#Probe Accuracy varying standard deviation

11 messages · Page 1 of 1 (latest)

rigid vault
#

Im currently working on my 2.4 Matterhackers Rev D kit, I have had some issues with my probe accuracy's sample tolerance failing a lot during QGL's and Bed meshes. I started kinda high level like Probe_ calibrate and Z endstop calibrate, this didnt resolve the issue so when I tried to figure out where it was messing up I ran 2 different probe accuracy's. The first had a standard devation of 0.002610 but the second was 0.006847. These were conducted within 30 seconds of each other. My Klippy log is attached

📎 klippy_3.log
rustic mantle
#

You probably should increase your samples_tolerance. 0.006mm is very small (6 micrometers!), and less than the accuracy of the z motor drive.

#

I run 0.02 and I'm perfectly happy with that. It's still an order of magnitude smaller than typical layer height and more precise than I'm ever able to adjust Z offset.

round vortex
#

I run 0.01 on my 350 2.4. Default is 0.0075. Roughly every other corner will retry due to high standard deviation once. Never failed a QGL unless something mechanical was wrong.

rustic mantle
#

It doesn't fail, it just has to hunt one or two more laps extra to randomly hit accuracy that isn't really there.

summer raft
rigid vault
#

Yeah I have! Thank you so much for the help!

summer raft
rigid vault
#

I am out for today and was gonna start working on a fix later tonight

summer raft
rigid vault
#

So as the update, I changed my new samples_tolerance up to 0.01 and have been trying to run QGL
It is now saying Probed points range is increasing, I went back to basics of making sure my belt tensioning is good, it's not perfect but should be plenty alright