On Apr 19, 2007, at 8:35 AM, Jon Elson wrote:

> This is mostly for John Kasunich, but anyone else with servo
> experience is welcome to chime in.
>
> My customer Stuart Stevenson has been having some problems
> with homing, so I had him send me his entire set of config
> files.  When I got them, I saw that he had what appeared to
> me to be VERY unusual PID settings.  When I run these on my
> machine, I see that the operation is "stable" but not accurate.
> By that, I mean that the machine continued moving for MINUTES
> after I took my finger off the jog key.  The following error was
> in the realm of inches.  Stuart says he has no following error
> to speak of, but I find that hard to believe.

Whether the creep is from P being too soft or the I value I don't  
have a clue.
However I've had good luck with the method below.

Wow! Every machine is different and the ideal  PID may also be 10X or  
more
different from another interface.
Set the gain on your amps so ~9 v in will give you max speed. This  
only works in
velocity mode. In torque mode follow the factory instructions.

Then:
First I'd set I  and probably FF1 to zero.
Then fire up halscope; plot following error and anything else that  
turns you on and start tuning.
Use short fast moves;  an inch at 200 ipm is fine.
40 for P would seem low for ppmc but that may simply mean that the  
gain on the
servo amps is turned up.
After I got reasonable stiffness from P but no oscillation then add the
FF1 back in and use it to make following error symmetrical around zero.
I tend to be VERY gentle with I & D.
Only after you have high speed (.5 to max speed of machine) with low
following errors; I get about 0.0005 at 200 ipm and 5-10X better at  
milling speeds,
should you try I and D. If they don't make much difference leave them  
out.
Simple is better.
This is just one approach; everyone has their favorite method. :-)

HTH

Dave
>
> Here are his settings :
>
> P =                              40
> I =                              4000
> D =                                0
> BIAS =                             0
> FF0 =                              0
> FF1 =                              1
>
> INPUT_SCALE is 10000
>
> Having so much I term just seems wildly wrong to my experience
> tuning this PID scheme.  I have noteced that the EMC2 D term is
> much more sensitive than it used to be.
>
> The servo interface is my PPMC analog servo board set.
> He is using Fenner drives from Servo Dynamics.
> I don't know whether they are set up for velocity mode or torque
> mode.  Since this is so different from my setup, I'm thinking
> maybe he has his amps set for torque mode.
>
> Does anyone have any ideas what is going on, here?
>
> Jon
>
> ---------------------------------------------------------------------- 
> ---
> This SF.net email is sponsored by DB2 Express
> Download DB2 Express C - the FREE version of DB2 express and take
> control of your XML. No limits. Just data. Click to get it now.
> http://sourceforge.net/powerbar/db2/
> _______________________________________________
> Emc-users mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/emc-users


-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
Emc-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to