Friday, 29 January 2016
Worth noting is how the blades are bent in one direction, that must be the way they pull the filament. Which means the way I inserted it must be the correct way, since up here is away from the gear wheel. Which matches well with where the filament goes. Excellent! I look forward to a silky-smooth extrusion tomorrow.
The thing at the bottom is most of what was stuck to the extruder hotend. It's almost but not quite a mold of the hotend.
I peeled off the bits and tried a smaller version. This one fit, but was severely underextruded:
I expect this was due to the previous print having given enough backpressure to cause the filament to strip. I'll check on that tonight. Alas, the new bolt from arcol.hu has not arrived yet, a week after it got shipped.
While I could prevent the situation in the first print by using flat clips or taping around the edge, I do find the fan placement to be a design flaw, the first real one I've found. Having the shroud around the hotend is cool, but the fan doesn't need to be in any particular place, so having it low and in front of the extruder is unnecessary. It makes it likely to hit things and blocks the view of the print in progress. It should be possible to place it vertically off to one side or something. There's a decent amount of space on the right side of the X carriage where it would fit nicely.
Thursday, 28 January 2016
I found a hex wrench (new sets again useful) that's about the right size for the upper hole (which for reference is 3.5mm). So now I heat up the hotend again while it's hanging in mid-air, being very careful to only grip the hotend with well-insulated pliers. The upper part is quite cool, though.
It's dripping black gunk off the side. Ew.
I was able to get the metal wire in from the top and pull a piece through. Only a little PLA came out with it. Then, after letting it cool down a bit more, I realized that was not the best approach. I should leave the wire in while cooling down and pull it at about 90C - essentially a wired Atomic. Didn't do much.
Tried another two regular Atomics. Didn't see much in the way of black stuff, but the filament seems to go in more easily now, and comes out straight.
Finally doing a test cube:
And we're back in business!
So to clean:
1) Remove nozzle from extruder, but leave it wired.
2) Use 0.3 mm metal thread from the top to pull gunk out
3) Clean wider part with a hex key
4) Reattach nozzle and do an atomic or two.
Monday, 25 January 2016
Monday, 18 January 2016
Saturday, 16 January 2016
The fan does make it difficult to see what's going on, and it may end up running into the clamps holding on the glass plate if the print is larger. But oh, my, does it print well now. Did a trio of previous designs, and they are certainly much cleaner. The holes are still a bit small, but with the fan on I probably should redo heat and extrusion amount testing.
But before that, I want to print stuff! Printed the hex bit, the fan mount, a new dog bag wire endstop (that I had to recreate almost from scratch, but got it better), and a dream weaver frame.
One guy at Make Munich today trying to sell my on his printer proudly proclaimed that the firmware and slicer software was all of their own making. To which I had to say, "You say that like it's a good thing". The Marlin firmware and Slic3r are both doing quite well for me.
I've had to move my machine into my own room, since the noises it makes gives my wife a headache. I hope the wobbliness of the temporary table doesn't do too much damage. On the bright side, it forces me to be in my room for other purposes than sleeping, so I'll probably end up cleaning up a bit:)
As a final print tonight, I wanted to do a battery holder to organize my charged/uncharged batteries. But the filament stripped partway through:( Not the first time it has stripped, either. If it happens more, I'll have to (argh!) take the extruder apart and clean the hobbed bolt, I guess.
Monday, 11 January 2016
Increasing by two full rotations on each, and doing an Atomic to get rid of any gunk and undo any stripping. BTW, the video of the Atomic method from Ultimaker shows a pretty horrible design: Temperature adjustment via a dial that has to be dialed many times to turn temperature from 260 to 90.
Looks pretty nice on the sides, corners still upturned a bit, and the top is still not fully connected. Another two turns of each.
Still upturned corners, a bit, even nicer walls, still a bit gap on the top. Two final turns, and it's almost as far turned as the springs allow. The result is really nice, though, much more even on the top.
I want my fan on, but the nicely designed fan piece that guides the airflow beautifully in to the hot end bends the heater and thermistor wires. For now, I'll make a smaller holder that puts the fan higher and at an angle. Less effective for sure, but also less in the way.
So, dive into the scad files for the Mendel90 itself! It's rather complex, with a lot of included files holding functions and definitions - and some of those files depending on other files having been included first. The definition for the fan mount is curiously in the same file as all the other X carriage parts, defined in a single module, inherently translated to an off-center position. That's rather odd. I'm impressed that nophead is able to make parts that good with code that messy.
But I manage, and create a version of the fan holder where the fan side is rotate 45 degrees, so it can be flipped and aim inwards. First print comes out ok, but the holes are too small. I also notice that there's hardly any place for the carriage mount screws to go in.
