Always supply DEFAULT_[XYZ]JERK in Delta printer configuration examples.

[Knifa]

Description

Fixes the user facing side of the problem for #14976 by always providing jerk values, even when using JD. I'm not really sure if there's a better core fix to be done here as I'm not too sure of the intricacies.

The file configuration_store.cpp sets all the jerk values to 0 if nothing is provided --- maybe this should just be above 0 by default?

Benefits

Users with delta printers won't experience mad stuttering when using JD --- see #14976 for more details.

Related Issues

Makes #14976 better for users but maybe doesn't solve the core issue.

[AnHardt]

Naah.

To activate JUNCTION_DEVIATION on DELTAs remove/comment

https://github.com/MarlinFirmware/Marlin/blob/bugfix-2.0.x/Marlin/src/inc/Conditionals_post.h#L56

Then it will really be activated and you don*t need the jerks.

But be warned. Most developers here believe in JUNCTION_DEVIATION is theoretically incompatible with not Cartesian kinematics.

[Knifa]

Naah.

To activate JUNCTION_DEVIATION on DELTAs remove/comment

https://github.com/MarlinFirmware/Marlin/blob/bugfix-2.0.x/Marlin/src/inc/Conditionals_post.h#L56

Then it will really be activated and you don*t need the jerks.

But be warned. Most developers here believe in JUNCTION_DEVIATION is theoretically incompatible with not Cartesian kinematics.

That makes it sound like you don't agree? 😄 Do you have some more info on this at all?

At the very least it probably still makes sense to provide sane defaults right? Otherwise you can end up with this weird stuttering problem without realising it.

[AnHardt]

It's a bit complicated.
I do believe the theory for junction deviation does not match delta printers. But i also believe in the reports you are able to print on a delta with junction deviation.

In theory all axes have to be perpendicular to each other. That's for sure not true for the deltas stepper axes. But for me this is not necessarily meant geometrically, but meant the axes are independent from each other. This is not completely true. Even if you claim the main masses are in the motors rotors, carriages and diagonal rods, the effector is clearly not moving independent. This could be compensated by smaller junction deviation radiuses - moving slower - for the worst case.

The second point is the non-linearity of the effectors movement. Worst case here is when a diagonal rod is becoming horizontal a step of that stepper causes nearly no motion of the effector. But as long the delta diameter is much smaller than the diagonal rods length the non-linearity is not that bad - about +- 50%, i guess.

Summary:
Junction deviations theory does not match delta printers, but printing will not necessary end in a complete disaster, as long not pushed to the extremes.
On the other side, also Marlins jerk speeds have their theoretical problems.

In my opinion every user should be able to experiment with both kinds of systems to find his/her favorite and the fastest/smoothest printing. Sadly most users are more limited in that than Marlin. Even experts like you and me are often turning the wrong wheels.

[Knifa]

This was a great reply --- thanks for being so informative!

It maybe sounds like it might be worth putting a warning somewhere instead? Like I said, this is the first time I've learned delta and JD is a contentious combination. Give everyone the tools to shoot themselves in the foot, but not necessarily encourage it? 😄

[thinkyhead]

I agree that JD is not compatible with these kinematics, so we will add a warning to prevent enabling it with DELTA unless you're into experimentation.