Delta calibration

Hi all,
I've been struggling with this for a while, and not making any progress, so getting really frustrated. Below are a couple of photos of a calibration object I'm printing;



Firstly, I have a "ridge" along the Y axis... you can see how thin the print is at the X extremes in the first image. Secondly, the object *should* be circular with a diameter of 100mm, but it is only about 95mm across along the X axis.The diagonals in line with the X and Y towers are a couple of mm each under size. Along the Y axis, it's right on 100mm. I'm guessing/hoping that these two things are somehow related. I could kindof understand if my print locus was somewhat dome-shaped, but it appears to be more cylindrical.

I've done some measurements and tried to calibrate, but don't know what could be causing this. I do know that all my rods are the same length, and swapping rods from one side to another doesn't affect the shape. I've checked and set my bed to be at right angles to all 3 towers. I know that one side (between the X & Z towers) is a couple of mm shorter than the other two sides, so I have tower positions set at 90, 210, 330 degrees. Previously I had 326 for the Y tower... changing to 330 didn't seem to help). But I can't think how a discrepancy in this tower position could produce an error that's symmetric about the Y axis.

I've tried using the Escher3d.com calibration wizard to at least get the endstops right, but after a couple of cycles where things seem to be heading in the right direction, it gives endstop correction values like 273 and 300 I want to print an effector to mount my dial indicator, to make it easier to do calibration for more factors, but I suspect that if the effector isn't equilateral, I'll just be complicating the problem.

[Edit] After doing some more work on this (turning the effector by 120 degrees (no change), a more detailed calibration object), I found the error is at 30 degrees to the X axis, not exactly along the X axis. i.e. it is aligned with the X tower. The way I figured this was by printing two "circles" with a hole at the centre. I put a drill bit through the centres as an axle, and rotated one circle relative to the other. It was very obvious where the maximum error was. I can now at least correct this error in Slic3r by scaling in the X axis by 105% and rotating by 30 degrees.

Any help will be much appreciated.

Frank

Edited 1 time(s). Last edit at 04/21/2018 07:30PM by frankvdh.

Regarding the Escher3d calculator...

I've just discovered that, for Repetier, it shows the endstop values in steps, not mm. To be fair, this is pointed out in point 2 of the "How to use the calculator" instructions. So my values of 273 and 300 correspond to a much more reasonable 3.4 and 3.75mm.

Now going to read the rest of the instructions...