× Forum Header

Help solving joint following error?

More
09 Dec 2010 13:01 #5939 by BWay
When I run my router I need to slow my rapids to about 100 in/min (depnding on my ferror value) so that i don't get a joint following error. I am puzzled by this since when i am in the Step Conifuration wizard I can run tests for each axis and run around 480 in/min (8 in/sec) with acceleration values of 100 in/secĀ². When i originally designed my machine I selected my encoders to make sure I wasn't going to exceed the capability of my computer. So here is my questions...why would step configuration wizard allow me to run faster than I can run an actually program? From much of what I have read the step configuration wizard is a good way to test how fast the machine can run. What am I missing?

Thanks!

Please Log in or Create an account to join the conversation.

More
09 Dec 2010 22:17 #5949 by BigJohnT
You have not given enough info to answer your question...

You mention encoders and the Stepper Configuration Wizard those two don't mix. Stepper systems 99.999% of the time don't have encoders and trying to use encoders for position feedback for steppers generally won't work due to the design of steppers. Encoders into the parallel port generally is not a good idea either...

www.linuxcnc.org/docview/html/common_Stepper_Diagnostics.html

The Step Configuration Wizard "test this axis" does not use encoders, it only drives the steppers.

John

Please Log in or Create an account to join the conversation.

More
10 Dec 2010 01:34 #5952 by BWay
I am using Gecko G320 drives with encoders. The Configuration Wizard does allow using the G320 as a driver and bacially it works with the exception of speed (and some problems with my gantry motors).

Please Log in or Create an account to join the conversation.

More
10 Dec 2010 14:51 #5957 by aike
I tested with FERROR variable. If this variable is low, the "joint follow error appears"
Also some time the motor lost the impulses and due to diffrerence between command value and encoder feedback
value the follow error appears.
If your computer is not fast enough, this error also may appear.
To avoid it I set FERROR = 0.5.
I make the accurate tuning in my lathe and during the programm execution, the FERROR not reached 0.005 mm.
I create custom.xml and display X-ferror and Z-ferror in real time in pyvcp window.
It looks like this
net X.ferror axis.0.f-error pyvcp.ferror

Please Log in or Create an account to join the conversation.

More
10 Dec 2010 17:46 #5962 by BigJohnT
The Gecko 320 is a step and direction drive like steppers and the encoder is not connected to EMC in any way. For a step and direction drive read the trouble shooting guide for steppers.

John

Please Log in or Create an account to join the conversation.

More
10 Dec 2010 18:10 #5963 by BWay
Thanks for the feedback. I checked out the stepper system error documentation and have a hunch that my acceleration values are too high casuing a following error. I tried slowing down the acceleration a few days ago and it didn't seem to help but it is worth another shot.

I will post the results later.

Please Log in or Create an account to join the conversation.

More
14 Dec 2010 13:04 #6026 by andypugh
Generally the only source of f-errors in stepgen-based systems is a base thread period which is too long for the requested step frequency.

The stepgen can only send one pulse per base-thread period (or per two periods if reset is not enabled on the step pins). Try a hand calculation of the step rate at peak commanded speed (including the effect of microstepping) and see if this is your problem.

Please Log in or Create an account to join the conversation.

Time to create page: 0.078 seconds
Powered by Kunena Forum