Strange ??extruding?? behaviour.

Hello, I own a Kossel mini delta printer and after some first prints I corrected the things that were wrong but I cant deal with this one. Just look at the picture, if you would have an idea how to solve this, or just try something, please help.

Thanks so much for every reply !




Edited 1 time(s). Last edit at 09/08/2015 03:53PM by Shepard48.

You need to level your bed. Where you have thick lines, the nozzle is closer to the bed and the extrusion is being smooshed down more.

It could be that your print bed isn't level. If the nozzle is lower, it squashes the filament out a little more, which could give the appearance of thicker lines.

If you print the same piece the other way around, are the thick/thin lines still on the same side?

Thanks for the suggestions, I tried releveling the printer by the paper drag method with the X Y Z endstop offset settings in Repetier FW in EEPROM settings, and also recalibrated the DELTA_RADIUS setting. And BOOM nothing.... Printed the same part, the same "bad" lines came out. I wondered what could be the problem and spotted something on the raft of the part. Please look at it. Thanks again !!

Chances are that your towers are leaning slightly or you have some other build inaccuracy, so that even though the nozzle height is correct near the towers and at the centre, it isn't at other places. Try printing a calibration spiral (there are some on thingyverse), which will show you how even the nozzle height is at various positions on the bed.

---

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].

I have/had similar problem (one of my towers is slightly closer to the bed than the others but still have the right angle) when I used Marlin, although not so obvious, I could correct the issue following the instructions and adding the lines of code from this page, did not use the test-object from that page as I had other issues which would not allow me to print it properly, but I used the object from this page

Now I have a duet with reprap firmware and this issue have again appeard, but it seems there is no fix for it as of now.

Okay, thanks for reply, I tried to correct some leaning by adjusting the screws on the top triangle and tightening the belts after it. After that I printed that handle again and it improved a tiny bit, not much. Lets say I dont like autoleveling and also my arduino performance doesnt like z correction and I tried setting up the probe for like a month and couldnt get it to work. So if its possible I wouldnt like to use autoleveling. Is there any other way, solution to leaning towers ? If its really caused by it. I saw some delta tower angle correction setting in repetier fw but I am not sure how should I measure my angle errors, since here a 0.5 degree angle can cause big problems. Anyone knows a tip or trick to do this ?? If the autolevel is the only/best way, then I will take another try to make that probe working. Thanks again.

Sorry Koenig, you wrote just before my post, Thanks for that site, will try that tomorrow and write what happened

I had a similar problem on my Mini Kossel, caused mostly by leaning towers I believe. The calibration spiral showed that there while the heights were the same in front of the towers and at the center of the bed, there were ridges between two pairs of towers and a valley between the other pair. In theory you can correct for this to a first approximation by adjusting the tower positions in the firmware. But I couldn't find a mathematical model that would tell me what tower position corrections to make.

I was already in the process of converting the machine to 32-bit Duet electronics, so I ended up implementing a least squares calibration process in firmware that computes the endstop corrections, delta radius and tower position corrections from a single 30-second probing of the bed. See [youtu.be] for the video.

At present, only RepRapFirmware supports this accurate auto calibration. However, I am told that Rich Cattel's fork of Marlin can often do a reasonable job of auto calibration if you allow it enough iterations. So you might want to try that.

Edited 1 time(s). Last edit at 09/09/2015 06:00PM 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].

Okay, I have been experimenting with the Z probing last days but I could not get reasonable results. I will do the calibration by the 2nd link that Koenig posted. Thanks for that.

Quote
Shepard48
Okay, I have been experimenting with the Z probing last days but I could not get reasonable results. I will do the calibration by the 2nd link that Koenig posted. Thanks for that.

You can download this fork of marlin, it already has the necessary lines added in the right places.

I noted that the main-fork of marlin also uses this system as a part in their auto-calibration.

Thats not a bad idea ! I used Marlin before but it has improved so much. I will have a try with that, I can set it up but if someone could send me an autocalibration guide for marlin, it would be really useful. Thanks Koenig!