[Emc-users] Curve in place of square edges...
Hi all, I use the EMC 2.1.6 with a Bridgeport style, three axis BLDC servo system. The servos are digital types with own position loops. Two parports, software PID loops and freqgen used, see that picture http://www.upload-images.net/imagen/390bf13788.jpg for more. I use this G code to test: G92.1 G61.1 G10 L2 P2 X0 Y0 Z-0 F500 G55 G1 X0 Y0 Z0 G01 X50 Y0 G01 X50 Y50 G00 X0 Y0 M30 It's qiute accurate, just a little bit slow. But the edges what it makes likes to a curves not a sharpe edge. After a lot of time to tune up, I have no idea about it... I try the G61, G61.1, G64, but nothing changed. When I check the pid.0.error (X axis) and pid.1.error (Y axis) on halscope it seems to me the axis.1.motor-pos-cmd don't wait to the axis.0 to reach their destination (X50) and starts to move. So the X and Y axis move the same time. I think the trajectory planner don't use the axis.N.motor-pos-fbto wait the exact stop. Just send the command one after other. Where is the bug?HAL, parameters, SW or where... Thx, Tamas Konya - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
[Emc-users] Curve in place of square edges...
I have also noticed a slightly similar thing but in a different context. I have a script in which a cut is made by the Y axis then this returns by means of a G0 and the work is turned by the A axis. The A axis always starts to move just before the Y axis stops. This is on a basic stepper machine and I am using the latest EMC2-Ubuntu iso release now - 2.1.6. Sorry Ray, didn't mean to break it just yet!!! ;-} lol -- Best wishes, Ian Ian W. Wright Sheffield UK The difference between theory and practice is much smaller in theory than in practice... - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Curve in place of square edges...
The lower you have your acceleration set - the more you will see emc2 'blending' http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?TrajectoryControl If you want it to be exact path you need to use G61. This will stop at every endpoint and follow the path exactly. You could do a G64 Px.xxx where x.xxx is how close you want emc to follow the actual path. emc will slow down during direction changes enough to stay within the tolerance you specify. - Original Message - From: Tamas Konya To: emc-users@lists.sourceforge.net Sent: Thursday, August 23, 2007 7:33 AM Subject: [Emc-users] Curve in place of square edges... Hi all, I use the EMC 2.1.6 with a Bridgeport style, three axis BLDC servo system. The servos are digital types with own position loops. Two parports, software PID loops and freqgen used, see that picture http://www.upload-images.net/imagen/390bf13788.jpg for more. I use this G code to test: G92.1 G61.1 G10 L2 P2 X0 Y0 Z-0 F500 G55 G1 X0 Y0 Z0 G01 X50 Y0 G01 X50 Y50 G00 X0 Y0 M30 It's qiute accurate, just a little bit slow. But the edges what it makes likes to a curves not a sharpe edge. After a lot of time to tune up, I have no idea about it... I try the G61, G61.1, G64, but nothing changed. When I check the pid.0.error (X axis) and pid.1.error (Y axis) on halscope it seems to me the axis.1.motor-pos-cmd don't wait to the axis.0 to reach their destination (X50) and starts to move. So the X and Y axis move the same time. I think the trajectory planner don't use the axis.N.motor-pos-fb to wait the exact stop. Just send the command one after other. Where is the bug?HAL, parameters, SW or where... Thx, Tamas Konya -- - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ -- ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Curve in place of square edges...
On Thu, Aug 23, 2007 at 08:51:09AM -0500, Sam Sokolik wrote: The lower you have your acceleration set - the more you will see emc2 'blending' http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?TrajectoryControl If you want it to be exact path you need to use G61. This will stop at every endpoint and follow the path exactly. You could do a G64 Px.xxx where x.xxx is how close you want emc to follow the actual path. emc will slow down during direction changes enough to stay within the tolerance you specify. I tested G61 and did not find any problems. If on a servo system you are getting rounded paths still with G61, your PID tuning probably needs attention. Check your following errors. If you see sharp corners in the backplot but round corners on your part, it's certainly a tuning problem. Chris - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
Re: [Emc-users] Curve in place of square edges...
Hi, Thanks everybody the answers. I made a picture about pid.0.error (X-axis), pid.1.error (Y-axis), axis.0.motor-pos-cmd and axis.0.motor-pos-cmd. You can see it in: http://www.upload-images.net/imagen/c6ebc76f47.png , an other after the acceleration changed smaller value: http://www.upload-images.net/imagen/f8e3a1376b.png The red and blue are the error signals, green and violet are the motor position commands. You can see the second axis starts moving before the first stop. Sure, the machine has a continuous following error, but I though the EMC2 will tolerate it. If the trajectory planner use the feedback signal to detect the exact stop in place of own position output, it could work. So, it's seems to me, I have to reduce the following error to get sharp edge. Like Chris sad, I have to tune my PID more strong, than I get worse accuracy :( (Backplot has sharpe edge.) Thx, Tamas On 8/23/07, Chris Radek [EMAIL PROTECTED] wrote: On Thu, Aug 23, 2007 at 08:51:09AM -0500, Sam Sokolik wrote: The lower you have your acceleration set - the more you will see emc2 'blending' http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?TrajectoryControl If you want it to be exact path you need to use G61. This will stop at every endpoint and follow the path exactly. You could do a G64 Px.xxx where x.xxx is how close you want emc to follow the actual path. emc will slow down during direction changes enough to stay within the tolerance you specify. I tested G61 and did not find any problems. If on a servo system you are getting rounded paths still with G61, your PID tuning probably needs attention. Check your following errors. If you see sharp corners in the backplot but round corners on your part, it's certainly a tuning problem. Chris - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/___ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users