There is a Z_PROBE_LOW_POINT setting in the firmware (configuration.h) that was also changed by makertech3d a while ago. I'm not completely sure about this, but if I understand it correctly, it limits the maximum range of the Z-probing from the printer's original z=0.0 point. If that's the case, it might limit probing into the bed when auto-leveling. It's currently at 3mm which should be sufficient if you follow the procedure from the build manual when mounting the z-probe sensor (resulting in about 1mm distance between sensor and hotend). It might fail when the sensor is mounted too high.

Not sure if you still have those issues and it's difficult to judge w/o additional information. to 1) not sure what you mean, are you saying you cannot remove the print from the print base/bed? to 2) shifts are probably related to mechanical issues like some of the axis being to loose or to tight, some grease might also help; it's also advisable to leave the bracket at the top of the z-axis - even if some say otherwise to 3) could be the temperature settings and/or thermistor seatin, although it looks more like too less than too much temperature; also the filament might be simply bad - it looks like you are using the makertech3d supplied red filament

Falls du nur eine feste Führung und 3 exzentrische Führungen für die Z-Achse bekommen hast, solltest du vielleicht ein Ersatz bei Makertech3D anfordern: https://www.makertech3d.com/pages/missingpartform Ansonsten sollte die Führung (aller Achsen) so fest wie möglich, so locker wie nötig sein. Allerdings ist auch irgendwann Schluss was man mit dem Design des AXIS überhaupt erreichen kann - ist nicht wirklich ideal aus mechanischer Sicht. Sollte zwar weniger was mit dem Z-Wobble zu tun haben, aber die Deckplatte für die Z-Achse sollte ebenfalls mit leichtem Druck auf die Z-Achse drücken.

Yeah. Congratulations to the results. Looks darn good to me and I guess everybody would have hoped for getting similar results just out-of-the-box. Have not touched my printer for some time due to lack of time to fiddle with it.

Die Ergebnisse sind nicht wirklich berauschend aber zumindest druckt er "etwas". Auf jeden Fall ist ordentlich Z-Wobble zu sehen. Du solltest die Führung (Guides) an der Z-Achse kontrollieren und gegebenenfalls nachjustieren. Falls du noch die originale Wellenkupplung aus Plaste benutzt, wäre auch ein Tausch gegen eine aus Aluminium empfehlenswert.

@Thomas S M503 würde alle aktuellen Parameter anzeigen. Mit M502 und M500 kann man einen Factory Reset machen. Wenn man nur M501 nimmt, werden Parameter aus dem Speicher neu geladen, also wie nach dem Start - wobei dann ungespeicherte Einstellungen verloren gehen.

@Thomas S der Abstand der nach G28 zwischen Druckkopf und Druckplatte verbleibt, wird mit dem Z-Offset korrigiert. Am einfachsten ist es, wenn man noch einmal mit Z-Offset=0 beginnt, d.h. M206 Z0 M500 G28 M420 S0 Im Display (bzw. mit M114) sollte Z=0.0 angzeigt werden. Danach reduziert man die Z-Höhe bis das Blatt Papier noch gerade so bewegbar ist. Die abgelesene Z-Höhe stellt man mit umgekehrten Vorzeichen als Z-Offset ein, also z.B. M206 Z1.8 bei einer abgelesenen Z-Höhe von -1.8mm. Speichern mit M500 nicht vergessen. Danach kann man mit G28 und G29 das Auto-Leveling starten, dort auch wieder mit M500 speichern. Noch ein Hinweis: Nach einem G28 ist immer ein Abstand zwischen Druckkopf und Druckplatte wie er ohne den eingestellten Z-Offset wäre. Der Z-Offset wird erst beim Drucken benutzt.

I think for small skews it is fine to simply add multiple skew factors together. The error is relatively small compared to adding the arctan() i.e. exact skew angles together.

