[Solved] Calibration not working

[jeffhuber]

I wanted to do a routine calibration to iron out some of the kinks in my models. So i fire up the computer and pull up the trusty assembly guide and scroll down to the calibration portion. Let's do this.

So I run scripts 1, 2, 3, 4 just like it's listed in the calibration manual. I've been adjusting the horizontal radius in increments of .2. I'm trying to get it as close to perfect as possible. I get to a state where I'm pretty happy with how everything is working. I run script 1, 2, 3, 4 and the paper seems to have about the same clearance under each one. NICE!

but wait...

Then I run an alignment script I had seen someone on here post:

Code: Select all

G28
G1 Z0.2 F15000
G4 S1
G1 X-77.94 Y-45 F2000
G4 S2
G1 X0 Y0
G1 X77.94 Y-45
G4 S2
G1 X0 Y0
G1 Y90
G4 S2
G1 X0 Y0

The idea is that you watch the hot end head that is .2mm above your print surface and look for changes in height. Makes sense to me. So I run it and here is the part that makes me . When it lowers down to 0,0,0.2 everything is looking good. But then as it moves to each tower and back, first to X, then to Y, and then to Z, the hot end grips the paper more and more. Basically it seems like the more it moves horizontally the more it lowers the hot end. At the end the Z height is still supposed to be at 0.2, yet it is gripping the paper like crazy.

Has anyone experience this behavior?? It is incredibly frustrating!!

[redlight]

It sounds like your horizontal radius is still off a little.

This is the method that I use. It's tedious, but it does work.

I run a gcode file that sets the nozzle at the center of the bed at 0.2mm and I adjust the height so that the nozzle is a smidge lower than a 0.2mm feeler gauge (used for setting spark-plug gaps). I call that 0mm.

I then run a script that starts at the center at the new 0.2mm (really 0.39), pauses for 10 seconds so that I can test it there, it then moves to the middle tower (Z I believe). On the gauge set I have I use a 0.2, a 0.15 and a 0.05. I have the 0.05 on top and it will flex if the gap is too tight. I use the 0.4 height as a safety and when using proper gauges the calibration height doesn't matter.

I set the middle tower first because it affects the other two equally. I then re-zero the height and test the middle, left and the right, before making any adjustments.

If they all need adjustment in the same direction, I change the horizontal radius. I too start with 0.2 changes, but the one I did tonight actually required a 0.02 to make it perfect.

Every time I adjust an end-stop, I re-zero and start the whole dance over. Often once you get everything really close, only one tower needs tweaking.

Finally, when I have them all set, I run the gauntlet script which does them all. I have a 10 second pause at each tower so that I have time to test properly.

I hope this helps you. I really dislike the process, but haven't found a better method.

[jeffhuber]

https://www.dropbox.com/s/aqft0hz97zjxr ... 5%20PM.mov

I made this video to further clarify the problem (only 26s long). This is me running Guanu's calibration script. The hot end is supposed to stop .2mm about the bed. Look at where it stops at the beginning of the video (looks like right at .2mm) compared to where it stops at the end of the video. Looks like 0!!!!

This is really infuriating. I ran run the z-height calibration over and over, adjusted the horizontal radius by .1mm over and over, and had all the tower scripts exactly aligned. But then I run the easy calibration script and it highlights the problem. Whenever the printer moves, the lower and lower it gets.

What is going on????

[guanu]

set your z=0 about .5mm off the bed for the first round of adjustments.. that first move looked like it dug down pretty good... get it till its "about" even, then reset your z=0 closer to the glass to make finer adjustments... this should help level it out.. and I agree, I've calibrated over 400 machines, and since that script only has start and end points rather than fluid motions, as it goes it does get progressively lower by a tiny amount, but as long as you compare the center pause, to the tower it goes to , then "reset" your eye when it pauses in the center again and compare THAT pause to the tower, you will have a good level... and reset z=0 when you make screw adjustments as that changes your height... but that first move in the vid is dropping good, so you have larger adjustments to make than the .2 will be visible, so I'd set your z=0 to a bit of a gap to get the large adjustments out of the way..

Guanu

[jeffhuber]

Well here is some more data in the WTF is going on with my printer.

guanu - i did just as you said and then loaded up some nylon bridge I had meaning to try. Set up my hot end to 230, bed to 60, loaded in a hollow cylinder to test the alignment and hit print.

Took a bit to stick but then things were running merrily for the first few layers and then i hear BAD noises.

I look over and the hot end has moved to the side of the cylinder and is literally grinding into the glass. (youll notice it's off center, repetier put it there and i didnt center it. it didnt shift on the bed - hair spray was in full effect)

Here are the pics:
[img]http://i.imgur.com/p1zcAPyl.jpg?1[/img]
[img]http://i.imgur.com/4dmKukcl.jpg?1[/img]

I'm curious - could this be because a stepper motor is going? The deviation is in the direction of the X tower (I think - X is directly right of LCD right?)

Thanks Guanu!

[PhoenixNZ]

The deviation is in the direction of the X tower (I think - X is directly right of LCD right?)

You mean the Y tower?

X is left of the "front" of the machine, Y is the right, and Z is the center-rear. Unless you;ve wired your machine different to the manual

[jeffhuber]

Yes, then Y. Thanks Phoenix! My mistake

[jeffhuber]

This video CLEARLY shows the problem I'm having. it has to do with the X axis...

https://www.dropbox.com/s/294qpjdi22j86 ... 8%20PM.mov

is this a bad stepper motor?? Looks like maybe the belt is slipping?

[Gr8Scott]

I had a similar issue where my printer would not hold a consistent Z height. Took a LONG time to figure it out. My belt pulleys were slipping on the shaft of the stepper motor. I would do a setup and have everything really working great one moment and then the effector was dragging the plate in no time flat. Looks like two of your pulleys are slipping to me.

[jeffhuber]

that sounds VERY familiar. after the machine is "homed" it's fine, but slowing loses fidelity on the z-height from that one tower.

By pulley - do you mean the ball bearings?

Is there a "sure sign" that is the problem? What did you do to fix it?

[Mac The Knife]

I think he means the drive pulleys, You may want to check that the setscrews are tight.

[jeffhuber]

Great diagnosis! It was the drive pulley on the X tower that was loose. Tightened it - and I'm not seeing the same problem now. Going to cross my fingers, mark this solved, and try to print something!

[Gr8Scott]

My Z axis was the first to slip consistently. I figured it out by moving the cheapskate plate up and down the rail by hand and noticed a slight "thunk" in the feel when I changed direction moving it up or down. I could tell something was slipping because it had a delayed effect before the "thunk" when the drive pulley finally caught on the stepper shaft and stopped slipping. Gravity will usually cause the slippage to move away from the drive pulley that is slipping. That is the surest way to diagnose the problem. Took me several months to figure it out.