My temp calibration process is as follows, though note the changes I want to make at the end. I have a custom designed part, but after going to all the trouble to make it I found someone had already done almost exactly the same part:
I create 9 copies of it in slic3r, and I choose the option to complete each item individually. I then slice it, and hand-edit the gcode:
- Search for Z0.20 (where 0.2mm is your layer height). If you don't have z-lift enabled, this should appear 9 times in your code before it starts each piece
- Right after Z0.20 (important to do this after, otherwise your nozzle hangs in the air while heating up to the next temp step and oozes like crazy), add the code M109 S175
- Search for the next Z0.20, and add in M109 S180 ..and so on
The only tricky thing is, when "completing indivudal objects" I haven't found how to force it to complete them in a particular order. I usually sit and watch the printer and just mark each with a piece of tape as it finishes. It would be nice if it started at the bottom left of the matrix and worked left to right...instead it hops all over the place. To give an idea how it finishes, in the attached photo you can see my test objects with their temperatures written next to them - and in the top right you can see a box with "205 185 180" in the bottom row - those were my notes as they printed.
The problem with this is, and perhaps it's a sign I need to get a gym membership, the parts are too hard to break. The 175 degree one usually breaks easily and cleanly (see photo), but for example my black stuff from 180 to 210 were unbreakable. Even with pliers they are hard to break, so hard that I can't tell the difference.
Based on that I started printing the MM 1.5 parts at 210 (because thats what worked so well with red), and they were so brittle. I think a better design for the parts would be more of an upside down T extruded shape, with the tall part being 3.5 or 4mm thick (like the MM 1.5 parts that printed so brittle). This would give a nice long path for the extruder to travel along simulating a better real print (I think the layer time on the nail shapes might be too short leading to really great adhesion). Also to simulate a real printing environment, I will slow down the printing and up the fan so the previous layer is cold when the next goes on, as I think that has a lot to do with the adhesion.
I quickly threw together a model (attached) - I haven't tried this yet but it is what I have in mind.
