-nuts were $7.50 (£4.85) for 100 from robotdigg.

This makes the total cost (6 x 0.161) + (5 x 0.0484) + (3 x 0.0394) + 0.0697 + (6 x 0.0041) + (3 x 0.0485) + (11 x 0.0041) = £1.61 ($2.50).

These were then installed on the top of the frame and adjusted to fit one of my spools of filament.
I then relocated the extruder stepper motor and fitted the hot end and a length of 4mm OD, 2mm ID PTFE tubing between the push fit connector on the extruder and the one on the hot end.
I am still deciding on how thick an acrylic printing surface I want to use, I will either settle on 10mm, or 5mm and a 2mm aluminium sheet underneath it for additional rigidity.

5mm will be cheaper at half the price, but also half the rigidity.

I have designed some end stops for use with my hall effect sensors and magnets in Sketchup.
For some reason Slicer did not like the countersinks I had put in the holes, it was convinced that there was a hole in the solid, so I decided to print them without and manually countersink them afterwards.

The slot for the hall effect sensor curves upwards towards the rear to help the pins exit the hole.

Here are the end stops hot off the printer.

The 3 end stops used 1572mm of filament for a total cost of £0.06.

Here  is one with the hall effect sensor installed and held in place with some blue tac.
I have pre-insulated the pins with some electrical tape to prevent unwanted contact of the neighbouring pins.

View from underneath.
Here are all three after I removed the sensors and then manually countersunk for the M4 screws with an 8.5mm drill and then re-assembled.
I actually turned the drill by hand rather than under power on my mini mill to give me more control of how much material was being removed.

I have two sets of very sharp, precision ground HSS metric drills, from Chronos in 0.1mm increments covering 1-6mm and 6-10mm respectively. I use these whenever I require a very sharp or accurate drill, most of the rest of the time I use a more generic set of drills.

The hall effect sensors were $4.00 (£2.58) for 10 from ebay seller gc_supermarket, the M4 T-nuts were $7.50 (£4.85) for 100 from robotdigg, and the countersunk M4 screws actually came with hinges I bought for my laser cutter where I used some slightly longer screws instead, so these were effectively free.

The total cost for the completed endstops is therefore 0.06 + (3 x 0.258) + (3 x 0.0485) = £0.882 ($1.35).

Here are some coils of wire I made up for the endstops, they are each about a metre long and are made from 3 cores of a 50m reel of 6 core alarm cable I bought from Wickes for £12.49 ($19.35) back when I was building my Prusa I3, so about £0.25 each.
Mixing up the various cores allows each cable to either be the same or different depending on what you require at the time.

I tend to leave the outer protective sleeve on for the 4 core motor wires, but twist any signal wires to help avoid any cross talk interference, should the signal wires need to run together with the motor wires.

Saturday, 2 May 2015

Mini Kossel Build part 3 - The Spider

I decided to simply bite the bullet and make all of the carbon tubes the same length, so I set up an end stop on my milling machine and held the shortest and one other tube in the milling vice, whilst I cut the longer one down with my Dremel.

Next I made myself a quick jig to ensure that all the rods were exactly the same length once assembled.

This was simply a piece of left over laminate flooring with two 3mm holes drilled through it 228mm apart (it would have been 230mm if I had not needed to cut the tubes down) and then some long M3 screws secured through the holes to act as guides for the rod ends.
I then tried several different glues to attach the rod ends, there must be something wrong with my cheap Wilko brand super glue, because the only thing it seems capable of sticking these days is skin!

It had no interest in securing the grub screws to the carbon fibre tubes, in fact it was not even interested in drying out!

I eventually decided to use hot glue, this worked fine, although there was very little working time before it set.

Here are three of the completed arms mounted on the jig.
Next, I attached one end of each of the arms to the effector to make a 'spider'.
Then finally attached the spider to the sliders on the frame.
The level of lateral movement on the sliders has now vastly reduced.

At most I can move the effector 0.2mm in a circle without actually moving the sliders up or down, but since nothing will be attempting to push the effector out of position, it remains to be seen if this causes issues when printing.

I have been designing some mounts for the hall effect sensors in Google Sketchup, although I am not 100% convinced I like them yet.
Here it is shown on the model in relation to the slider.
The wires from the hall effect sensor would come out in front of the screw and then be hidden within the 20x20 profile slot.

I would prefer the wires to come out at the back, but then I have no way of routing them past the T-nut, or past the T shaped base of the slider, that would be blocking the path of the wire.

Instead I will route the wire to the top of the frame and then back down the side of the 20x20 profile.

I likewise can't use the centre of the profile  for routing wires, as this has a foot screwed into it at the bottom, blocking the egress of the wires.

I may yet also have issues routing the wires past the tensioning screws at the top of the frame.