Setting end-stop screws for level Z

[TN Yankee]

Greetings, all. I've been lurking here for over a year (bought my Rostock V2 in 11/2014), and have enjoyed and appreciated all the advice and wisdom I've been able to pick up.

I used my printer extensively for the first few months, then it got packed away last spring for an upcoming house move, and I'm just now getting it set up again. I upgraded to the current ball joint carriages and arms. Everything is functioning, but I'm trying to get the end-stop calibration (i.e. flat Z) dialed in better. I purchased the usual Pittsburgh/Cen-tech (Harbor Freight) digital depth indicator, and drew up a mount for it, based on the current ball joint effector plate STL. I run a modified "all towers cal" from the current build instructions, adjusting Z heights and speed to accommodate the gauge.

All in all, with learning the best Z height range for the gauge, a few turns on the wrong tower, etc., I recorded 42 iterations. The gauge reads to 0.01mm, claims accuracy to 0.001" (0.0254mm), and so I consider +/- 0.01mm to be as good as I'm going to get. The gauge does seem to be very repeatable, which for this test is much more important than accuracy. Each time I got close to equal readings (i.e. XYZ all within 0.01mm of each other), I did 2 more iterations without making adjustments to check for consistency - they were always very stable. Then I tweaked EEPROM horizontal radius, and started over again if necessary.

After dialing everything in better than I thought I would be able to, I decided to rotate the effector plate to another set of arms to verify. Bad news: stable, repeatable values, but not level (X now 0.3mm high, Y 0.4 low, Z 0.2 low, but center still spot on); rotating to the other arms gave me different but still very uneven values. I thought "OK, the plate/ball clips/gauge aren't perfectly level, I can live with that." I then moved it back to the original mounting arms and it came out uneven again, though not as bad as with the other mounting locations. Now I'm wondering how can I trust any of these measurements? Due to having to swap ball clips from the dial plate back to the hot-end plate, I haven't yet run a thin print to visually check flatness.

How do you guys run your dial/digital indicators? Do you verify by mounting in different arms?

Thanks,

Mike

[TN Yankee]

Moved to http://forum.seemecnc.com/viewtopic.php?f=77&t=9810