Following error and chorderror problems, so strange

[awen]

I did some experiment using stepper motor control by 7344 using closed loop control. I developed a interpolation algorithm and I generated position points for the half circle and check using tow stepper axis. I get the position data and following error data from the software. The Chord error is calculated using this way, say (x,y) is the position data I got and the Radius is R, Chorderror=R-sqrt(x^2+y^2); After experiment I got a very strange result. The chorderror I calculated is smaller than the following error. The data I attached.  Is there anyone can explain this strange problem? Thanks a lot!

The first file is the data file, RMS-Fe means Root Mean Square Following Error, RMS-CE means Root Mean Square Chord Error. M-Time means total machining time and Average-Vel means Average Velocity.

The second file the Figure generated by plotting the position, velocity, following error and chord error.

[shoneill]

If I understand correctly, the following error is calculated basically along the circumference of the circle of rotation.

The chord error (basically the distance between two points which don't neccessarily lie on the circumference, is this correct? (I'm assuming the center of your circle is 0,0).

Draw a diagram of what you're actually measuring and it'll be clear why the two lengths don't have to correlate.

In case you don't have time to do this, here's an approximation (in your excel file).

If your chord error is large, this means your circle center os offset or your XY trajectory is simply not a circle.  Otherwise this returns more or less your measurement accuracy.

Your following error is the difference between the IS position and the SHOULD position ALONG the axis of travel.  If you have a perfect circle, and the center is 0,0 then basically your chord error should ALWAYS be 0, even if the following error is huge.

Hope this helps

Shane.

EDIT: After looking at the word file, I can say the following.  The chord error is basically telling you how circular the actual motion path is (of how offset the center is) whereas the following error is telling you how well the motor is tracking the position you're telling it to take.

The two values are almost completely independent of each other.  There is no reason to assume a correlation.

Message Edited by shoneill on 04-07-2006 10:20 AM