Rostock Max Z Axis help

[truenorthtrader]

Hello folks

I'm having problems calibrating the machine

No matter what I do I'm getting the same measurements at the Z axis...about 8 mm or so higher than the middle.

Is there a video on step by step calibration somewhere? I've gone by the manual and the wiki site but I'm obviously missing something.

thank you
A

[wonderemporium]

This sounds like it could be the same as a problem I was having although my variance was only about 2mm from the center to the edge. If you are seeing 8mm difference you may have a more serious mechanical issue but check out this post...
http://forum.seemecnc.com/viewtopic.php?f=37&t=1024

[barnett]

just some suggestions:
- obvious, but I'll note anyway: you have to send G28 after adjusting the endstop actuator screws.
- are you sure the endstop actuator screws are going in/out when you turn them? Can you measure them moving?
- do changes to the endstop screws on your other columns produce results?
- Is anything slipping? After measuring nozzle to bed at the center and then measuring at the XYZ columns, if you move back to the center, do you get your original measurement? If not, then you may have chewed belts or timing pulleys slipping on on the motor shaft
- Is your bed flat?
- could there be sticking where the arms attach to the u-joints?
- I might also check to make sure there isn't any gap between the T-slot columns and the top or base. Not exactly sure what kind of trouble that would cause, but it might show up in calibration?

[Polygonhell]

If you write say 20 lines of gcode to move the carriage backwards and forwards over the center of the machine, can you see it arc in BOTH directions?
Something like

G1 X0 Y0 Z10 F5000

G1 F1000
G1 Y70
G1 Y-70
G1 Y70
G1 Y-70
G1 Y70
G1 Y-70
G1 Y70
G1 Y-70
G1 Y70
G1 Y-70
G1 Y70
G1 Y-70
G1 Y70
G1 Y-70
G1 Y70
G1 Y-70

If the platform arcs up or down in the middle, then I'd start by checking that the extrusions are seated properly, and that the cheap skates are assembled correctly and sitting on the extrusions.
It's most likely a geometry error, the firmware has to know the radius of the machine and the lengths of the arms to calculate the corresponding carriage paths for a given requested linear path, an 8mm error seems like a huge amount to me, which is why I'd start by suspecting mechanical assembly, if you rule that out I'd try the fix in wonderemporiums response.

[truenorthtrader]

Would have been nice if all this was in the manual in enough detail to work it out if you're not an engineer or had previous experience.
1st time the screws were NOT fully tight...but I have the ACRYLIC verisons and I'm getting CRACKS on various parts that I over tighten.

For Barnett all points ok except the slipping which I fixed and the bed being flat...I saw nothing in the manual about measuring the bed flatness, I assumed it was engineered with all the ports in the Acrylic that once tightened it would be flat, but I put a metal ruler across 2 ways and there were no gaps.


I noticed the Idler Mounts slipped a bit from each position for first time.
Once re tightened I did the Z Axis, recorded it, send to the board, did all axises and everything and it went well, only a couple mm differences at first.

I figured it out how to record the info in the Cofig h tab but after putting the Z value in and doing all axis, doing PID Autotune settings and sending them to the machine, I extruded plastic, then did my first cube print by the manual step by step and it the build plate.

It was rubbing a tad against the PLATE and wasn't even able to extrude anything, it just melted the pattern into the plate.
I even did a G1 Z0 F1200 to be sure it was still above the build plate and I could fit paper in the space just before the print!

Then it pushed DOWN INTO the build plate a short time after it started even after I set the Z axis...I assume this is the LOWEST it's allowed to go under any circumstances but it melted a hole twice into the plate while pushing into it.

I had to shut off and move it off the plate but it melted very fast.

My belts look fine I just put it together.
I checked Idler mounts and rail alignment and nothing has moved.

Why after setting the Axis does it go LOWER than it at ANY TIME in a print?
I thought the point was to restrict if from doing things that are dangerous?
Note I manually do a G28 before ANYTHING, then I opened the cube in slicer and Ran it after making sure the extruder was at 240, bed was at 55, and extruding a bit of plastic first.

I could have destroyed it so I'm not touching it until a video is put out with details from the Axis measurements to the first print as I'm too scared I'll destroy it.

Thankfully John says they are working on a video for Noobs like me.

Thank you guys so much for these details.
I want to run those codes Polygonhell but I'm a bit unnerved at the results at the moment but will soon.

[Polygonhell]

After you do a G28 IF the Zheight is set correctly in the firmware 0 should be "on the bed".
Having said that if you calibrated with a cold machine, when the hotend warms up 0 will be blow the surface of the bed.
Last time I looked at the manual the process omitted setting the ZHeight again after finishing calibration, when you adjust the screws to level the motion it also changes the ZHeight.
Most slicers (skeinforge is the exception) will not produce GCode with the head < 1 layer height.
For whatever reason Skeinforge always prints the first layer at 0.

[truenorthtrader]

OMG you're right! I heated everything then G1 Z0 F1200 hit the bed. Important info to know, I guess they'll get a more detailed manual with time.
I re-adjusted to the heated height THANK YOU!

I hope others new users like me are reading this.
Info could have saved me from ruining my Build plate. :/
Once I'm confident everything is working safely I can flip the build plate over to hide the small holes I guess.

[geneb]

I didn't include instructions on re-setting the Z-height because the Merlin firmware throws out fractional mm values and the new height is likely going to be a fractional value.

When I get the chance I'm going to re-write the calibration section to use a much simpler and easier to repeat method that doesn't involve using a caliper.

g.