#TL;DR
- given that the "ideal" numbers are +255, +255, +255
- adjust screw A til #1 is close to +255
- adjust screw B til #2 & #3 are close to equal
- this may require adjusting screw C if #2 or #3 start +ERR-ing
- adjust screw C til #2 & #3 are close to #1
I bought it used from my friend. Before I paid he wanted to make sure I could calibrate it and do some test prints so I could see that it worked.
+ERR, +ERR, +ERR
I observed that the extruder wasn't touching the bed when it was trying to calibrate point 1 -- at this point I can't tell if that is what was actually happening. It also never left the first calibration point and threw the +ERR before moving on to the other two calibration points. So we loosened the screws as much as possible, though not so much that the springs lost tension.
I also cleaned the crap out of the metal calibration points (using an eyeglass cleaner cloth and applying a lot of pressure with my thumbnail but also keeping my other fingers underneath the heat bed so I didn't apply too much downward force) on the corners of the heat bed, and scrubbed the shit out of the extruder with the copper-bristled brush.
That still returned all +ERRs but at least the printer was finally moving on to the other 2 callibration points and not stuck on the first one, so we tightened each screw by the same amount because the bed seemed like it was generally too high as it was callibrating (it was hitting the pulleys during the callibration process).
+445, +ERR, +ERR
I don't remember what we actually did because we were in full experimentation mode, but here is the sequence of values we got:
+ERR, +380, +ERR
+ERR, +382, +305
+ERR, +286, +ERR
+439, +306, +ERR
+415, +298, +ERR
+375, +289, +ERR
+371, +299, +284
We were so stoked to finally get all 3 non'ERR' values that I didn't realize we weren't nearly done. This is what the full screen said:
CALIBRATE
FAIL
+371, +299, +284
[OK] TO RETURN
FAIL
From that point on, we never saw +ERR again. But we did calibrate it about 13 more times with various ideas and hand-wavy concepts that we weren't familiar with.
Turns out we both slept on it hard. These were our results:
My friend is a wizard, so he whipped up some formulas:
Given that 50c = r (50 calibration points equals one rotation), assigning the screws to A, B and C and calibration point readings to X, Y and Z respectively; I believe we may be able to calculate the correct rotations for the screws with these three equations: Xc = Ar, Yc = Cr+Ar-(Ar+Br)/2 and Zc = Cr+Br-(Ar+Br)/2
I came up with these steps:
- given that the "ideal" numbers are +255, +255, +255
- adjust screw A til #1 is close to +255
- adjust screw B til #2 & #3 are close to equal
- this may require adjusting screw C if #2 or #3 start +ERR-ing
- adjust screw C til #2 & #3 are close to #1
The final three values from the previous night were:
+289, +235, +315
I wanted to try out my new steps so I tightened screw A a bit, guestimating how much it would take to get #1 to +255
+274, +227, +325
I tightened screw A some more (maybe 1/3 turn), tightened screw B too but not as much, tightened screw C too even less.
+266, +222, +313
I tightened screw A 1/3 turn, and screw B 2/3 turn.
+266, +243, +312
I tightened screw A 1/3 turn.
+244, +235, +319
Nice! Now untighten A 1/6 turn.
+252, +236, +312
Cool! Now tighten B 2/3 turn to try to get +236 & +312 to closer values as each other.
+251, +253, +300
Closer! Now tighten B another 2/3 turn.
+250, +271, +288
Even Closer. Now tighten B 1/3 turn.
+256, +286, +290
A mere difference of 4. So it was time to tighten C! By a smidge more than 2/3 turn.
+260, +249, +225
Hmm, my theory is starting to waver... I will loosen C by 1/3 turn because I don't know what else to do other than backtrack. Perhaps C doesn't need to be turned as much to get as far because of its positioning.
+255, +256, +250
SUCCESS!!!!!
Full screen readout:
CALIBRATE
SUCCESS 250
+255, +256, +250
[OK] TO RETURN
SUCCESS🎈 🎉 🎈
🎉 💃 🎉
💸 🐷 👍