Hello Guest it is January 26, 2020, 09:39:29 PM

Author Topic: Mach4 & ESS and G43 issues  (Read 2887 times)

0 Members and 1 Guest are viewing this topic.

Mach4 & ESS and G43 issues
« on: February 08, 2016, 09:47:18 AM »
I decided to test MACH4 and my ESS after waiting another six months to see if it is now stable.  I have installed Build 4.2.0.2872, the latest ESS plugin and have the machine moving fine and homing works fine as well.  I decided to load a small Gcode file that I have run on MACH3 fine as just a little test to see if things look good before getting the license for MACH4.

My little Gcode file is to build a small bracket and it uses four tools.  I do not have an automatic tool changer so for the test I did not even turn on my spindle, nor did I have any tool in the spindle.  I referenced all axis and that worked fine.  Then I put in a offset of three inches in X and Y and zero those DROs.  I set the Z0 about two inches off the table as this was just going to be an air cut to see how well motion was working.

The first part of the code worked just fine.  It hits the first tool change and I click ok.  There is no tool so I just tell it to proceed with cycle start and it does the next segment just fine as well.  On the third tool change things go wrong.

N480 M01
N490 T3 M6
N500 G0 G90 G54 X.8028 Y.6692 S8000 M3
N510 G43 H3 Z.1
N520 G99 G83 Z-.407 R.1 Q.1 F4.28

On line N510 the Z axis goes full speed all the way down into the table.  No tool broken and I hit the emergency stop to avoid damage to the ballscrew.

This segment of code is pretty much identical to the two previous parts of the code after a tool change.  The DRO reads something like Z 1.5 when the spindle is crashing into the table.

T1 M6
G43 H1 Z1.5

T2 M6
G43 H2 Z1.5

T3 M6
G43 H3 Z.1     *** this one crashes....

I just looked at the release notes for this version and it says they cleaned up fixture offset corrupting the tool table.   Well something seems corrupted for sure....

Keep in mind this code works perfect under MACH3 with no changes.  Since I am not using the automatic tool changer, I never populate the Tool table.  I assume and seems to be confirmed by others that MACH should use a zero offset and use the Z setting created when using G54.

This is a very short program and it is next to impossible to troubleshoot the issues due to the fact they have the timeout value for the DEMO version so short you can't do much testing.  The previous Gcode lines of code limit seemed to work much better at least you could test to see if everything seemed to be working.  Anyway this appears to be a bug unless I have just done something completely wrong.  Any help or suggestions would be much appreciated.  Thanks

Russ




Build 4.2.0.2872 - 01 February 2016

1) Update core API documentation covering threading, rigid tapping,
   and general recommended practices for plugins.

2) Updated handling of canned cycles to address issues with spindle
   speed override and soft limits violations.

3) Fix problems resulting from use of Unicode characters in file names.

4) Added infrastructure needed to support user controllable CV modes.
   No user interface yet though.

5) Cleaned up some issues with fixture offsets corruptings tool table values.

6) Allow jog rate override to work while jogging is active.

7) Added plugin to allow send and receive from RS-232 serial port.

8) Convert LUA environment to use dynamic runtime to support loadable
   LUA modules. LUA RS232, LFS file system, LUA Sockets with TCP modules
   now provided.

9) Update SIM plugin to support CALLBACK function allow SIM to be used
   to monitor other plugins.

10) Added mcPluginPanel control to allow a plugin to have a panel that is
    integrated into the screen set as opposed to a floating window. If
    this is used, a button must be provided on this panel to access the
    normal plugin configuration dialog if needed. SIM plugin updated to
    include example.

11) Provide autofill of function call paramaters while using LUA editor.

12) Added new control type to provide gage control to replace the default
    Windows gage (which could not set colors, and had flicker issues).

13) Added ability to turn menu bar on and off in screen sets from a screen
    script.

14) Fixed screen offsetting issue when editing a screen in a scrolled view.

15) Fixed tool path mouse events not working after a screen edit.

16) No screen set updates in this release.

Offline BR549

*
  •  6,925 6,925
    • View Profile
Re: Mach4 & ESS and G43 issues
« Reply #1 on: February 08, 2016, 11:27:27 AM »
The 3rd section is VERY different from the first 2 notice where you are sending Z in teh 3rd section. It goes to Z.1 not Z1.5 that there is a 1.4" difference in height so you are sending it towards teh table AlSO what are the offset values that you have in the tool table for these tools ?? IF teh offset difference between 2 &3 was enough it would be very easy to send the tool into the table.


Before you think about Mach4 double check the math of what you told it to do.

Just a thought, (;-) TP

Offline ger21

*
  • *
  •  6,291 6,291
    • View Profile
    • The CNC Woodworker
Re: Mach4 & ESS and G43 issues
« Reply #2 on: February 08, 2016, 12:59:30 PM »
If you're not setting any length values in the tool table, then you shouldn't be using tool length offsets in your g-code. Just because it works in Mach3, doesn't mean it's right.
Gerry

2010 Screenset
http://www.thecncwoodworker.com/2010.html

JointCAM Dovetail and Box Joint software
http://www.g-forcecnc.com/jointcam.html
Re: Mach4 & ESS and G43 issues
« Reply #3 on: February 08, 2016, 01:00:06 PM »
BR549,

Thanks for your response.  I went down and looked inside of the tool table and discovered it had numbers populated for that tool.  The developers were probably testing the previous issues reported and forgot to zero them back out before posting the release.  I zero'd those out and the code ran perfect right through all four tools without crashing.  Next to get the spindle working and then my pendant and then I will be ready to jump on the MACH4 band wagon.  Thanks

Russ
Re: Mach4 & ESS and G43 issues
« Reply #4 on: February 08, 2016, 01:02:44 PM »
Gerry,

The POST processor put in the Tool Length stuff, never noticed that before.  Anyway it gets ignored and uses what was set if the tool table is empty.

Russ

Offline ger21

*
  • *
  •  6,291 6,291
    • View Profile
    • The CNC Woodworker
Re: Mach4 & ESS and G43 issues
« Reply #5 on: February 08, 2016, 01:33:53 PM »
It's the user's responsibility to know what the post processor is doing.
Remember that next time you update the software, and it drives a real tool into your table.
Gerry

2010 Screenset
http://www.thecncwoodworker.com/2010.html

JointCAM Dovetail and Box Joint software
http://www.g-forcecnc.com/jointcam.html

Offline BR549

*
  •  6,925 6,925
    • View Profile
Re: Mach4 & ESS and G43 issues
« Reply #6 on: February 08, 2016, 02:15:05 PM »
HIYA RUss not a problem that is an easy one to miss if you are not used to proofing teh Gcode.

Been there done that one a few times myself (;-)

(;-) TP