GearHeads Forum

Thank you very much Art

I always do an 'air cut' to estimate the maximum effektive feedrate achieveable for the job, bacause you explained and recommended this earlier here in the forum.
So this is not causing the effect. Regarding the foil - and if I think about it - you are of course right: it either melts somwhat on the surface or then (likely suddenly) cuts through. Not linear but binary, so to say. So I do what you said and try slower but also with less power and/or set the minimum in the settings to something above zero.
Will be back if I find out some consistent effects, maybe others have observed the same effect.

Thank you

Best regards
Roman

Hi Roman:

In essence, the dist correction tries to maintain a solid fire power, the way it does this is worth noting though, as it may affect what your doing.

First, the speed is taken for the upcoming ms of motion. This is computed as a percentage of the FULL SPEED request, that is the ultimate feedrate
in effect. Then the power is adjusted to be 0-100% based on the percent of speed to be done.

This is important for this reason.. say you set 15000mm/sec as the feedrate in a curvy program, but it never reaches 15000, say it reaches 5000
at max speed due to the curves and corners. The power will be computed off of a speed of 15000, so the power will be much lower than
one expects. At full speed it would be 1/3rd full power. (I know this doesn't explain your trouble but it may be related as Ill explain).

Corrections on a laser aren't linear typically, each laser differs in just how linear they are, but Auggie expects it to be for this calculation.
If the actual max feedrate isn't close to the set feedrate max, then this nonlinearity is accentuated, made worse by the assumptions.

For example, if a program is running at 30mm/sec in a curve and then at 100mm/s in a line, with a proper max feedrate of 100, then the slower speed will 1/3rd of the max power seen at the 100mm speed. But..if the feedrate is set to 1000mm/sec, then the difference between the 100mm segment and the 30mm/s segment will be very small. (Because as a function of percentage the difference of 70mm/s is only about 7 percent of the total speed curve.

When I'm being picky, I run the program in the air to see just how fast the program gets at max speed. I then set feedrate to there, or lower, and the lower the set feedrate is, the better the speed correction works.

This is nonintuitive, but the math problem is one too many unknowns for the speed/power equation, so the feedrate setting is assumed to be
the actual max speed , so if its off, the distance correction gets off by a percentage.

Film cutting in particular can be very sensitive to power levels, where wood or plastic may not well see a 5% variation,
films of many types will react very non linearly to power. Try varying the feedrate lower ( and power of course ) to see if this smooths out the differences. The usual problem with distance correction is a blanked corner, this is due to the next ms of speed creating a power level way too
low to cut. (that nonlinear problem I mentioned.), In the settings you can select a minimum for the power calc to aid in that.

Art

Hello Art

The Eroor 7871 is solved since a longer time. I tried so many things that I do not remember....

I just apendend to this old conversation the new question regarding the speed/laser power corelation.
Yes, the "DistanceCorr" is on. Because I thoght this is one of the 'essences' of Auggie ... :-)

Best
Roman

Hi Roman:

Error 7871 is a macro C string execution error. Did you recently modify a script or macro?
Its a generic error, all it really tells me is the macro system executed a string in the C interpreter,
and then after it was sent for execution, something crashed. Id check my macros.. when does it happen?
Does it occur at startup, or when you try to run a file?

>>Courved, slower sections: no cut, just little burn.

The selection of "DistanceCorr" on the main screen turns on distance correction, which adjusts power to
be equal to the actual motion speed, its meant to remove as much as possible the slowdown in curves and corners
to stop them from burning. Is it on? I think there are a couepl settings as I recall that set a minimum amount
of motion time before firing.

I haven't run Auggie in a while, but I started mine up just now tro check it all still works here. I really must
get back to using my lasers..

Art

Hello Art, me again...

The above problems where meanwhile all solved (new, faster Computer probably solved it. Fingers crossed).

Somehting else I observed now by chance:
I cutted a thin masking foil. On straigt, faster moves the foil was cut just like it should. Courved, slower sections: no cut, just little burn. If I raise the power or lower the spead the curved lines ar cut just right, but the straight ones are overdone, burned.

Means that the linear relationship between actual speed and Auggie-regulated laser power is not good here. Is there a way to influence this?

Best regards
Roman

Roman:

Sorry for the delay in answering.

  I think of two things, acceleration, and jerk setting. They are codependant and may be the likely
source of the problem. Try with a lower jerk, or a higher jerk. Then try with higher or lower accleration.
Getting accel and jerk right is Auggie can be hard as few understand the real relationship there,
try a few experiments at 1200 velocity with high or low jerk to see if it makes a difference.'

Art

Hello Art
After 500mm/min I tried 600, 700, 800, 900, 1000. At 1000 it crashes. Repeatedly. Went back to 900 and it does not crash. Repeatedly.
There is nothing else running, ther is enough CPU and enough Memory.

No clue.....

Best regards
Roman

Hello Art
I left Lookahead at 300 and also everything else as it was.
So the very first thing I did now is run the file at only 500 mm/sec instead of 1000 mm/sec. Works perfect, no fault. The Queue flickers at 299/300/301.

So it must be something speed dependent I assume. Of course I would like to be able to go back to higher speeds.

Roman 

Roman:

  Yes, you need the spindlelib for actual laser, but its sometimes usefull to
run with the system only thinking the laser is on if you can arrange it. I usually go
about it by eliminating what I can.

  On the other hand, the profile contains pretty much everything that can crash a system
other than its scripts.. so replacing the profile will probably do it.

Art

Hello Art
The first thing I did is that I went down to lookahead 100 (normally running at 300). No effect.
Yes, I have backups of everything.
Will now
- Reload Profile from Backup
- Turn off Spindle Lib  (You means SpindleLib-Laser under GCode Scripts? Isn't this essential for Laser usage with Auggie?
- Try to run it slower, means 500 instead of 1000mm/sec

Will report back any observations.

Thanks for your help

Roman