dimensional accuracy off in only one direction?

[cwilson]

Hi all,

I'm a new Rostock Max v2 owner, and am having a lot of fun with it. I noticed while doing the test outlined in this video: https://www.youtube.com/watch?v=FSOPsRiiOZk to find the optimal print temperature for PLA, that I have a dimensional inaccuracy between the X & Y towers, but not from the Z tower (see the diagram below). The cuboid is supposed to be 25mm square. I would have expected any error to be the same on both sides. Although .4mm doesn't seem that bad (or is it? what do you all experience?), what are the possible causes and potential remedies of this?

[img]http://www.footprintgraphics.com/person ... iagram.png[/img]

Thanks,

cwilson

[Generic Default]

I had this same problem when I had inconsistent arm lengths from me failing to cut them to the right length. I don't know why you would have this if you're using the injection molded SeeMeCNC arms, since they should all be the same length. 0.4mm won't affect the looks of your print too much, but if you need cylindrical parts that rotate, 0.4mm is too much.

It might be that one tower is leveled wrong and the plastic is be squished more on that side, causing it to be larger than it should in that dimension.

Try a triangular print with the triangle edges lined up with the towers, then measure that.

[cwilson]

Thanks for the reply, Generic Default. I should have mentioned that I have upgraded to Tricklaser's carbon fiber arms. I did check their lengths with my digital calipers before installing, and they all seemed consistent. I squared the towers to the glass bed with a machinist square, and all seemed pretty good (though admittedly the aluminum towers and melamine base/top do allow for some flex with even moderate hand pressure). Could it be something as simple as slight variations in blue painter's tape thickness on the bed?

I'll give the triangular print a try, too.

-cwilson