Kossel calibration issues...

Hi there, I have been continuously calibrating my kossel for the past weeks but I cant seem to get my nozzle to move parallel to the printbed in all directions.
To be honest, it is getting pretting frustrating
The problem here seems a little weird to me. I have followed Minow's calibration guide and used the following calilbration technique:
1) Adjust the endstops to make sure they all give the same height in front of the X, Y and Z towers.
2) Check the Z height at coordinates x,y = (0,0)
3) Adjust accordingly, by either increasing or decreasing delta radius, (done by editing smooth rod offset)
4) Check and repeat

So far I have managed to get the nozzel to move quite parallelly to the plate.
I brought the nozzle down to the lowest position and then moved it in the positive-X, negative-X and positive-Y directions and they all seemed to move rather flatly to me
However, when I try to move it in the Negative-Y direction, the nozzle would move into the printbed after a certain distance.

Another issue I noticed is that sometimes, if I were to move the nozzle in the negative-X direction, it would move flatly with respect to the bed. But if I were to move it in the positive-X direction, it would start dragging against the bed. At any point if I were to move the nozzle in the negative-X direction again, the nozzle would once again be above the bed, even at points were it was previously dragging against the bed when moving in the positive-X direction...

I have checked and the bed should be level with respect to the ground. If it were not level, I would think the the nozzle would move away from the printbed when I move it in the positive-Y direction, but that is not the case.
Also, my parallel rods are all of same length and the effector does not rotate when moving. It stays perfectly flat.

Anyone ran into such an issue before?
Real sorry for the long and confusing post. Thanks in advance for any help that i can get
Cheers!

Edited 1 time(s). Last edit at 01/17/2016 06:38AM by Wiireless.

Quote
Wiireless
So far I have managed to get the nozzel to move quite parallelly to the plate.
I brought the nozzle down to the lowest position and then moved it in the positive-X, negative-X and positive-Y directions and they all seemed to move rather flatly to me
However, when I try to move it in the Negative-Y direction, the nozzle would move into the printbed after a certain distance.

That suggests that your tower positions offsets need adjusting. Try my delta calibration calculator at [escher3d.com]. Caution: if you select Repetier firmware, you need to reverse the sign of the endstop corrections (I will fix this soon).

Quote
Wiireless
Another issue I noticed is that sometimes, if I were to move the nozzle in the negative-X direction, it would move flatly with respect to the bed. But if I were to move it in the positive-X direction, it would start dragging against the bed. At any point if I were to move the nozzle in the negative-X direction again, the nozzle would once again be above the bed, even at points were it was previously dragging against the bed when moving in the positive-X direction.

It may just be that when the nozzle was dragging, one of the motors was missing steps.

Edited 1 time(s). Last edit at 01/17/2016 10:19AM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Hey, thanks alot for the reply!

I have checked out your calculator although I think that I may be doing something really wrong since the problem is getting worse and worse with each change I make...
I am using Marlin's firmware by the way. The latest revision from the github, Release 1.1.0
I understand that the endstop adjustments can be made using the EEPROM and M666 commands. For the delta radius and z height, I manually adjusted the firmware by changing the value of DELTA_SMOOTH_ROD_OFFSET and Z_HOME_POS respectively.
However, I am not really sure how to go about editing tower position angle corrections.
I tried adding new definitions for SIN and COS under marlin_main and then changing the respective delta_tower_x and _y values.
How do I actually go about adjusting these values?

Cheers!

Wiireless,

If the problem gets worse with each change then something is clearly wrong. Did you select the correct firmware type in the calculator? Did you enter the correct initial values into the calculator? Are you entering the correct sign of the height error, i.e. positive where the nozzle is too high and negative where it is too low?

If you believe you are doing these things correctly, please try using the calculator in 4-point mode. Take measurements in front of each tower and at the centre, and compare what the calculator suggests with what you would do if following Minow's calibration guide. Post the results of both here.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Hi David, okay I will try the calibration process again hopefully tomorrow and will update again.
However, I would just like to confirm if what I am doing to adjust the "tower position angle corrections" is correct. A.K.A changing the sin and cos values in Marlin_main

Cheers!

Quote
Wiireless
Hi David, okay I will try the calibration process again hopefully tomorrow and will update again.
However, I would just like to confirm if what I am doing to adjust the "tower position angle corrections" is correct. A.K.A changing the sin and cos values in Marlin_main

Cheers!

