John Kasunich wrote: > Thomas Kaiser wrote: > > > Today I measured the time (with a stop watch) to move 200 mm in > > full speed (G00 Z-200.0) and got +8 seconds. Looks like the it > > does not move at 3000 mm / min. > > > > [EMCMOT] > > EMCMOT = motmod > > SHMEM_KEY = 111 > > COMM_TIMEOUT = 1.0 > > COMM_WAIT = 0.010 > > BASE_PERIOD = 100000 > > SERVO_PERIOD = 1000000 > > BASE_PERIOD = 100000 means you can never move faster than 100 > microseconds between each step pulse. That is 10000 step pulses per second. > > > [AXIS_0] > > SCALE = 400.0 > > Max step rate of 10000 steps/sec, divided by 400 steps/mm, means a > maximum speed of 25mm per second. At 25mm/second, it will take 8 > seconds to travel 200mm, exactly as you measured.
That looks right. I am familiar with PLC programming and there we have to check the cycle time of the PLC program. That means I got It. This configuration was done with stepconf. So, I just entered the value in stepconf and thought it is ok. The 40000 step on time can not be achieved in the base_period and never in the servo_period cycle :-( I use Linux since 1998, but its the first time I use a real time kernel, now. > > > MAX_VELOCITY = 50.0 > > But your configuration is asking for 50mm per second. It is impossible > for the step generator to generate that speed. It will fall behind the > commanded position and cause a following error - unless you set the > following error limit to a crazy value: I don't know the correct velocity of these drives, so I just put something there. > > > MIN_FERROR = 10000.0 I add this to avoid the following error. I know I have to change after I could set up this machine right, but is this really needed? I don't have any feedback from the axis, So how can EMC2 detect a following error? EMC2 is only sending pulses to the axis and assumes that the axis moves this distance. I am wrong on this one? > > The following error limit tells EMC "if you ever find yourself this far > away from the proper position, STOP!" > > When you set following error to 10000mm, you are telling EMC "it is OK > if you cut 10 meters away from the proper location". NOT GOOD! See above > > If you get following errors, you should NOT fix them by raising the > following error limit. You should figure out the real reason for the > errors and fix it properly. I know, but when you starting with this and you don't have the "big picture" it helps at the first try to move axis, doesn't it? > > I think the reason that the jogs are running after you stop is that the > commanded position (moving at 50mm/sec) has gotten ahead of the motors > (moving at 25mm/sec). After you release the button, the commanded > position stops, but the motors keep running until they catch up. After reading your explanation, I think it is this. > > Fix the speed problem and the crazy following error limit, and the jog > problem will fix itself. Your right, too. But I have to start somewhere. Only reading documentation does not help. When some other person move the light to an other direction helps sometimes more. Thanks a lot. Thomas ------------------------------------------------------------------------------ This SF.net email is sponsored by: SourcForge Community SourceForge wants to tell your story. http://p.sf.net/sfu/sf-spreadtheword _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
