Mesh Bed Compensation doesnt work well
hopefully, i'll find some answers here.
i've build a large Delta with magnet Bearings and a Chimera Hotend (aircooled) the printing Volume is about 315mm high with a printing diameter of about 260mm (the bed) but i can increase it to reach the 320mm dia.
the Effector and all other parts are created with CAD, so i could measure all the offsets in CAD and just copy them to the config file. (i have an Y offset of about -25mm between nozzle and the BL Touch Probe, and if the pin is deployed, i have about 1.30mm between the tip of the nozzle and the pin of the probe (Z offset)
now i'll try to explain my problem: i started with the G32 (S4,6,7,8 and 9 (yes, all of them, because i am confused about my problem.... )) and repeat it for about 5 to 10 times.
The errors are at 0.05, so its quite good. after that, i saved the values with the M500 (M501 at the end of my config file).
now i started to print and realized, that the diff between the bed and the nozzle is quite good in the middle of the bed, but at the y Tower, the distance is to short, so the filament is stocking inside the nozzle. at the other towes, the distance is quite good. so it means, the bed is not flat enought.
so i started to use the mesh bed compensation just to so if its better or not. i get really good results (see the pictures of the hightmap below)
BUT: it doesnt work. as u see at the pictures, there arent any big gaps or hills at the hightmap.
so i tried to slow down the probing speed, use a R1 command to wait for the reset of the probe before each point and started again. so again, the map was very clear and without big surprises.
i did it again and again, and everytime i start to print, thers the same problem. at the middle it is quite good, near to the X and z tower also, but for example 5cm away from the middle to the direction of the Y tower, the nozzle is to close to the bed.
at every gcode file, i use the G29 S1 AFTER HOMING, so it should work, but it doesnt.
here u have some pictures from the hightmap and also my Config.g the overrides and the bed.g.
Your issue is almost certainly caused by the effector tilt varying with XY position, causing the relative height of the nozzle the the bltouch to vary. Mount a spirit level on the effector to check.
It's possible to compensate for effector tilt in the bed.g file, but it's tedious because you need to measure the trigger height at each probe point. It's better to use a nozzle-contact probe, such as the Smart Effector or Precision Piezo.
Thanks for this little hint. i just figured out, that the Effektor is tilted depending on the position at the bed.
so this could be a reason. now i have an other question. i dont really want to recreate the effector and use an other Z probe (a friend of mine has the same printer with just a few differents in rod lenght and hes just using one instead of two hotends), and his printer works quite well.
so what i need to know is, how i can decrease the tilt. in my opinion, the tilt could depend on different rod lenght or some differents at the cadridge or the effektor itself.
i know, that all of the rods have the same lenght (diff: less than 0.1mm)
do you have any ideas to decrease the tilt? i just can measure the geometry again and again but i thought it was as good as possible.
after a few days of testing, i figgured out, that the problem wasnt caused on one of these mechanical reasons.
the problem was quite obviously, but sometimes you dont see the forest for the trees...
I use the TWE 15 Igus system, and the carriages are adjustable. Because of the "Big" Distance betreen the joints (65mm) the whole effektor was tilted at the Positions in front of the towers. i adjust the bearings of the Carriages and calibrate again.
the problem was gone. now, without any Meshbed compensation, there arent any Problems with the first layer.
im Very happy
I'm glad you solved it. In what way did you adjust the bearings of the carriages?
there are two screws for adjusting the carriages. i adjusted both of them until the carriages are still able to move up und down (call it the Z axis) but not able to tilt around both of the other axis.