Hello Guest it is June 17, 2019, 11:28:51 AM

Author Topic: Jerking with short segment G-code  (Read 4009 times)

0 Members and 1 Guest are viewing this topic.

Offline spunk

*
  •  56 56
    • View Profile
Jerking with short segment G-code
« on: January 20, 2010, 05:17:30 AM »
Dear all

Have read all the topics about this and haven't come across the solution for my app.

Let me give you an idea about what i'm willing to accomplish:

I'm about to apply a coating on a round surface using the tangential option in Mach3.
I output the G-code using VX cadcam. because this to-be-coated surface has complex upper contours it appears that VX cannot output G2 or G3 segments (allthough it all looks like its round)
The resulting code is cut in to about 1mm pieces.

Problems so far

Without "tangential" turned on I get a rather smooth run apart from some bumps here and there...
When I turn on "Tangential" the whole Machine starts to shake like it's dancing the samba!

checked speeds and accelerations, all set to a high enough value... according to all the topics i read everything should be OK.

any ideas?

i'll add the example code for reference...

The XML will be for later on when i get the chance to download it.
Re: Jerking with short segment G-code
« Reply #1 on: January 20, 2010, 06:06:11 AM »
Hi, I'm not sure about tangential mode but have you tried CV mode, With G64.

Jeff

Offline spunk

*
  •  56 56
    • View Profile
Re: Jerking with short segment G-code
« Reply #2 on: January 22, 2010, 04:03:19 AM »
so, turn off tangential and just put G64 in the header?

Offline spunk

*
  •  56 56
    • View Profile
Re: Jerking with short segment G-code
« Reply #3 on: January 22, 2010, 06:05:09 AM »
tried the G64 but nothing...
Then tried to program it by hand... when only 3 axis are driven there's no problem, when the fourth (A) kicks in the jerking starts...

anyone?

Offline stirling

*
  • *
  •  2,188 2,188
  • UK
    • View Profile
    • www.razordance.co.uk
Re: Jerking with short segment G-code
« Reply #4 on: January 22, 2010, 12:43:57 PM »
am I missing it? looking at your code in #1 there are no A axis moves, just x,y,z. Also can you explain what your a axis is supposed to be doing with regard to your coating requirement - and why tangential? - just trying to get a grip of the problem.

Cheers

Ian

Offline Hood

*
  •  25,843 25,843
  • Carnoustie, Scotland
    • View Profile
Re: Jerking with short segment G-code
« Reply #5 on: January 22, 2010, 01:40:28 PM »
Afraid I have never used Tangential Knife control but I have read a few posts regards it, if you do a search there may be some help. Even if you dont find the answers there may be names of members that use it that you could PM to see if they could help.
Hood

Offline spunk

*
  •  56 56
    • View Profile
Re: Jerking with short segment G-code
« Reply #6 on: February 02, 2010, 06:11:17 AM »
found the "problem"

if you make a short segmented g-code with G1's the knife will get in the right angle BEFORE the movement takes place.
thus jerking it's way through the code.

made some nice long G2 and G3's and all goes perfect!

cheers
Re: Jerking with short segment G-code
« Reply #7 on: June 23, 2011, 10:28:41 PM »
Greetings All,

Sorry to revisit this after a long time but we are having the same problem. I gather from the last post that you are suggesting that for
the short steps the system stops the X and Y axes, turns the A axis ( knife ) and then starts again.  This sounds reasonable except that
I have seen our machine do a 90 degree turn without hesitating ( at high feed rate ). This was a mistake in the G-code but Mach3 handled
it without a problem.  It seems to me there is much more to this.

Could it be possible the small steps are messing with the tangential angle calculations?  I imagine if the resolution of the calculations is poor the
angles calculated could fluctuate a lot. I haven't seen the knif swing wildly back and forth so maybe this isn't the case.

Unfortunately I am trying to diagnose this from a distance as the machine is at my brother's workshop 1700km away.

The system is a cutting table with stepper motors driven through a smoothstepper.  Mach 3 is running in 2.5D mode with tangential steering of the
knife. The Z axis is wired to the spindle on/off output and is driven by pneumatics.

Can anyone offer any insight into this?

Cheers,
mike