In the second print, I have fudged the hole diameters and lowered the plate, so there's more room for the screws. It fits onto the carriage screws now, but hits the belt tensioner.
Offsetting the screw slots a bit to factor that in.
I'm making good use of the # operator in OpenSCAD to figure out which part of the code I'm looking at. Too bad I can't just put my cursor somewhere and then it highlights it immediately, that'd be nifty.
It's fun to print parts for the printer itself. Once I have the fan mounted, I can re-print this part in higher quality because it is doing what I wanted it to. Then I can replace it with the higher-quality version. It's like building compilers, just with physical things.
It still hits the tensioner a bit. I can add 3 mm more before hitting the other side, but I'll wait for the working model before doing that.
Mounted it now, rather fiddly mount really, and the carriage mount came off a bit while mounting the fan, so I had to re-tighten that through the fan blades. But it sits there, just a bit above the bed, and works! Here's a test cube done with it:
There's corner lift on the far side where the fan doesn't reach as well, but the layers on the front and top are absolutely beautiful, and the top is a lot better, too.
Reprinting the fan holder, though not moving the holes since during the mounting proces I found there was just the right amount of space on either side of the nuts.
I printed the bit I designed the other day. One side is perfect, flat hex sides, nicely rounded top, the other is weirdly rounded and the top leaning towards the good side. Proof enough that I need to do the proper all-round fan.
Changing the design of the fan duct is hard. Easier to change the fan holder part a little.
The first print of the elongated fan holder has one side of the vertical part being really crappy and the other being fine. I suspect it's a retraction issue - the bad side is the one being printed right after the middle, so there's a much smaller retraction time.
Maybe if I increase the retraction? Trying double retraction, didn't make a noticable difference. The holes look funny here because I sanded them a bit to get the screws in.
Ended up doing two extra prints because I didn't realize I was saving into the wrong directory. Since I had the modified holder ready, I took off the kludged fan, and only then realized I had used the wrong size settings, dibond instead of sturdy. So I printed again with correct settings, but no fan - and it came out nicer! Not perfect, but not as bad either. Here it is, mounted on the fan.
Unfortunately, while trying out the new fan, the two screws mounting it to the X carriage came out, and in turn the nyloc nuts they go into fell out, into the underside of the carriage. That's going to be a pain to get back in - either I do some really clever fudging, or I take the entire X carriage off. Might have been a better design to drop those nuts in from above.
That's probably a sign I should stop here for the weekend.
Saturday, 9 January 2016
But first, I realized from looking at The Calculator that the Z steps/mm was a little off - it should be 2560 but was 2600. Not a huge difference, but might throw the layer height off a smidgeon. So I fixed that, which in turn made the Z offset wrong. So I have to re-calibrate that. 19x10mm first, then to reach the 10.43mm hex bit took another 4x1mm plus 5x0.1, so total 194.4 + 10.43 = 204.8, 3 millimeter higher than before (then - 0.2 to allow for hotend expansion, according to nophead). No wonder that test print curled up on itself.
There was a rumbling sound when homing X+Y, I should probably recheck the fastening of the gears if I see any sign of X or Y slippage.
The result is good, seemed to have slight initial sticking problems, still has a layer or two with gappiness and the top is still not fully closed. I will try slowly increasing amount of extruded material. (Cube #18)
Reducing my fudge factor to 50.0 / 30.0 and increased height to 204.7. Decent print, still with some adhesion problems, but the top looks a little better. (Cube #19)
Reducing to 50/27 and increasing to 204.8. (Cube #20). Adhesion pretty good now, but worth adjusting once more. Also less gap at the top infill.
Trying one more adjust: 50/25, 204.9. Now it's starting to get rising corners and thick layers. Somehow I manage to get overextrusion and underextrusion at the same time:(
Thursday, 7 January 2016
After this, a calibration cube printed almost as well as before - a few layers are gappy. Would another Atomic help? Or is it a question of adjusting extruder tension?
|Cube #15: Slight wall stripping|
One of the troubleshooting pages mentioned hole shrinkage, which I think I've been seeing. So I went to the original page on the subject and found a nice piece of OpenSCAD to do a test print (except it was using the outdated assign() operator, easily fixable). But even though it's but 3mm of print, it's slated to take one hour and 45 minutes! That's a long time. Maybe this would be the right time to test what happens if I speed up printing. Let's try a cube at 200%:
|Cube #16: Wibbly-wobbly, half-timey-wimey|
|Cube #17: No help for the top layer|
Wednesday, 6 January 2016
Should be easy: A plate that can be sewn on, a cone with an appropriate hole. Starting with the plate, which is just a hull of four circles, one with a larger diameter (here shown after a bit of getting the syntax wrong and without the appropriate offsets to make it square):
Why won't this compile? This is perfectly cromulent code, but it balks at the assignment to radius, even though variables have been assignable in any scope since the 2015.03 release. Oh. I have the 2014.01.29 release. How altmodisch! Well, I'll just compile it, then (this being a work laptop, I'm not too happy about installing binaries from random repositories). How hard can it be?
