Help Calibrating Delta - Printing issue

Hello All,

I have had this delta for a while and have not been able to get it working properly. I built it from scratch, it is a cherry pi.

Here is a picture of the problem

http://imgur.com/NQAwIB0

Basically the left is the x tower, right is the y tower and rear is the z tower

The x circle has about 1/4 that is not close enough to the bed -

The y circle has about 1/3 that is waaay to close and prevents extrusion

The z circle has about 1/4 that is not close enough

The problem I have is they all seem to contradict one another. I have no idea how to level that to fix it. I have levelled/calibrate each tower individually and the center, and with the paper test they all see reasonably close. I can't understand if the first layer of y is good for more than 1/2 the circle how could 20mm away it be so close to prevent extrusion. I think more likely it is a geometry problem, but I have been back and forth thru all that calibration, making changes and not been able to solve this issue.

Any help would be much appreciated.
Thanks!

Have you tried printing a large circle that is centred on the middle of your bed? This will show any miscalibration clearly.


It looks to me like your print surface is crinkled like this:



If so, then this is caused by an incorrect steps per mm value in your firmware. The easiest way to measure this is to:

Home your printer
Zero a pair of calipers on the top rail, and the carriage
Move the carriage down 20 mm using your computer
Adjust the steps per unit value in your firmware until you get a perfect 20mm movement.

Now go through your standard calibration routine, measuring your diagonal rods carefully, and calibrating endstops and delta radius again.


I hope this helps.

Edited 1 time(s). Last edit at 02/02/2016 11:22PM by nebbian.

Yes, exactly, I feel like it is a wavy washer but it is glass, and I have checked and it is flat, at least to the best of my tools for measuring. I hooked the printer up to pronterface. I told it to go down 20 mm and it went down 19.97 very close I then homed and told it to go 100 and it went 99.97mm so to the best of my measuring the steps are correct.

this is from my config file
I am running ramps 1.4 Marlin, 8825 drivers 1.7A 1.8deg nema 17 motors

The size of the circles I printed came out as expected, so I do not see how it is the steps that are wrong. But I can not figure it out.

Below is a pic of a circle from the center 1 layer
I had to put a white paper behind it so you could see what is happening
It is like it has a high spot thru the center

Pic

Edited 1 time(s). Last edit at 02/03/2016 12:23AM by n9jcv.

By print surface, I mean the shape of the path that the effector travels along. This is different to your heatbed or glass that you print on


Hmmm well if your steps per mm are spot on (and it does appear to be so), then you might try carefully measuring your diagonal rods (centre to centre). Be sure to measure from the centre of one ball to the centre of the other ball on the same rod. Ensure that your diagonal rod distance in firmware matches this value.

Also measure the width between the diagonal rods both at the carriage and the effector -- these should be exactly the same, otherwise you get effector tilt.


It doesn't take much miscalibration to get that crinkly surface effect.

Quote
n9jcv
Yes, exactly, I feel like it is a wavy washer but it is glass, and I have checked and it is flat, at least to the best of my tools for measuring. I hooked the printer up to pronterface. I told it to go down 20 mm and it went down 19.97 very close I then homed and told it to go 100 and it went 99.97mm so to the best of my measuring the steps are correct.

this is from my config file
I am running ramps 1.4 Marlin, 8825 drivers 1.7A 1.8deg nema 17 motors

The size of the circles I printed came out as expected, so I do not see how it is the steps that are wrong. But I can not figure it out.

Below is a pic of a circle from the center 1 layer
I had to put a white paper behind it so you could see what is happening
It is like it has a high spot thru the center

Pic

8825 steppers on a ramps 1.4 um are you using 1/32 micro stepping i wonder and if so would suggest that you may be missing an awful lot of steps the Ramps/Mega combo is known to struggle with Delta maths at higher speeds and 1/16 micro steps I have never run a ramps on a delta so don't take it as authoritative just it may point to the issues.

Doug

Have you calibrated tower positions, as well as doing the usual 4-point calibration of endstop adjustments and delta radius? The height error pattern you are getting does not appear to be symmetrical with respect to the towers, which suggests that you have either a geometry problem with the rods (see nebbian's reply), a tower position error, or a slight lean of the towers. You can use my calculator at [escher3d.com] to calculate the tower position corrections.

---

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].