Bruce Klawiter wrote:
> OK it appears I have my servos tuned as good as they are going to get.
> See images 5, 6 and 7: https://sites.google.com/site/bmklawt/home/pid-tuning
> Adding any I or D to any axis make things worse to violent.
>
>   
> Anyway I have terrible oscillation or dithering on the Z axis and if I move 
> the axis a bit the dithering changes, how can I stop this.
>   
Well, you should use some D.  You may have to start with values as low 
as .01 and work your way
up.  There will be a valley that gives the best results.  Definitely too 
much D causes oscillation.
Also a little deadband can help the system settle when idle.

BUT, there is a limit to how well it can do.  Any servo needs to hunt.  
If you have an encoder
with 50,000 counts/inch, then a jitter of +/- one encoder count will be 
almost undetectable.
On your machine, the encoder count is 10 um is approximately .0004" 
which is a lot.

The only advice I might give is to reduce P so you can put a little D on 
it, and then tighten
up the following error with FF1 and FF2, if needed.

But, your situation is why I always tell people who want to use cheap 
linear scales with
DRO-class resolution that they will run into exactly this problem.  You 
say the old
Anilam control did not have this problem, but I will BET that it also 
did not achieve
as tight a following error as you are now achieving.  That is the 
downfall, if you want
tight position control, you get dithering.  If you turn the P way down, 
the dithering
becomes a sluggish crawl back and forth that is undetectable, but the 
following error
is not kept small.

Jon


------------------------------------------------------------------------------
The demand for IT networking professionals continues to grow, and the
demand for specialized networking skills is growing even more rapidly.
Take a complimentary Learning@Cisco Self-Assessment and learn 
about Cisco certifications, training, and career opportunities. 
http://p.sf.net/sfu/cisco-dev2dev
_______________________________________________
Emc-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to