Hypothetical question regarding acceleration and latency

3DPMicro

Cartesian, Duet 2 Ethernet (or Duet 3), X axis - external driver and servo with S curve acceleration off the Expansion Breakout Board, Y axis - stepper w/ trapezoidal accel. driven by onboard driver.

Could this scenario cause print issues because of mismatched motion profiles or latency between the two?

o_lampe

@3dpmicro
IMHO, everytime one motor runs slower than jerk setting, the mismatching ramp-up/down profiles come into play.
~~Eg. at the turning point of circles~~
//edit It would probably show most, when both motors run below jerk speed?

3DPMicro

@o_lampe thanks for the input. There's also potentially a speed/distance range where the move is in accel/decel for a significant amount of time that might cause issues. Outside of theory I may have to just try and see.

o_lampe

@3dpmicro
OTOH we are allowed to use different accel and jerk settings for X and Y axes. That alone would cause non-conform motion profiles.
But I'm not aware of any complaints. Mostly because we set them identical. (and G1 always picks the slowest setting?)

dc42

@o_lampe said in Hypothetical question regarding acceleration and latency:

OTOH we are allowed to use different accel and jerk settings for X and Y axes. That alone would cause non-conform motion profiles.

Not so. The actual acceleration and jerk used on X and Y will be in proportion to the movement distances for that move.

o_lampe

@dc42
That's clever! So we don't have issues with linear acceleration, but external drivers with 'S' curved accel would be a problem?
At least when they are mixed with linear acceleration drivers

dc42

@o_lampe said in Hypothetical question regarding acceleration and latency:

@dc42
That's clever! So we don't have issues with linear acceleration, but external drivers with 'S' curved accel would be a problem?
At least when they are mixed with linear acceleration drivers

External drivers that have a latency between receiving step pulses and moving to the requested position could cause mismatches in the XY tracking. If both X and U used those drivers and they had equal and constant latency then X and Y would track properly.