Some versions of Marlin have tower angle corrections in the configuration.h file. Otherwise I guess it's probably angles in the main code, but I haven't checked, and you would need to be careful to get the signs of the corrections right.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Are you using Astrosyn dampers? I noticed that on my delta which uses them I had the belts too tight and that caused the X stepper to flex upwards and the belt to ride up on the lip of the pulley as it moved.

I'm hoping that once I fix this my weird calibration issues will be a thing of the past.

Edited 2 time(s). Last edit at 01/18/2016 05:38PM by WZ9V.

Wiireless, I can pretty much guarantee that your steps per mm is off.

To measure this, home your machine, then get a set of calipers and place them so that they will measure the carriage movement precisely (one jaw against the frame, another against the carriage). Now move the carriage down by say 50mm using pronterface.

I can guarantee that your carriage won't move down 50mm. Calculate the percentage difference and adjust your firmware to suit.


Now home again, and you'll have to go through the whole calibration process again. But this time when you get it going flat towards the X, Y and Z towers, it will be flat everywhere else.



* I know because I had exactly the same issue

Edited 1 time(s). Last edit at 01/19/2016 01:59AM by nebbian.

Quote
dc42
Wiireless,

If the problem gets worse with each change then something is clearly wrong. Did you select the correct firmware type in the calculator? Did you enter the correct initial values into the calculator? Are you entering the correct sign of the height error, i.e. positive where the nozzle is too high and negative where it is too low?

If you believe you are doing these things correctly, please try using the calculator in 4-point mode. Take measurements in front of each tower and at the centre, and compare what the calculator suggests with what you would do if following Minow's calibration guide. Post the results of both here.

Hi, sorry to only update after so long.

I tried using the 4-point mode and it was much better (but the results were still not optimum). Somehow then 6-point mode would make my the calibration worse and worse with each iteration.
However, the 4-point calibration does not always work immediately. From previous manual calibrations, I found that my delta radius is around 111.3 range. The calculator however will sometimes constantly decrease the radius futher and further each time to around 109 or less (Yes, the values that I entered into the calculator are correct). If i were to remove all endstop offset values and try again, it would then usually work. The flatness is somehow still not there though, but this is enough for me to print at around a distance of 60mm from the center in all directions, before the printhead becomes too high (I have a 170mm bed).
Also, I cant seem to get the deviation to less than 0.09mm. It is usually withint 0.11 to 0.14mm. Is there still anything that I am missing out??

Quote
nebbian
Wiireless, I can pretty much guarantee that your steps per mm is off.

To measure this, home your machine, then get a set of calipers and place them so that they will measure the carriage movement precisely (one jaw against the frame, another against the carriage). Now move the carriage down by say 50mm using pronterface.

I can guarantee that your carriage won't move down 50mm. Calculate the percentage difference and adjust your firmware to suit.


Now home again, and you'll have to go through the whole calibration process again. But this time when you get it going flat towards the X, Y and Z towers, it will be flat everywhere else.



* I know because I had exactly the same issue

Hi, I tried measuring and the steps per mm seems to be correct. They are all the same values for all 3 axis and the microstepping are also all at 1/16 so I dont think there would be any discrepancies in the values.

Quote
WZ9V
Are you using Astrosyn dampers? I noticed that on my delta which uses them I had the belts too tight and that caused the X stepper to flex upwards and the belt to ride up on the lip of the pulley as it moved.

I'm hoping that once I fix this my weird calibration issues will be a thing of the past.

Nope, I am not using at dampers with my current printer. But I think maybe I should though. Never hurt to cut away the noise

I personally got nowhere with minows guide. For one thing, there's some stuff in there that seems like it must be realy dated and obsolute. He seriously advises to avoid the EEPROM and instead go through the brain damage of editing configuration.h, recompiling, and re-uploading with each change?!?!! You have to be kidding!!
This video on the other hand really helped me out: [www.youtube.com]

Quote
BlasphemousBits
I personally got nowhere with minows guide. For one thing, there's some stuff in there that seems like it must be realy dated and obsolute. He seriously advises to avoid the EEPROM and instead go through the brain damage of editing configuration.h, recompiling, and re-uploading with each change?!?!! You have to be kidding!!
This video on the other hand really helped me out: [www.youtube.com]

Yea, I sort of gave up on adjusting the manual endstops after awhile and started using EEPROM to adjust my endstops, though I still had to adjust the firmware to change the z_height. I think it would be possible and much easier if I had the endstops with adjustable screws to change the height.

Anyway I think the wise thing for me to do now is just to rebuild the whole delta and I dont think whatever I have currently can be saved with software alone.