Indicator Leveling showing odd numbers...

[Zero3D]

Hello all. So, Ive had my printer for about 6 months, I have had great success with smaller parts and am now realizing I need to calibrate this machine a lot finer than what I have right now for larger prints. I currently have a dial indicator setup off to the side of the extrusion head (about 2 inches from where the extrusion would touch the bed). I'm a little new to all the different settings and software adjustments but this is whats happening... I level the bed using the end stops and a piece of paper. When i get them set i write down the indicator number. I do this for all 3 end stops. Shouldn't my numbers be consistent? The rearward end stop is out by .8mm from the front right. The front left is out .2 from the front right. Nothing is adding up. I have played with the horizontal radius to get my center point to touch as well which is also not level to any of the end stops as well. I'm not fully understanding this. Maybe someone can give me some insight on how they are using a dial indicator to level everything. I'm not sure if the bed needs to be shimmed, if I'm getting that much of a "bowl" shape. Also Ive ran the Onyx Bed Leveling print and in one corner of the bed it wont print on because the head is too tight to the glass... I'm so confused! Thanks guys

[Nylocke]

The indicator is a better replacement for the paper method. You use it because it gives you a quantitative number instead of a feel. You should adjust the screws so that they are all consistent according to the indicator. If x is .2 off of y, then adjust x so it's not. That's badically how to use it. It also helps you dial in the radius to damn near perfect. Same procedure as with the paper, you're just trying to hit the same mark at every point.

[Zero3D]

Alright, so this is what i figured the indicator would help with but.... when using the indicator (which is offset from the location of the hotend)... trying to get all the points to match would make the hotend crash into the bed or vice versa. I'm not fully understand the geometry of the delta yet but it appears its tilting the effector. Why i think this, is because, with the 2 inch offset from the hotend to the hotend, the hot end is touching meaning the center of the effector is the same distance from the glass as the other 2 axis's but since the effector is tilting its giving me odd numbers. To fix this i will be making a plate that will support my dial indicator directly in the center of the effector to reflect the hotend as much as possible. I did not understand that this is how the deltas work, maybe this isnt how they are supposed to work but its what mine is doing.

[Nylocke]

Most people do directly mount their indicator in the middle of the effector, they level their machine, then they put their hotend back on and zero out the machine.

[Zero3D]

Also... This is my plan of attack for getting this to work, please let me know if I need any additional steps. I will be adjusting the belts, Getting the printer within its original tolerance (small prints looked great), printing a center mount dial indicator attachment. Setting the z-height with the hotend. Then switching over to the dial indicator. Using my z height measurement on the dial indicator as a baseline for the z height. Putting the set screws to the default location. Reset the Z-height to get the same reading i was previously getting. Then moving on to using that z height on the indicator to match my X Y Z tower adjustments. That's the plan... Let me know if you think I have something out of order. Thanks.

[Nylocke]

Your Z height is irrelevant when leveling with a dial indicator, since you can always set it by touching off with the extruder when you remount it. All that you're concerned with is "is the dial indicator reading the same at the 4 points, center, x tower, y tower, z tower". Just focus on making sure your machine moves parallel to the bed and you can worry about the z length later.

[Zero3D]

I thought that afterward. Thanks Mike! If anyone has any G-code that makes calibrating easy let me know. I wouldn't mind g-code for running the the outer perimeter but i guess i could just use the Onyx bed leveling sketch dry.

[Nylocke]

The routine that gene has in the manual should be fine to use.

[guanu]

Calibrating the Guanu way.... this is how I calibrate orions (and when I build an assembled rostock max here)

for me its easier to watch a gap between the nozzle and bed to check movement up or down... this will also cover horizontal radius for pics check page 25 of the orion manual (the only difference is the calibration gcode)

1) Preheat hot-end and bed to the material you use. If you use both, just use the “Preheat ABS”
in the printer settings menu on the LCD.

2) Once the hot-end and bed are up to temp, on the LCD, click the knob to bring up the menu,
go to Advanced Settings (at the bottom of the list). Click the knob.

3) Go to the bottom of the next menu to Calibrate Z Height. Click the knob.

4) Now you are in the calibration menu. Click Home Towers. The machine will go up and hit all
the end-stops.

5) Go down to Z Position and click the knob.

6) Turn the knob Counter-Clockwise to lower the nozzle. Bring the nozzle down till its a couple
millimeters from the glass like the following pic.

7) Turn the knob counterclockwise SLOWLY and look eye level with the bed, and bring the
nozzle down till it just barely meets the glass (this is easiest with a white wall in the background
so you can watch the reflection of the nozzle – if you don't have a white background to use, you
could tape a sheet of paper between the towers to give you enough contrast). When the nozzle
touches the glass, click the knob.

8) Go down to Set New Z=0.00 and then click the knob.

