And they saw that it was good,and they named it… Run CNC
Yes, that night has finally arrived – we have started printing whole objects. After three long nights of details and more details, re-working small finesses and taking well thought through short cuts (well, we think they were very well planned!), the steel machine is beginning to make sense. And to make things.
As I write, CH and CM are watching the twitchy printing of a decorative butterfly. Like two fathers watching their newborn, there is even a special warming radiator attached to the frame for extra cuddly warmth.
So where were we last week? We had most of the the mechanical bits in place, we had chopped down, milled and machined, drilled and threaded two dozen linear bearings saved from the scrap pile at some engineering shop, ground out a few planning errors in the toothed belt positioning, tested motors, reassembled the extruder too many times in various configurations, built a debugging display board and kept ourselves far too busy with details and possible extra details.
This week things got serious. The rods were replaced with ones that were hardened, so the bearings would not dig themselves in. The extruder was attached to the extruder sled. The belts were attached to the X and Y axis sleds (i.e. extruder and build platform) and tensioned. The extruder was reassembled. Then we had to attached beltsto the sled, so off with the sledbearings, hammer out the bearing from the block, cut a thread into the block and reinsert the bearing using the drill press and a pressing tool.
Don’t try this at home.
Tonight we learnt what PID stands for (Proportional, Integral and Derivative, somehow like Present, Past and Future – see wikipedia for details), configured the beta and other temperature coefficients, wrapped our build platform in Kapton (Actually the cheaper Koptan tape from Deal Extreme) and started extruding.
The first extrusions were all a mess. We started with the standard – a 20mm cube. On a raft. I have yet to work out exactly what the raft is for. Except that the mungy mistakes happen at the edge of the raft and are thus kept away from the rest of the print.
We used a heat gun and then a bar radiator to heat the platform from above. The radiator works better because the air movement of the heat gun seems to cool down the extruder. The Koptan surface of the printing surface was roughed up with some 280 grit sandpaper, then dusted down afterwards, for better grip. We ended up setting up the nozzle to be exactly 20mm above the print surface, then using a G92 X0 Y0 Z20 G-code command to let the machine know it was X-Y centered and 20mm above the surface. Thus worked better than trying to have the nozzle touching the surface at the start – we ripped two holes in the Koptan surface doing that.
It is not an amazing thing, but tonight we finally managed to keep the raft sticking and the ABS flowing and ended up with a cube. HUZZAH!
So what comes next? A lot of things.
When we measure the temperature right at the tip of the nozzle, it is often over 230 degrees – and the ABS seems to be too hot, it is coming out not as shiny as it should be. So we have to work out what to do here. At the moment we have set our operating temperature to be 205 degrees, because the thermistor measures that (correctly!) on the outside of the tip even though it seems to be warmer inside.
In the longer run we want to develop a “bowdenzug” style extruder, getting as much weight and volume off the X axis sled as possible. Then we can move faster, possibly more accurately, and have more space internally. Since we want to print our own parts for this, we want to get the printing process in a very stable state. The cube is solid, but we need to work out how to play with operating mechanical parts that should fit together, align with axles and work effectively.
So the main next steps are more building! Trying out techniques, building simple and not so simple things. Let’s see what happens here.