@Thomas S you don't need to adjust the sensor trigger point with the paper method. All you need to ensure is that the sensor triggers with some distance from the printing bed (plate) to give some room for the leveling. To adjust the sensor, simply manually lower the hotend to touch the bed and get some piece of 1-2 mm thickness. Lower the sensor with the extra 1-2 mm piece between the sensor and the printing bed and fix the sensor with the 2 nuts. The hotend and the printing bed doesn't need to be heated. After that, you can follow the Z-offset procedure with G28 and the bed-leveling procedure with G29 from the manual. German: Es ist nicht notwendig den Sensor mit der Papiermethode auszurichten. Der Sensor sollte mit etwas Abstand zur Druckplatte auslösen, sodass genügend Platz für das automatische Nivellieren vorhanden ist. Zum Einstellen des Sensors reicht es aus, den Druckkopf manuell soweit herunterzufahren bis er die Druckplatte leicht berührt, dann sucht man sich etwas mit 1-2 mm Dicke. Mit dem 1-2 mm dicken Zwischenstück zwischen Sensor und Druckplatte reduziert man dann den Abstand und befestigt den Sensor mit den 2 Muttern. Druckkopf und Druckplatte müssen dazu nicht aufgeheizt werden. Danach kann man dann der Prozedur zum Einstellen des Z-Offsets mit G28 und dem Druckbettausrichten mit G29 aus der Anleitung folgen.

@Jay M the bed skew compensation corrects for errors on a pair of 2 particular axis not being perpendicular so in 3D space you have 3 such corrections. Usually, the x,y axis pair has the most errors on the AXIS printer - at least in my case. The bed leveling corrects for differences in z-heights between the plane with z=0 and the actual bed surface. The more uneven the bed surface is, the more points you likely need for correction. I'm not sure if 100 point bed leveling really helps as there are more factors that go into the equations, especially the mechanical accuracy and stability + temperature. Personally, I have only increased the number of bed level points to 20 with the default bed leveling method so far. The steps per unit (mm) was updated from 98 to 85 for the extruder in the firmware (configuration.h) but it should as be possible to just update the values on the printer with the M92 E command. The current value is listed with M503.

@Jay M It seems you also have some bed skew on the AXIS (same as I did). If you want to correct for it you can check the instructions on that forum post. Makertech also changed the steps/mm firmware settings for the extruder recently. I haven't been able to do anything with the printer currently but it might be worth to check if that would make any difference.

@Michael G Thanks for posting your projects :-)

Ok, trying to comment on a couple of things, except the quality thing - there have been enough comments already ;-) "When homing it moves a couple cm then stops." >> If you get a funny noise from the motors it would mean that the guides are too tight and the motor isn't able to move it further. You can try to loosen the guides a bit - or (carefully) increase the current provided to the motor with the drivers. "Moving the axis works (though had to invert the direction from what was posted on the guide)" >> I think that's normal when changing from A4988 to TMC2208 drivers --The display cables themselves are okay (I tried some spare cables from other printers/parts) and they were oriented the same way leading me to conclude it may be on the display board end itself where it may be flipped. >> The important thing is where pin 1 is on each end. The connector could be orientated correctly but still the cable could be assembled in the wrong way --Alignment of the axes is horrendously difficult. Also I don't quite understand the use of solid plastic linear guides (as opposed to some kind of bearing) >> Agreed. The use of plastic guides is because they are most likely cheaper than real bearings --Also would be nice to have some way of belt tensioning >> Same as above, would make the printer more expensive. You could print belt tensioners yourself or there are also some spring-like to buy from shops. --Would have been nice to have some of the wiring/cable/tubing precut/assembled. A touch dissapointing in that this is supposed to be a "kit" but part of it is "solder this yourself," "using a craft knife do this yourself," "cut this tubing exactly as described super precisely or it'll clog and bad will give you hell later" "cut off the ends of this wire and tie it together to another wire " (why not give a proper connector and attach it to the board via a splitter/connector?) >> Agreed, but again, the printer would probably be more expensive - but, maybe one would also say that it could have been cheaper with all the things one had to do themself. --On a similar note why is it that we have to PRINT a couple pieces (fan shroud, pi camera piece) for it to be fully functional? Could this not have been included in the kit? >> I guess that's also pretty common in that price range to print some accessories yourself - at least those things are not essential for using the printer. --I got the TMC 2208 add on and got a connection error. Turns out that for some reason it wouldn't work on E1 so had to revert to the A4988 for both E0 and E1. Upon further inspection the pin spacing on the board for E1 seem different (making it difficult to place the jumpers and maybe this caused it not to work? idk) >> That's strange. The manual was also not well written in this regard as it would been a lot easier to first install the drivers on the board and only after that installing the board into the printer housing. I understand this is more of a "budget" printer but it's definitely not even close to being remotely beginner friendly. But more advanced users who could troubleshoot through all of this likely would be willing/interested in getting a more expensive but advanced kit anyway. >> Yes and yes :-)

@Thomas S Sorry, I have a hard time following your explanations. The sensor is there to stop the hotend before it goes into the printing bed. This should be the case for homing as well as for auto bed leveling. If it already crashes into the bed then your sensor is mounted to high. The Z-offset is applied when actually printing something but the same principle applies. Too much Z-offset will let the hotend crash into the bed while too low will let the printer print in the air above the bed.

