[natural]

Hello All,
I've done my best searching for a solution to this but I'm at about the end of my knowledge (and wits!). Machine is a custom built 3 axis router, 49x97 cutting area using steppers. I've got an older Flashcut signal generator going to Gecko 201 10u microstepping drives, and I'm running Flashcut 2.1 software.
As the title says, my issue is that it will always pause when any axis switches directions. Linear cutting this doesn't bother me so much, but cutting a circle for example, it pauses every 90 degrees. It is a momentary pause, so I'm not sure I can quantify the actual duration. There does not seem to be any ramping up to or after the pause, it is extremely abrupt especially at higher feedrates. I have changed my direction change delay down to 0.001 seconds or 0 seconds. I have continuous contouring tolerance set to 15 ipm. Starting feedrates are 5 ipm on all axis. I have my acceleration set relatively low as the gantry and z carriage are on the heavy side.
Not sure what else to check! I have also tried many different g-codes and tried importing a simple circle DXF for flashcut to generate the code. Still pauses!

Any help would be much appreciated! Last thought: Is my logic flawed? Am I looking in the wrong place? I know the stepper motor signals need a tiny delay to allow the direction signal to reverse, but I understand the units of that time delay to be imperceptibly minuscule. In any case, is this pause just something that's inherent with my stepper/driver setup? Or is it possible that my direction lag / setup signals are too short and I'm actually losing a step on these pauses (because the motor isn't being given enough time to change directions)?

I apologize for the novel! Thanks for your patience if you're read this far.

Cheers,
natural

[natural]

Ran some more tests this morning. Set my start feeds & acceleration really low and found out it IS ramping to/from these pauses. Does this mean my the "continuous contouring" is not working?

Thanks again,
natural

[natural]

Well, I'm not sure anyone cares, but I thought I'd reply with some new insight on the matter.

Turns out, I'm an idiot. As I did not use any kind of mechanical anti-backlash devices on the x and y axes, I was relying on the control software to compensate for the some 0.002" backlash that my ballscrews have. This isn't bad per se, but it is responsible for the original problem I posted. When backlash compensation is turned on, the machine has to pause every time an axis changes directions so that it can add that backlash back to that axis without deviating from the programmed path. If it did not pause, the actual cut path would deviate from the programmed cut path by an amount relative to the backlash on that axis.

Hopefully this helps someone in the future!

-natural

[Fastest1]

This thread is very old and I am not familiar with FC yet however generally a backlash compensation will have a time function too. There are a few ways to improve those issues. Look ahead helps and gives the pc time to plan the trajectory.