On Dec 11, 2008, at 6:11 AM, Howard Chan wrote:

> Dear Jon,
> Today I tune the PID again.  Would you mind help me to check the value
> and give me some suggestion?  Thank you!
>
> [TRAJ]
> DEFAULT_VELOCITY =      13.30
> MAX_VELOCITY =          33.34
> DEFAULT_ACCELERATION =  45.0
> MAX_ACCELERATION =      50.0
>
> [AXIS_1]
> MAX_VELOCITY =          50.0
> MAX_ACCELERATION =      50.0
> BACKLASH = 0.000
> INPUT_SCALE =           -2835
>
> When P=20, I=50, D=0, FF1=0, FF2=0 and click 'F2'
> http://myweb.polyu.edu.hk/~icwfchan/p1.png
>
> When P=20, I=50, D=0, FF1=0.198, FF2=0 and run 'Y=20'
> http://myweb.polyu.edu.hk/~icwfchan/p2.png
>
> When P=20, I=50, D=0, FF1=0.198, FF2=0.00021 and run 'Y=20'
> http://myweb.polyu.edu.hk/~icwfchan/p3.png
>
> Howard
>> Howard Chan wrote:
>>
>>> Dear Jon,
>>> Thanks you give me some suggest.  Today I try to use small value  
>>> of the
>>> P. But there are an error when I click 'F2'.
>>> http://myweb.polyu.edu.hk/~icwfchan/error.png
>>> Would you mind telling me what has happen?
>>>
>>>
>>>
>> WOW, very strange.  I think it is integral windup.  Your  
>> parameters of
>> huge I and small D are quite odd.  Set I to zero first, and see if  
>> this
>> gyration when you hit F2 goes away.  I think it will.
>>
>>> http://myweb.polyu.edu.hk/~icwfchan/x.png
>>> I would like to ask you is it the accel/decel too high and how can I
>>> find the best value of DEFAULT_ACCELERATION
>>>
>>>
>>>
>> Find out what the drive is capable of, by increasing it in moderate
>> steps until a large error suddenly appears during the accel/decel  
>> parts
>> of the move.  Then, back off, say 20 to 25% from that value.  On my
>> Bridgeport mill, if I turn up the accel too much, I break timing  
>> belts.
>> I know that the belts are undersized for the mass of the machine.
>>
>>> My setting:
>>> [TRAJ]
>>> DEFAULT_VELOCITY =      10.0
>>> MAX_VELOCITY =          37.6
>>> DEFAULT_ACCELERATION =  112.8
>>> MAX_ACCELERATION =      138.0
>>>
>>> [AXIS_0]
>>> MAX_VELOCITY =          56.4
>>> MAX_ACCELERATION =      112.8
>>> P =                     4.000
>>> I =                     100.000
>>> D =                     0.00
>>> FF0 =                 0.000
>>> FF1 =                 0.089
>>> FF2 =                 0.0019
>>>
>>>
>>>
>> Jon
Hi Howard,

        You are making progress. :-)

        However, it my understanding that [TRAJ] values are global and it is  
a good idea to set [AXIS] values below the globals.

        I usually tune by plotting following error which is the difference  
between commanded position and actual position.
        I think you have the potential for somewhat better following than  
you are getting.
        
        Try different values for FF1 and FF2.
        IIUC the ideal value for FF1 is 1.0 but that will not necessarily  
give good results on any given machine.
        Using somewhat more FF2 may help suppress the accel and decel spikes  
at the beginning and end of your moves.

        Good luck and keep at it. You are getting there.

Dave
        
>>
>> --------------------------------------------------------------------- 
>> ---------
>> SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las  
>> Vegas, Nevada.
>> The future of the web can't happen without you.  Join us at MIX09  
>> to help
>> pave the way to the Next Web now. Learn more and register at
>> http://ad.doubleclick.net/clk;208669438;13503038;i?http:// 
>> 2009.visitmix.com/
>> _______________________________________________
>> Emc-users mailing list
>> Emc-users@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/emc-users
>>
>>
>
>
> ---------------------------------------------------------------------- 
> --------
> SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas,  
> Nevada.
> The future of the web can't happen without you.  Join us at MIX09  
> to help
> pave the way to the Next Web now. Learn more and register at
> http://ad.doubleclick.net/clk;208669438;13503038;i?http:// 
> 2009.visitmix.com/
> _______________________________________________
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users


------------------------------------------------------------------------------
SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada.
The future of the web can't happen without you.  Join us at MIX09 to help
pave the way to the Next Web now. Learn more and register at
http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/
_______________________________________________
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to