Struggling with calibrating delta printer

[jonharri]

I recently assembled a delta printer kit. When I started to calibrate, I found that when I brought my print head to 1mm above the print bed, the Z height read 255. I pulled up the firmware, and #define MANUAL_Z_HOME_POS 295.4 is what is showing. I subtract the 255.4 from this and enter MANUAL_Z_HOME_POS 40 and upload it.

I also had to change " #define min_software_endstops false // If true, axis won't move to coordinates less than HOME_POS." from true to false to get the print head to get to the print bed. It wanted to stop 40mm below the end stops because it thought that was bed (or 0).

Now I have a new problem. When I try to home the printer after it has been lowered to the print bed, it raises all three axis at the same time until it gets about half way, and then only one axis continues. This make the print head crash into the guide rails. Any idea why it does not raise all three axis all the way?

[Roxy]

I'm bringing up a new Delta printer and struggling through the calibration issues. Most of my problem is I start writing new firmware to help me do it.

Install Maxima (from SourceForge, and go get the calibration.wxm spreed sheet. I haven't gotten through it, but it is supposed to get you really far up the curve very easily.

[LuckyImperial]

I've spent the past several weeks walking a new guy through the calibration of his delta printer in this thread: http://3dprintboard.com/showthread.p...e-serious-help

I would encourage reading through that entire thing, or at least the posts of mine.

I assume you're using Marlin firmware? I would highly encourage you to convert to Repetier Firmware.


The "MANUAL_Z_HOME_POS" should be the printable height of your printer, which is likely around 200mm. "Z_MIN_POS" should be set to 0.

A common error is that people use their Z-Probe as a Z_Min Endstop. It's not an endstop, it's a probe. This implies that all minimum endstop pullups are disabled. Under probing configuration you pull up z_min and set those pins as your "probe". The reason your printer wouldn't move down is because the min endstops were pulled up and were sending a high signal...so non existent endstops were telling the firmware that they were being hit.

[LAFilament.com]

Please post solutions!

I have had my firmware corrupt a number of times, requiring that I reflash

[LuckyImperial]

Please post solutions!

I have had my firmware corrupt a number of times, requiring that I reflash

What firmware do you use? Marlin or Repetier?

Also, I recommend you start your own thread. This one was started by jonharri and is specific to his issue.

[Alibert]

Install Maxima (from SourceForge, and go get the calibration.wxm spreed sheet. I haven't gotten through it, but it is supposed to get you really far up the curve very easily.

I have tried this route (as well as several other algorithms I programmed myself) and it is no panacaea. The problem is that with a limited number of points (7 for the link you refer to) there are multiple solutions. You will get a result, but if you use a more stochastic algorithm around it and run it thousands of times, you get a host of possible 'solutions'.

My next attempt (to automated calibration as opposed to auto-levelling) will be to implement a z-probe on my delta printer and modify the firmware to scan and output all data points of the whole bed (>100 or so) over the serial port so I can log them and use the whole matrix for a downhill simplex optimization. That will take some time to organize and program, so until then I am stuck with the manual method, as are all of us I guess....