forum.cncdrive.com

Hello guys,

I experience a (supposed) strange behavior while executing G31 :

*** Supposing no inertia problem : G31 is executed a reduced speed anyway ! It is supposed that the speed choosen for G31 is compatible with instant stop *****

The feature i ask here should be optional , for sure,

I would like that the UC-300ETH (and other intelligent BB from CNCdrive) instantly stop sending steps when the probe is hit.

Second : the Fxx percent correction should not apply within G31. I broke a series of 0.8mm PCB drill bits until i understood that G31 Zxx Fyy is influenced by the 232 field.

For now i made the correction in my plugin, but...

Seb

Instantly stopping the axis can cause lost steps for most of the machines is why the probing does not stop with instant stop, but with decceleration.
However the probed position is still precise, because the probed position is registered when the probe is hit and not when the axis stops.

If the G31 would not depends on the override then how could you adjust it's feedrate without rewritting the F code in your g-code file?
I mean then there is no way of fine adjusting the feedrate which is not too good in my opinion.

Sebmay

What is the steps per unit each of your machines axis

And what is the acceleration setting of each of your machines axis (X,y,z,a,b,c)?

Hello,

Sorry for the late feedback : i am not at the cnc place, but i can tell that the regular "fast" feed for Z is 500mm/min, and because the head is very heavy, the acceleration is poor, I.E. it takes time to stop. But i always do a G31 Zxx F100 as fast probe, then go back up relatively 2mm, then G31 Zxx F30.

But the logic is this : when i use the percent acceleration for feed, this is mostly because i missed to set up a reasonnable fast speed when generating Gcode. I compensate by setting the feed multiplicator. Or because i feel like the result of cuttings are so good that i may accelerate what was at first guessed.

I never do the same work twice : there is no option to correct for the next time.

But in my M6 (generated dynamically with each job), i cannot anticipate the coming multiplication of the feed rate.
For now, i solved the problem in my M6 generator by adding a push of the multiplicator coefficient when entering the M6, setting 100%, doing what has to be done, then pull back the coefficient.

I took note of the fact that the returned value is the one sensed when the detector trigger.

Anyway, between 100mm/min and 30mm/min, i get almost 1/2mm of difference. It's huge when carving a PCB. That's why my question is.

Thanks for your attention.
I will send my accell/max speed value later tomorow.
Sebastien