what causes x/y skew?

I finally have a flat bed and perfect scale on the X and Y axes.
but when I measure the diagonals they are off about 2mm on 100mm square. That would be about 1.4 mm of lean to one side.
thanks

If the towers are not equidistant from each other, or the diagonal rods are slightly different lengths, that will cause skew. Depending on what firmware you use, you may be able to either tell the firmware exactly where the towers are, or enable axis compensation.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

^ what he said. This is by no means a very scientific way of doing things, but I like to use an 'asterisk' shape for calibration. The idea is that you have an asterisk, whose arms run parallel to each of the towers. This is to isolate the movement of each tower (or more specifically, carriage) and will allow you to see which one needs correcting. Assuming your rods are all of equal length, the way to correct for tower displacement is by adjusting delta radius. If you just think of each tower as forming a right sided triangle (diagonal rod, delta radius and carriage z position [the uknown]), you can quickly see the effect changing delta radius for one tower will make. As a general rule, if you increase delta radius, it will lengthen the 'line' produced parallel to a tower.

thanks for the feedback,all of you.
I know of the extra parameters to adjust (RichCattell firmware) although I'm not sure how to manually calculate them.
BTW, I removed the nozzle and stuck an ink-pen into the heat-break, held in place by the filament pushing on the opposite end. Makes a great quicky plotter. I just manually jog the tip around with Pronterface on a piece of paper, takes me seconds to measure, adjust a parameter and redraw. The slowest part is re-homing each time!)
Anyway, just to be clear, here is top-down image of the bed:
The lengths of the edges are perfect in length and scale (100.03mm!). It's not obvious from this image though how any individual tower-radius would cause this. I just build up new rods and the maximum difference of the six rods is 0.17mm.

Quote
shadowphile
The lengths of the edges are perfect in length and scale (100.03mm!). It's not obvious from this image though how any individual tower-radius would cause this. I just build up new rods and the maximum difference of the six rods is 0.17mm.

1. Have you measured the distances between the towers, to see how close they are?

2. Have you checked that the towers are perpendicular to the bed? A sloping bed or tilted towers will cause the delta endstop adjustments to be at different heights, which will cause XY skew (and of course there will be Z skew as well).

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

This is the calibration asterisk I use. You'll likely find that the arms are uneven lengths and widths - printing (or plotting) this should give you a few more clues regarding the nature of the skew.

Edited 1 time(s). Last edit at 02/11/2015 08:29PM by iira.

I measured the outside width of the three towers, down around the base. They differ within a .5mm range.
I also plotted 3 lines pointing from the center toward each tower and they differ within a mm range!
i am going to try to physically straighten them out but in the meantime can someone tell me how to interpret and use the individual tower adjustments.(radius and position).
thanks

Hi,

I have the same issue with my new Kossel Mini. The layers of my prints are farther away from the Z axis for each new layer. Did you manage to fix this ?
Thanks