Z-Height Calibration Area  

Eminent Member


Today while observing the first print on my freshly cleaned PEI Powdercoated Bed I had a moment of enlightenment regarding the 9 (49?) Point Z-Height Calibration procedure that is happening at the start of each print that got me wondering - Why the hell does the Probe calibrate itself against the entire 250mm x 250mm of the Bed if the contact-making first layer of the to be 3D Printed model is only 45mm x 45mm?

Wouldn't it make more sense to reduce the Z-Height Calibration Area to the size of the area of the Bed contact-making first layer plus a 5mm border? So 45mm x 45mm would be 5+45+5mm x 5+45+5mm and obviously also taking the Probe offset into consideration.

After all, the Printer SHOULD know where the first layer will end up and what the area it is making contact with will be, no?
Heck... We could even go as far as to just choose a Z-Calibration Spacing ( say 15mm ) and the Printer would only calibrate itself in those areas where first Layer Bed contact would happen instead of the whole Bed.

Posted : 24/03/2020 3:41 pm
Illustrious Member

If you're doing 7x7 mesh bed leveling with 5 probes at each point, the entire process takes something like 45 seconds. If you cut this in half, you're saving all of 23 seconds on a multi-hour print. The biggest challenge is that leveling is currently triggered by a single gcode command. To make it behave differently would require firmware changes which are likely to introduce unexpected side effects and you'd have to update the slicer to insert the non-standard gcode parameters. This would also not benefit anybody not using a PrusaSlicer-Prusa printer pair (e.g. using Cura as a slicer).

It's an interesting idea, but it's not likely to make a big difference to overall print times. I wouldn't expect this to be a top priority.

My notes and disclaimers on 3D printing and miscellaneous other tech projects
He is intelligent, but not experienced. His pattern indicates two dimensional thinking. -- Spock in Star Trek: The Wrath of Khan...
Posted : 24/03/2020 4:03 pm
Estimable Member

I just happened to stumble on a developers thread on this topic yesterday. A lot of thought and testing went into 7x7 MBL.

They investigated various grids (4x4, 4x6,....). They needed to avoid the bed magnets and standoffs.

Check here for github thread


Quality is the Journey, not the Destination
MK3S, Nylock mod, Modified Y-axis bearing mounting
Posted : 24/03/2020 4:28 pm
Eminent Member


Besides saving 23sec on a multi-hour long print I'm actually more annoyed with the thought of measuring my entire house to place a vase in my kitchen using a ruler made of 1m increments. The more you think about it the less of sense it makes ( to me ) when the obvious choice would be to only measure the kitchen and using an appropriate 1cm ruler.

As it stands now, small parts like the one in my example only benefit from 1 of 9 ( ~11% ) or 5-9 of 49 ( ~10-18% ) probe points depending on the shape of the first layer and grid size I choose which, at least in my book, is kinda terrible xD

But thanks to @kennd s post it looks like there are issues with my suggestion, like the magnets and standoffs, I wasn't aware that would need consideration.

Posted : 24/03/2020 9:37 pm

Please Login or Register