@Johannes Helm I hear you :-). That's why i'm still using the default bilinear bed leveling with just an increased number of points (7x5 pts).

Not sure what you mean with "adjusted the Z arm". After the G28 command, the hotend does not need to be paper-size close to the bed unless your Z-offset is really 0.0. Usually, you will have some Z-Offset because the sensor is not at the same level as the hotend - and it also does not need to be. As Jay M mentioned and if you followed the manual, a good distance is about 1mm between hotend and the sensor. If you don't have a Z-Offset set (=0), the difference between the hotend and the bed after homing will become your Z-Offset. Also note that with a Z-Offset set, the homing will end up at Z=0 without the Z-Offset applied. That means the Z-value you read with M114 or on the LCD will be the actual Z-Offset. This is true when the bed leveling is of (M420 S0) - otherwise it will differ by whatever the bed level correction at the homing position is. Example procedure: - starting with Z-Offset=0, do a homing with G28, disable bed leveling with M420 S0 => Z-value should show 0.0 - do the folded paper test and lower the hotend until grabbed - read the Z-value with M114 or directly on the LCD, let's say it is -1.5 - save the Z-offset with flipped sign using M206 Z1.5 - now save setttings to flash with M500 - M503 will show current settings including Z-Offset After that you can use G29 for the bed leveling and don't forget to save that as well with M500. You also don't need to worry about the previous bed level disable (M420 S0). It will be re-enabled after G28 and there is also an explict enable in the CURA start g-code for the printer (along with the G28 homing).

The initial Z-Offset configuration is described in Stage 9 - Steps 22/23 of the build manual. The final adjustment is described in Stage 10 - Step 11. In short, the M206 Z command will set the Z-Offset where offset is your reading but with changed sign. If the hotend is too low, you would decrease the Z-Offset value.

M114 command will just show the current position so it might be misleading depending at which point you use the command - and what Z-offset is already in use. I would recommend to use M503 to check for the configured Z-Offset. Also make sure to input decimal points as '.' not ','. In case you need the original start and end g-code: ;----------AXIS Start G-code---------- T0 M104 S{material_print_temperature_layer_0} ;Set hotend temp. M140 S{material_bed_temperature_layer_0} ;Set bed temp. M109 S{material_print_temperature_layer_0} ;Wait for hotend M190 S{material_bed_temperature_layer_0} ;Wait for heated bed G28 ;Home M420 S1 Z4 ;Activate auto-level and set Z-fade. G1 F8000 X10 Y3 ;Move hotend to bottom left corner G1 Z{layer_height_0} ;Set to first layer height M82 ;absolute extrusion mode G92 E0 ;Zero Extruder G1 F1500 X290 E12 ;Prime Hotend ; ----------AXIS start G-code---------- ;----------AXIS End G-code---------- G91 ;Relative postioning G1 Z5 F600 ;Raise Hotend G90 ;Absolute postioning G28 X Y ;Home and present print M104 S0 ;Hotend off M140 S0 ;Bed off M84 ;Motors off ; ----------AXIS end G-code----------

Congratulations. You must have a different printer..well maybe it is a different printer with your modifications ;-). Someone should release a AXIS v2 with all those improvements. Something like the i3 bear upgrade :D.

@Johannes Helm if you want to manually probe those points you can use the G29 P4 T Rn Xx Yy command. It will not directly go to a mesh point with x/y specified, if no x/y is specified it will use the closest mesh point from the current position. R defines how often this is repeated (for multiple mesh points). I assume this could be combined with G42 Ii Jj command which actually goes to a mesh point directly i/j are the row/column indices. If you really want to manually edit a mesh point there is also the M421 command. With M421 you can also invalidate a mesh point so it might be possible to delete and then (auto) fill again with G29 P3 command. Note that I have never used UBL so this is just what I get from the documentation.