Z axis moves same direction when jogging - Marlin RC7 Upgrad

[ryaneb85]

Hi all,

So i just updated to Marlin RC7 as I installed a BLTouch and want to get mesh bed leveling working correctly. Everything generally seems to work ok, but when I jog the Z axis, it appears to work correctly in the +ve direction (bed moves down), but when I jog it in the -ve direction, it goes the wrong way (also moves down), but seems to move slower and makes a funny noise. I can't for the life of me find which setting I may have missed, so any suggestions?

Ryan

[ryaneb85]

It looks like it moves as expected when jogging in small increments such as -1mm, but if i try -10mm, the bed drops and I get the funny noise. Thought it might be an acceleration issue, but the settings are the same as previous.

[ednisley]

seems to move slower and makes a funny noise

Check the Z-axis acceleration & max velocity settings: most likely, those innocent jogs now require more oomph than the motor can deliver.

Either the default values changed or you've got a typo in there that could happen to anyone...

[ryaneb85]

This is what I have. I think it's as it should be:

[ryaneb85]

Ed, do you know if there are any settings I need to update from default in the configuration_adv.h tab? I haven't changed anything in there as from what I know it should all be ok as is.

[ednisley]

Because I can never remember that stuff, I wrote up what I knew when I installed the fine-tooth filament drive gear:
https://softsolder.com/2016/04/14/maker ... 1-1-0-rc5/

Those files change their format & content without warning, but perhaps you can run a diff to figure out what's new & different...

[ryaneb85]

It looks like the max acceleration for the Z should be 30, whereas my previous version was 100 (and working fine), so I dropped it down and all is well.

Thanks for your help. Now I just have to try and get this BLtouch configured... but that's a job for tomorrow morning!

[ednisley]

my previous version was 100 (and working fine)

There are mysterious linkages between the various constants, so I wouldn't be surprised to find something else has changed every so slightly.

A factor of three sounds about right for a "makes a difference" acceleration change. I don't know any algorithmic way to figure those numbers out, because all the forces depend on things you not only can't measure, but can't even guesstimate their magnitude.

Dabbing a bit of lube here & there on the Z-axis slides & leadscrew may be in order, too, if only because we all tend to forget that task. Then you can increase the acceleration by factors of two until it stalls out on longer moves, back off the last factor, and run with it.