Anyone have any problems as to setting up a Kossel Mini? I still need to tune the motors, use OpenDACT or manual calibration, and start printing cubes…
created
Jun '16last reply
Jun '16- 13
replies
- 2.5k
views
- 5
users
- 1
like
- 4
links
Anyone have any problems as to setting up a Kossel Mini? I still need to tune the motors, use OpenDACT or manual calibration, and start printing cubes…
Here a procedure a friend if mine from 3D Printing Academy in Hong Kong sent me:
During our last workshop at CDNIS, we have developed a new, streamlined way to calibrate the Deltas without the need of the Autolevel Probe! It looks complicated but if performed carefully, it should not take much longer than 15 minutes per printer. We did 10 printers in one afternoon:
New Calibration without Z-Probe
a. Important: wash Glass Bed with soap to remove any residues from factory. The residues render the UHU glue useless.
b. Start Printrun and enter: M666 H265 X-5 Y-5 Z-5, store with M500
c. Home, then move Head to X: G0 X-73 Y-42 Z5 F6000 (you can program it in Printrun)
d. Manually move head down until nozzle is paper thickness above glass bed
e. Example: If display shows 2.1mm, then 5 - 2.1 = 2.9 enter: M666 X-2.9
f. Home, and repeat for X until nozzle exactly 0.1mm above glass bed; store with M500
g. Home, then move Head to Y: G0 X73 Y-42 Z5 F6000
h. Repeat d) e) and f) for Y
i. Home, then move Head to Z: G0 X0 Y73 Z5 F6000
j. Repeat d) e) and f) for Z
k. Repeat again c) - j) for X, Y, Z and adjust if necessary
l. Store with M500
m. Once X, Y, Z are all exactly 0.1mm above glass bed, home and move head to center: G0 X0 Y0 Z5 F6000
n. Manually move head down until nozzle is paper thickness above glass bed
o. Example: If display shows 0.1mm, then (250 – 0.1 = 249.9) enter M666 H249.9
p. Store with M500
Your printer is now perfectly calibrated and does not need G29 Autolevel anymore!
Remarks: If you forget the settings, you can always check with M666 L.
If you get inconsistent results then most likely one or more of your spur gears are slipping and/or endstops are not well fixed. Check and tighten.
I have tried doing it with that method and the delta radius with the dish and bowl was always throwing it off.
Following this thread.
I have a mini Kossel with a MKS Mini which doesn’t seem to want to accept the configuration I acquired from github. Also, controller software is an issue, Repetier and MatterControl don’t play well so it’s one or the other - the Rostock wins hands-down. A couple of days ago I purchased an android app to control the mini, the OTG cable is on order and I’m excited to try that approach. Any advice on any of this appreciated.
Additionally, the arms are nothing like the samples I’ve found online so my measurements are based on my best attempts and could be an issue.
Anyway, I’m looking forward to reading any responses on this subject as the reproach forums seem to be deficient on info these days…or my researching skills need honing.
What method did you use to make the arms? Did you use a paint stick jig?
I meant diagonal rods, not arms. I didn’t make the rods, they came with the printer and I’m not certain who makes them and most importantly, what measurement I should use…end to end, eye to eye…? That problem is secondary to the firmware and controller issue. Once I get past that I can work on fine tuning the configuration
It’s tie rod hole to tie rod hole. What’s the other issue?
On mine I have Ramps 1.4 and Arduino Mega 2560 with Repetier and Repetier Host. That’s what I’d suggest.
Hole to hole, that’s useful. The other issue is the controller, I use MatterControl for the Rostock which is not playing well with the mini. I installed Repetier which caused issues with MatterControl (an issue that has been plaguing my host system since). The configuration in Arduino for the mini is straight from github, and with no other configuration to use I’m stuck with the mini doing weird things. Such as, the head moves incredibly slow to the bed, and will not calibrate z height properly…I was frustrated and out of time before leaving it awhile.
These are current issues that I haven’t really looked at since my last round of troubleshooting. I have a mega and ramps, all the things I need to switch over. The installation and configuration is really the next step.
to comment on the delta vs Cartesian design as a first printer: the Rostock max is a great machine and hasn’t disappointed me yet, which was a deciding factor in my purchasing the mini
I would just use Repetier-Host because it works with my Folgertech Kossel Mini Rev A. As for the head moving slow to the bed, it seems like the feedrate is off. Another thing is that if you have a Marlin configuration I found a Repetier page that allow you to upload the Marlin configuration and have the values transfer over. I’ll let you know when I find a link.
Good choice on printer. I’ve used various kinds, and the Kossel Mini + that I built seems to work the most reliably. I respect the choices that other have made, by my personal experience has been positive. I’d suggest an auto calibration probe and a firmware that handles it for anyone building one. I like https://github.com/RichCattell/Marlin/tree/Testing 3 for the firmware, and the IR probe from maker DC42 is amazing, and distributed in the US through Filastruder: Mini differential IR height sensor - Filastruder 2. If your Repetier setup works, keep it. The probe has been amazingly accurate in getting my printer to pretty sick calibration, and I’ve had the bed off once or twice without even having to recalibrate.
I also love and recommend the Bondtech Extruder - it’s expensive, but I have yet to have any single issues with extrusion related to it. Once you get proper extrusion/retraction feedrate and E-steps dialed in, it’s extraordinarily reliable and even prints Ninjaflex through a 600mm bowden tube without any complaints.
It sounds like you’re going through the mechanical calibration and are interested or have purchased a probe, all of which will make for happy printing. The instructions left by Cello below are what I’ve seen used elsewhere and should get you very close. I found that the additional probe firmware calibration made things perfect. I’m not sure what your issue with delta radius is - it may be prudent to re-measure the parameters for whatever Repetier uses to calibrate it.
For calibration objects, I’ve done away with cubes for the delta and moved on to this object: Calibration Reuleaux Triangle by senormeow - Thingiverse 3. If it prints and the inside diameter is on and the point-to-point distance from each of the vertices is the same, you’re golden. If not, you can orient the object so that each point faces a tower and then are able to quickly identify if your rods or off or something’s not right otherwise with a tower offset calibration number. For delta radius calibration and tweaking, something like these concentric circles works (and also is helpful for getting the bed mechanically flat): Bed Levelling Spirals by clefranc - Thingiverse 1. If your radius is slightly off they will be noticeably different in height from the perimeter to the center, and any high-side on the bed will be obviously apparent. If it’s trying to drill your hotend into the center of your print bed, I’d get that fixed before trying these.
That would be great, I’ll scour in the meantime and hopefully locate something useful
So I’ll do the same. My kossel mini is still not correctly tuned for 90° angles. I hope as I can start this week.
Yeah, calibration is an issue even nowadays. Anyone ever used OpenDACT?