Hello Guest it is June 12, 2024, 10:04:12 PM

Author Topic: reset home with G code  (Read 38755 times)

0 Members and 1 Guest are viewing this topic.

vmax549

*
Re: reset home with G code
« Reply #30 on: December 23, 2008, 11:20:29 AM »
HUM it passed right throught the macro here and continued on with the code???

ANother test ,(;-) on a gcode file you know it happens try reducing the Fspeed to 1/2 normal (fro) and see if it still occurs.  With a file that rotates the cutter CW & CCW it could loose steps and LOOK like it gained steps(+3-10) in the end.

IF the cutter stepper(A) is undersized it may have trouble with low torque and fast moves. IF so a stronger steppper motor OR gear it 2:1 for more torque IF you can get by with  lower speed on rotation.

ALSO check your stepper drive motr settings if it is set lower than max amps for the motor the motor will have less torque to work with.

Just a thought, (;-) TP
« Last Edit: December 23, 2008, 11:22:11 AM by vmax549 »
Re: reset home with G code
« Reply #31 on: January 02, 2009, 05:10:13 AM »
I have been working on the machine today trying different things.

The macro continues to pause ater the a axis homes, I will read up tonight on a cycle start command to add to the macro.

Most of my code runs CCW and at the completion of one file the A was +20 deg out and another file it as +12deg, this was check by MDI'ing a A0 command then jogging backwards the axis until the homing limit is hit . I worked on reducing my step/deg number and found that 5.535/deg had the filet hat was + 20 out at the finish now finishing at 0 however the other gcode file is -7 deg when zeroed with these settings. These test were ran with the blade removed so no load on the stepper. I did another file wth 5 different shapes like a square, circle, pentagon and eclipse of varying sizes and at the completion the A axis was ok ie no lost steps. I tried running the problem files with reduce feed rates and vel and Acc settings on the A axis with similar gains of posttion as above. BTW I have I and J codes in these files. Meaning I have drawn the shapes using arcs and fillets between straight lines. I have done this as the speed of the machine is not compromised as the A axis runs in Exact stop mode even though I have a G64 in the code. This may have to do with the way the A axis code is outputed.

I may be wrong here but I thnk the A code is developed like this
A90
x100 y100


If it was done like this it may work better??
A90 X100 y100

My next process is to set the chains for the parts to run in the CW diection and see how the position ends up.

Thanks for the help and Happy New Year

vmax549

*
Re: reset home with G code
« Reply #32 on: January 02, 2009, 11:32:43 AM »
Just a thought but,  A90 x100 Y100 will give you an interpolated move that has the knife turning the entire length of the straight xy move that should create a drag on the knife as it is NOT in angular position at the start.

I would think the angular rotation should occur at the radius points where the knife has to make the radius transition. AND the knife should be in position at the end of the arc to make the next straight move. Or at a sharp coner

A90 X100 Y100 R.25 ( knife follows the arc)

X150 Y100

A120 (sharp angle no radius)

Y150

Just a thought, (;-) TP
Re: reset home with G code
« Reply #33 on: January 05, 2009, 09:49:43 PM »
Ok more testing today.

Happy New year to you all.

I still have the code pausing with the  M1099 A home command. I tried a dobutton (0) after this however it did not excute the A home. I tried adding a Code "G4 P5" after the A home and before the cycle start. Still it did not do the A home:)

I have run some simple 5 shape tests with F10000 and it is OK with refernce to A axis positionng.

I tried a 20 part cut in the CW direction and the error is exactly the same as the CCW being +15 degs out.

I switched my A axis stepper to Unipolar and change the current set resistor. Reduce A Axis motor tuning to V5000 A500 and run a test at F2500. The problem still exisits as if it was with high V and A setting and feedrates of F10000.

I am not sure where to go from here??

Offline Graham Waterworth

*
  • *
  •  2,691 2,691
  • Yorkshire Dales, England
    • View Profile
Re: reset home with G code
« Reply #34 on: January 06, 2009, 06:57:09 PM »
I think your problem is an accumulation of part steps on complex shapes, on simple shapes you do not have the problem as the a axis is only moving a few times.

I would put a belt drive on the a axis that changes the steps per rev to an even number, so if you ran at 200 steps per rev the 360 degrees / 200 steps = 1.8:1 ratio.

You could then microstep and the step error would be reduced by the fixed ratio drive.

Just my thoughts.

Graham
Without engineers the world stops
Re: reset home with G code
« Reply #35 on: January 06, 2009, 09:54:02 PM »
Thanks Graham,

I like our idea. Wha are your thoughts on using a 0.9 deg stepper so that the step count/deg would be 4000/360 =11.1111111

Or are you thinking that a whole  number of steps/deg would reduce this error.

If this is the case do you think that a 3.6:1 would be better still with a higher step count.

I'm tossing up the amount of work required to do the reduction drive compared to sourcing a 0.9 deg stepper.

Thanks again

Offline Graham Waterworth

*
  • *
  •  2,691 2,691
  • Yorkshire Dales, England
    • View Profile
Re: reset home with G code
« Reply #36 on: January 08, 2009, 05:29:19 PM »
In my opinion there is no better drive than a whole number of steps per rev, no loss of part steps and no rounding errors.

The best machines out in the industrial world use encoders that have multiples of 360 pulses per rev. They don't use these for fun :)

Graham
Without engineers the world stops
Re: reset home with G code
« Reply #37 on: February 21, 2009, 06:03:29 AM »
would it not be g 91 g 28 zo xo yo ?