9) Go to Home Towers. Click the knob. You have now set the z height, and now it is time to
calibrate the towers.
The next steps can be done through a computer or SD card. I find it is easier on a sd card so
you can keep close to the machine to watch the movement.

10) Open notepad or a text editor of your choice.

11) Paste the following into the text editor: ***modified for Rostock Max!***
; Tower endstop calibration script
G28
G1 Z.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
;

12) Save the file as “Towers.gcode” and place it on an SD card, or if you are going to be using
a computer, load it into your host software that you use.

13) Again with the bed at eye level and looking close, Run Towers.gcode from either your host
software or SD card. Watch closely as the machine will home, then the nozzle will drop to
0.2mm ABOVE the glass.
After it drops to the center like that, it will travel to the X tower. DO NOT pay attention to what
the nozzle does while traveling, what you must pay attention to is when it pauses.
The nozzle will pause at the X corner, then return to the center, and move to the Y tower and
pause. After the pause, it will again move to the center, and move to the Z tower, pause, and
return to the center.

14) What you want to remember is the nozzle when it pauses. You want to compare the
movement to the gap at the center.
– If the nozzle at the tower RAISES (compared to the CENTER), you want to adjust the screw
that hits the end-stop by turning it counterclockwise (moving the screw up).
– If the nozzle at the tower LOWERS (compared to the CENTER), you want to adjust the screw
that hits the end-stop by turning it clockwise (moving the screw down)

15) After you have adjusted the screws, You have changed the Z height, so you want to go back
and re-set the Z height the way you did earlier in this guide.

16) Re-run Towers.gcode and watch as you did before, again noting the changes in the nozzle
at each pause. You will run through these steps till the nozzle remains the same height at each
tower when compared to the center.
If the nozzle goes the same direction on all 3 towers (such as you have the gap in the center,
and at every tower the nozzle lowers. Or you have the gap at the center and at all 3 pauses the
nozzle raises), you will adjust the radius in the following way

17a) If from the center gap, the nozzle goes DOWN toward the glass at ALL 3 TOWERS, load
your host software and bring up the EEPROM information. You will look for Horizontal Radius.
You want to RAISE that number. I suggest raising it by 0.2 and run towers.gcode to see the
change, and keep raising the number till the gap evens out (changing this number will not make
you need to re-set your z height, it will just raise the outer edges where the nozzle pauses).

17b) If from the center gap, the nozzle goes UP toward the glass at ALL 3 TOWERS, load your
host software and bring up the EEPROM information. You will look for Horizontal Radius. You
want to LOWER that number. I suggest raising it by 0.2 and run towers.gcode to see the
change, and keep raising the number till the gap evens out (changing this number will not make
you need to re-set your z height, it will just lower the outer edges where the nozzle pauses).
After doing this, you will see any changes where one tower may be higher than another, if this is
the case, go back to adjusting the end-stop screws as before.
Typically it can take anywhere from 5-10 or so re-runs of the tweaks to get the gap to remain the
same at all 3 pauses compared to the center of the machine. Once the gap is the same at each
tower compared to the center, your machine is calibrated and ready to print!


Guanu

[Zero3D]

This is awesome! Thanks I really appreciate it!!!

[bot]

That is a perfect description of the calibration process (though you have a typo in 17b where you say the nozzle "travels UP towards the glass" (instead of away from...)).

I would add a few steps myself and augment the process slightly.

I would use the towers.gcode commands to set pre-defined scripts in repetier host (or pronterfacem etc) just to make it easier to go back and forth from positions.

I would also add positions OPPOSITE to the towers and use the EXACT SAME METHOD of calibration (except "opposite") for the opposite positions. I have experienced (and others seem to have, as well) a condition where the three tower points and the center point all seem perfectly level, but the "opposite" points differ wildly. I adjust the endstops and radius to make sure the "from tower-to-opposite-of-tower" travel of the nozzle is flat, rather than simply the "center-to-tower" travel.

[Polygonhell]

That is a perfect description of the calibration process (though you have a typo in 17b where you say the nozzle "travels UP towards the glass" (instead of away from...)).

I would add a few steps myself and augment the process slightly.

I would use the towers.gcode commands to set pre-defined scripts in repetier host (or pronterfacem etc) just to make it easier to go back and forth from positions.

I would also add positions OPPOSITE to the towers and use the EXACT SAME METHOD of calibration (except "opposite") for the opposite positions. I have experienced (and others seem to have, as well) a condition where the three tower points and the center point all seem perfectly level, but the "opposite" points differ wildly. I adjust the endstops to make sure the "from tower-to-opposite-of-tower" travel of the nozzle is flat, rather than simply the "center-to-tower" travel.

The problem with doing this is even if you have the issue you don't have enough degrees of freedom to tune it out, the 3 end stops allow to to effectively rotate the plane, the center Z point sets the vertical offset and Delta radius changes the radial movement effect. You'd need an extra variable to tweak, either the tower positions or the arm length.