![]() | |
| Home Page | Mark Forums Read | Today's Posts | My Replies | Classifieds | Reviews | Photo Gallery | Web Links | Share Files | Advertise With Us | Ad List |
| |||||||
| LinuxCNC (formerly EMC2) Discuss LinuxCNC (formerly EMC2) Controlers here! |
| This forum is sponsored by: |
![]() |
| | LinkBack | Thread Tools | Search this Thread | Display Modes |
|
#1
| |||
| |||
| I'm sorry if this has been posted befor. I searched for it fairly extensivly, but maybe I didn't know which words to use to search for. I'm fairly new to EMC & haven't completely dug into the software inter-working parts. Some of our post-processors kick out some very tiny small G3 & G2 arcs. It looks as though if a rounding error sometimes causes the EMC processor to think the arc end position is the same as the current XY position, then a complete circle is executed around the IJ coordinates instead of a .0001 move. This is very likely to cause an undesired cut through the majority of the stock. Sometimes small circles are cut, sometimes large circles are cut, depending on the I&J coordinates. This breaks bits and destroys stock. The EMC backplotter does not detect, or graph this out before it executes, so nothing stops, or detects it before happening. The move might easily drive the XY mill bit position to the hardstops as in this example. Notice the last line's Y position is close enough to the previous that a rounding error will calculate it at the same position. The I coordinate will attempt to execute a massive circle instead of .0001 mm move. G1 X64.5751 Y137.1004 G3 X64.7079 Y135.4314 I220.773 J16.7207 G3 Y135.4313 I220.6402 J18.3896 It's pretty obvious that I should keep my post-processor resolution at a level of that of my EMC config, but I was wondering if there was a more robust solution for EMC? This is kinda a dangerous problem with the bit wandering around where it should not be, not to mention frustrating when the work is destroyed. Why is the backplotter calculating differently, or at a higher resolution than the execution process. If you would like screenshots, I can get those, but it might take 24hrs. It took a lot of troubleshooting to get the info I have, any further info would be greatly appreciated. Thanks! |
|
#2
| |||
| |||
| you dont say which cam software your using or post processor , i'd like to see any machine that can cope with a .0001 move ! correctly i'd suspect the cam package post processor is not handling small moves correctly is this a servo driven machine or stepper ! |
|
#3
| |||
| |||
| What version of EMC are you using? I vaguely remember some obscure problem that would create huge circles. that would probably be in EMM 2.3.X. You might want to send this to the EMC maillist. You will get a better answer - most likely. |
|
#4
| |||
| |||
| @ cncbasher, I think you are missing the point. I use several post-processors, the post processor that happened to create this code was CamBam. The machine is a stepper running EMC 2.4.3 I have already verified the post-processor is working correctly, and even narrowed down the problem in EMC. Please look up the G3 I & J codes if you do not understand this. The point is that EMC did not graph the large circles with the backplotter, nor did it error before attempting to execute large circles, and nothing indicated that it would do so ahead of time. I have increased the resolution of my post processor, but I may accidentally create another file on another computer with high resolution, and I'm not the only person post-processing for my machine all the time anyway. I don't expect a .0001mm move to execute correctly, but keeping my machine from exploding is in my best interest. @ chester How do I send this to the EMC mail list? Do I do that at LinuxCNC.org, or use sourceforge? I see a version 2.4.4 is out, but I've had difficulty upgrading due to my machine not being networked. I can temporarily connect it to the internet to do so, though. I don't see it in the version notes, upgrading may or may not fix it. I'd like to do this if I can before I pester developers. I'm not really sure how to safely upgrade anyway. Do I simply use the EMC repository through Ubuntu's upgrade manager? Thanks much! |
![]() |
| Tags |
| arc, crash, g2 g3, rounding error |
| Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
| Thread Tools | Search this Thread |
| Display Modes | |
| |
Similar Threads | ||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| Problem- Crashing | wileybrett | BobCad-Cam | 2 | 05-14-2011 09:16 AM |
| Esprit crashing | kip0569 | Esprit | 2 | 04-08-2011 12:29 PM |
| Need Help!- V23 crashing | RCE | BobCad-Cam | 15 | 03-12-2009 07:47 AM |
| Can't use Ctabdef or POLY commands, unreleased option used error | Ruselgun | G-Code Programing | 0 | 03-03-2009 04:19 AM |
| MCX crashing | baran3 | Mastercam | 1 | 03-26-2007 07:09 PM |