Hard, it appears. OpenSCAD was not made with home-compilation in mind. It doesn't use autoconf like in the good old days, but rather has a home-made script that downloads the sources from github (unnecessarily, since I have the sources here and want the released version), then lists some libraries that must be installed, and then has a script that checks if they are of appropriate versions. That's ... not what I'm used to. The script that checks dependencies finds nothing except make, and has an error about "-query not found"!
- [Qt4 (4.4 - 5.4)](http://www.qt.nokia.com/)
- Available as libqt4-dev, version 4.8.5
- [QScintilla2 (2.7 - 2.8)](http://www.riverbankcomputing.co.uk/software/qscintilla/)
- Available as libqscintilla2-dev, version 2.8.1
- [CGAL (3.6 - 4.5)](http://www.cgal.org/)
- Available as libcgal-dev, version 4.2
- [GMP (5.x)](http://www.gmplib.org/)
- Available as libgmp-dev, version 5.1.3
- [MPFR (3.x)](http://www.mpfr.org/)
- Available as libmpfr-dev, version 3.1.2
- [cmake (2.8 - 3.0, required by CGAL and the test framework)](http://www.cmake.org/)
- Available as cmake, version 2.8.12
- [boost (1.35 - 1.57)](http://www.boost.org/)
- Available as libboost-dev, version 1.54
- [OpenCSG (1.3.2 ->)](http://www.opencsg.org/)
- Available as libopencsg-dev, version 1.3.2
- [GLEW (1.5.4 ->)](http://glew.sourceforge.net/)
- Available as libglew-dev, version 1.10
- [Eigen (3.x)](http://eigen.tuxfamily.org/)
- Available as libeigen3-dev, version 3.2.0
- [glib2 (2.x)](https://developer.gnome.org/glib/)
- Available as libglib2.0-dev, version 2.40.2
- [fontconfig (2.10 -> )](http://fontconfig.org/)
- Available as libfontconfig1-dev, version 2.11.0
- [freetype2 (2.4 -> )](http://freetype.org/)
- Available as libfreetype6-dev, version 2.5.2
- [harfbuzz (0.9.19 -> )](http://harfbuzz.org/)
- Available as libharfbuzz-dev, version 0.9.27
- [Bison (2.4 -> )](http://www.gnu.org/software/bison/)
- Available as libbison-dev, version 3.0.2
- [Flex (2.5.35 -> )](http://flex.sourceforge.net/)
- Available as flex, version 2.5.39
- [pkg-config (0.26 -> )](http://www.freedesktop.org/wiki/Software/pkg-config/)
- Available as pkg-config, version 0.26
|Two versions of the bit. Fanless on the right, with fan on the left. Notice the left having a nice flat hex part at the top and the cylinder is fairly straight, while the right one is more rounded and the cylinder is wobbly.|
Update: Pulled out the filament. It doesn't look badly stripped, but there are marks on it. Tried extruding with more or less tension, with no obvious difference, speed is about 1/2 of before. There's a slight clicking every few seconds, but it can't account for this amount of difference.
Friday, 1 January 2016
As I am AFP (away from printer) for a few days, no new prints for a bit. But I want to point out a web site I just ran across.
Walter at thrinter.com has a lovely overview of many types and makes of filament, including his own (extensive) experience with many of them. It's a valuable resource for anyone curious about which filament to get for their printer.
It's also an interesting source of inspiration for what and how to design, and a good place to look if you're not sure what a 3D printer is good for. In the PLA and ABS sections in particular, he shows plenty of example prints, from the artistic to the immensely practical, often with design tricks I hadn't considered and that you rarely see in commercial products. Well worth a look.
Z offset Tuesday morning: 1.0. Air pressure 1017hPa. Humidity 26% Z offset Tuesday evening: 1.1. Air pressure 1016 hPa. Humidity 30%. Mov...
Since my printer is currently in a good shape and simply printing useful things (and some less useful), I've thrown myself at another pr...
As I've been printing this and that, my Z offset has gradually increased to over 2mm. I have two remaining hypotheses for what causes th...