Hello Guest it is November 09, 2024, 04:32:38 AM

Author Topic: Inverse time feed bug?  (Read 5778 times)

0 Members and 1 Guest are viewing this topic.

Inverse time feed bug?
« on: January 05, 2009, 03:41:55 PM »
I am using Mach3 to control 4 axes (3 axis mill with rotary table), and can't get G93 inverse time mode to work properly.  Here's some g-code that does not work as expected:

Code: [Select]
G93
G0 X0 Y0 Z0.4 A-0.7
G1 F1000 Y0.00003 A0.7

The problem is the G1 move, which is specified to take 0.06 seconds but actually takes many seconds to complete a motion of 1.4 degrees in A and .00003" in Y).  When the insignificant Y axis motion is removed, Mach 3 performs the G1 in a small fraction of a second as expected.

Can someone confirm that this is a bug?  If it is, how do I report it.

What about a work around?  My best bet may be to go back to G94 (units per time feed mode) generating synthetic feedrates using the square root of the sum of the squares of the axis movements (inches for X, Y and Z, and degrees for A) divided by the desired time for the operation.

If it helps, I think may have updated my installation to the latest version of Mach 3, but I'm not certain.  The "about" screen just says R3.00.

vmax549

*
Re: Inverse time feed bug?
« Reply #1 on: January 05, 2009, 04:12:23 PM »
Ken the BEST way to report it is to make a copy of your test codefile AND a copy of your xml file and send it to his email . In the header use a short descriptor such as  "G93 BUG Report"

(;-) TP
Re: Inverse time feed bug?
« Reply #2 on: January 05, 2009, 09:59:42 PM »
Thanks for the bug reporting advice.  I got an answer back already!  G93 only works for XYZ moves, adding ABC is on their to do list.

I modified my g-code generating program to produce synthetic feedrates for A axis moves in G94 mode.  Seeing feed rates of F2000 in the g-code for a mill that moves at 20ipm (at best) looks very strange, but a test run on a piece of scrap works reasonably well.  This solution isn't perfect because the g-code has no control over whether Mach will choose a blended feedrate based on the rotation diameter setting, but I can cut my parts.