I have been having some issues after mounting maslow to a concrete floor.
Turns out the floor is not only sloped, but also curved!
This makes auto calibration not work very well at all, (I’m seeing fitness values around 0.3 in many cases.
I fixed this by measuring the mounting points myself and editing the config.
This made things dramatically better.
However, I am thinking its likely that a lot of position error can come from relative slope between the mounts and the work piece (which I have), and / or from curvature of the work piece(which I have)
It should be possible to account for this in software if it isnt alrerady accounted for.
1 Like
Michael DeLay wrote:
However, I am thinking its likely that a lot of position error can come from relative slope between the mounts and the work piece (which I have), and / or from curvature of the work piece(which I have)
It should be possible to account for this in software if it isnt alrerady accounted for.
for the mounts to the workpiece, that’s what the Zoffset values are for (we
don’t get enough information in the calibration to be able to compute these, but
it may be that a more sophisticated estimating algorithm could look and see if a
different Zoffset improves the estimates)
curvature of the workpiece is even harder, and I would say that the answer is to
get a stiffer wasteboard to avoid it 
David Lang
1 Like
Good to know that I can calibrate out relative slope by measuring z positions. Fairly easy to do to ±0.5mm accuracy using a self levelling line laser.
I’m juggling the idea of machining a custom shim set using the Maslow for the bottom of my waste board. It works OK already but definitely is a little suspect near the edges.
1 Like