…Continued from previous post…
I swear this is getting worse 
I ran 100mm through, got 101.3mm and printed the calibration cube. Measured x,y,z. Calc’d new values and uploaded them. Reprinted the cube and this is what I got.
Thanks for the link Steven but I am using Sprinter firmware.
It looks like one of your problems is your zsteps/mm. You see that the lower portion of your print is more orderly than the upper half. This is likely because you are going a little bit too far on each layer, until the nozzle is just dropping the plastic on top of the last layer, rather than sticking it directly into place.
@Michael_Hohensee but the cube is measuring the right height.
I rechecked the filament diameter which was a bit off. Increased that number in Slic3r and, I’d say, it has made about a 20% improvement. Going in the right direction, another clue?
It looks like there is something else going on beyond the software calibration issues.
Even if you are over extruding, I would have expected the layers to still bond with good alignment and consistency; which the first few layers already show.
You have not mentioned what speed you are printing and this could have a bearing with the issues you are having. It will be a good thing to rule out mechanical wobble before attempting calibration because an unstable printer frame will not give you consistent calibration measurements.
It seems unlikely for the filament to drop over the side if the nozzle was perfectly aligned over the last layer perimeter routes.
Chin up, more square than I got on my MM2 in the first month.
@Richard_Blackman Z steps/mm needs to be calculated, not measured. Are you sure your layer height is actually small enough? Go down a bit further than the 80% of nozzle diameter max you’re using, see what happens?
This really looks like really bad backlash. Loose belts, T5 belts, messed up pulleys, crappy bearings… Have you ever said what your machine is?
@Jasper_Janssen The measurement he’s talking about is in response to my comment on the previous thread recommending a method to check whether or not his E steps were correct.
@Richard_Blackman Has your hot end come loose? Like others have said, it could be backlash in your belts, or printing at too high of a speed, or possibly a slightly loose hot end. It may be ok at first, or at random parts of the print, as the head wedges itself into a stable position but then suddenly the head starts shifting again and you get odd misalignment layers and squishing effects.
It still looks suspiciously like overextrusion to me, though.
@Stephen_Baird I think you’re very close with the overextrusion idea.
I have checked the belts, pulleys etc. The only signs of wear are on the bearings but note that the problem is seen on both X&Y sides. I suppose they could have worn all the same way to the same degree but…
Anyway, I’ve tried printing at .1, then at .3 layer height with same results… then I thought I’d try the extrusion multiplier setting again… see next thread (geez, I wish I could add more pics to same thread)…
@Richard_Blackman never, ever, never, never, never, “calibrate” your steps/mm for any axis other than E! @Josef_Prusa has a nice calculator online that gives you your true steps/mm for XYZ. Then use @Richard_Horne 's guide to set filament diameter and steps/mm for E. If you’re still seeing the issue you’re posting about, you might need to adjust the extrusion multiplier in Slic3r (Skeinforge calls it flow rate).
That picture looks exactly like some of my first prints three years ago - it took me some trail and error to get where i am now.
@Richard_Blackman
It looks like you are at least 30% over extruding to me. bring the e-step figure right down 40% and you should be under extruding, then see if it’s also a mechanical issue.
Calibrate E-steps later, just rule out mechanical issues first.
Is this being printed as a 100% solid calibration cube?
Also print is slowly (25mm/sec) when you are trying to get things calibrated. Make sure you have cooling enabled and a minimum layer time of around 15 to 20 seconds.
Check the nozzle diameter is correct in Slic3r and confirm with hot-end supplier.
Let Slic3r calculate width and start with 0.3mm or 0.25mm
And make sure Slic3r is saving your changes, sometimes it can screw-up so if in doubt create a new profile.
Then also try printing just a hollow cube, that should tell you straight away if you are seeing mechanical issues, if layers are not lining up, sort out belts frame and any loose things first.