Hello Guest it is December 10, 2019, 05:57:38 AM

Author Topic: Cutter compensation - Found a bug and have questions  (Read 4238 times)

0 Members and 1 Guest are viewing this topic.

Cutter compensation - Found a bug and have questions
« on: September 01, 2010, 01:59:09 PM »
I've just recently started playing with cutter radius compensation in MACH3 Mill.  (R3.042.020)

The first thing I found was a bug, but I forget where to file a bug report.

At the top of each screen (to the right of the page tabs) is a display of the currently active modal G codes.  The display of cutter compensation codes (G41/42) appears to be reversed.  When my program calls for G42, the display shows G41.  Conversely, when my program calls for G41, the display shows G42.  Can someone tell me if this has been fixed in a more recent version or if I should file a bug report for this?

One other thing I noticed is more of an annoyance than a bug.  In the active G code display, the state of the various modal G codes that are being displayed does not reflect the current line of code being executed.  Instead, it displays the upcoming state based on the General Config value for Lookahead Lines.  For example, if line 50 in my program contains a G42 and my Lookahead Lines value is 20, the display will change to show G42 (actually G41 because of the bug) when line 30 in the program is executed.  I verified this by changing the Lookahead Lines value. I did find through testing that the Cutter diameter compensation is not actually activated until the code is executed, but it's somewhat misleading when the display is updated early.

Now for the questions.

I can take care of the annoying display problem by changing the Lookahead Lines value to 1 or 2, but should I?  The only things I've been able to find about the Lookahead Lines setting are rather vague.  Can someone give me a better explanation of this setting?

One other setting that I'd like more information on is the Advanced Compensation Analysis setting under the Mill Options tab on the Ports and Pins screen.  What exactly does it do?  What would the reason be to NOT have it set?

Thanks to anyone who can help!
Bill (the Cat) Shubert

Offline ger21

*
  • *
  •  6,291 6,291
    • View Profile
    • The CNC Woodworker
Re: Cutter compensation - Found a bug and have questions
« Reply #1 on: September 01, 2010, 10:18:53 PM »
Quote
The display of cutter compensation codes (G41/42) appears to be reversed.  When my program calls for G42, the display shows G41.  Conversely, when my program calls for G41, the display shows G42.  Can someone tell me if this has been fixed in a more recent version or if I should file a bug report for this?

I've been using comp for years, and have never noticed this. I just checked with 3.042.040, and it woks correctly. Perhaps what you're seeing is due to the next issue?

Quote
One other thing I noticed is more of an annoyance than a bug.  In the active G code display, the state of the various modal G codes that are being displayed does not reflect the current line of code being executed.  Instead, it displays the upcoming state based on the General Config value for Lookahead Lines.  For example, if line 50 in my program contains a G42 and my Lookahead Lines value is 20, the display will change to show G42 (actually G41 because of the bug) when line 30 in the program is executed.  I verified this by changing the Lookahead Lines value. I did find through testing that the Cutter diameter compensation is not actually activated until the code is executed, but it's somewhat misleading when the display is updated early.

This is because Mach3 is a buffered system. The code is always read in advanced and placed into a queue. I think the displayed code may be more accurate in Version 4, but for now, it is what it is.
I believe that reducing the lookahead will reduce performance in CV mode, especially with 3D work. I personally don't need Mach to tell me when I'm in G41/G42, so it's not a big deal to me.

Quote
One other setting that I'd like more information on is the Advanced Compensation Analysis setting under the Mill Options tab on the Ports and Pins screen.  What exactly does it do?  What would the reason be to NOT have it set?

I don't think there is a reason not to use it. I think when the advanced mode was added, the original was left in place. I'm pretty sure the original comp will gouge inside corners if they don't have a radius.
Gerry

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

JointCAM Dovetail and Box Joint software
http://www.g-forcecnc.com/jointcam.html
Re: Cutter compensation - Found a bug and have questions
« Reply #2 on: September 02, 2010, 04:16:28 PM »
Thanks for your reply Gerry!

I did some more investigating/playing and I think I might have found why it worked for you.  For Tool Zero (no compensation), or any tool that has a positive value in the diameter field in the Tool Table, the display is correct. The only time that the display output is reversed is when I have a negative value for the Tool Diameter.  My programs are written for a Tool Path Contour based on a nominal tool diameter.  If the tool I'm using is undersize, I need to use a negative value for Tool Diameter Compensation.  In this case, the G41/42 value in the display is reversed.  If you could verify this with your version of MACH, it still may need to be reported as a bug.  Just set the Tool Diameter value for Tool 1 to a negative value and make it the active tool.  Then MDI a G41/42 and see if your display is also reversed.

Thank you again for your answers to my questions, but, again, the information was still too vague for me to make an informed decision.

You said that you thought that reducing the Lookahead would "reduce the performance in CV mode"?????  I've read the CV Settings documentation and know that they recommend a value of 200 (the default appears to be 20). I have no idea what sort of calculations are being done (besides how to transition from the current move to the next), but I guess I could see a need to "lookahead" on complex 3D contouring that is accomplished with sequences of thousands of short moves. Since my profiles are not very complex (not large sequences of very short moves), I have a hard time understanding how much benefit there would be to using a large value for Lookahead.  In addition, the CV Settings documentation even states that too large a value could cause problems depending on the speed of the computer. As it is, I have no feel for how few is "too few" and how many is "too many".

In reference to Advanced Compensation Analysis, you stated that you didn't think there was a reason not to use it and that the original mode was simply left in place when the Advanced mode was added.  If there's no reason not to use it, why is it a choice?  Any time there is something to be gained....there is always an associated cost. There must be pros and cons....I just don't know what they are.

Bill (the Cat) Shubert

Offline ger21

*
  • *
  •  6,291 6,291
    • View Profile
    • The CNC Woodworker
Re: Cutter compensation - Found a bug and have questions
« Reply #3 on: September 02, 2010, 08:59:23 PM »
By using a negative diameter, it's effectively doing the same as the opposite offset, so that's probably why your seeing the opposite offset displayed. G41 with a negative offset is the same as a G42 with a positive offset. Why don't you just have Mach3 do the full radius offset, then it'll work like you expect it to. That's how I use it.

Sorry, but I can't check as I'm getting ready for a trip and don't have much time.

As for what I said about reduced performance, ai was really referring to what you said about setting lookahead to 2 or 3.

Unfortunately, I really don't have any more information for you.
Gerry

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

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