Dynomotion

Group: DynoMotion Message: 2481 From: skyluvr Date: 11/30/2011
Subject: Motor tuning blues
Hi Folks,

I have the motor tuning blues. I cannot seem to get the motors tuned on the x axis of my vertical mill. I am trying to tune a closed loop stepper system that is using the linear scales for feedback. I am starting to suspect that using the linear scales for feed back my be causing my problems. The mill has ball screws and I can detect no backlash with an indicator. Have others used the scales successfully? The mill currently runs Mach3 through the parallel port and works ok. I wanted to start using the Kflop to control the machine closed loop and get away from the mach driver.

I am using mirco-step controllers at 2000 steps/rev on 200step/rev motors and the scales are .0005 resolution. I am using a gain of (2000mircostepsX200step motor=400000/8000=50. Is this right? At a gain of 50 I can not tune the motor. If I adjust it to say 11 I can match the command and the position closely. But at 50 there is no way.

Any help would be greatly appreciated. Maybe a step by step on tuning somewhere?

Thanks in advance

Rick
Group: DynoMotion Message: 2482 From: Tom Kerekes Date: 11/30/2011
Subject: Re: Motor tuning blues
Hi Rick,
 
You didn't provide enough information for us to determine your uStep resolution.  We also need your leadscrew pitch and any other mechanics.
 
Its not clear what "gain" you are referring to.
 
But the first step is to set the InputGain0 parameter to match your encoder position measurement to your commanded uStep destinations.  Test this by running in open loop Step/Dir output mode and verify that the plots match within the mechanical accuracy of your lead screw and mechanics.  This is described here:
 
 
Let us know when you have that part working.
 
Regards
TK
 
 
 

Group: DynoMotion Message: 2604 From: skyluvr Date: 12/5/2011
Subject: Re: Motor tuning blues
Hi Tom,

I was able to get the axis to move the proper distance when commanded in Step/Dir mode. Using a Input0 gain of 2.65.

The motor is coupled to the axis with a toothed belt drive a gear reduction of 2.6:1 to a ball screw of .2/rev.

Rick

--- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@...> wrote:
>
> Hi Rick,
>  
> You didn't provide enough information for us to determine your uStep resolution.  We also need your leadscrew pitch and any other mechanics.
>  
> Its not clear what "gain" you are referring to.
>  
> But the first step is to set the InputGain0 parameter to match your encoder position measurement to your commanded uStep destinations.  Test this by running in open loop Step/Dir output mode and verify that the plots match within the mechanical accuracy of your lead screw and mechanics.  This is described here:
>  
> http://www.dynomotion.com/Help/ClosedLoopStep/ClosedLoopStepper.htm
>  
> Let us know when you have that part working.
>  
> Regards
> TK
>  
>  
>  
>
>
> ________________________________
> From: skyluvr <lovethebuzz69@...>
> To: DynoMotion@yahoogroups.com
> Sent: Wednesday, November 30, 2011 11:40 AM
> Subject: [DynoMotion] Motor tuning blues
>
>
>  
> Hi Folks,
>
> I have the motor tuning blues. I cannot seem to get the motors tuned on the x axis of my vertical mill. I am trying to tune a closed loop stepper system that is using the linear scales for feedback. I am starting to suspect that using the linear scales for feed back my be causing my problems. The mill has ball screws and I can detect no backlash with an indicator. Have others used the scales successfully? The mill currently runs Mach3 through the parallel port and works ok. I wanted to start using the Kflop to control the machine closed loop and get away from the mach driver.
>
> I am using mirco-step controllers at 2000 steps/rev on 200step/rev motors and the scales are .0005 resolution. I am using a gain of (2000mircostepsX200step motor=400000/8000=50. Is this right? At a gain of 50 I can not tune the motor. If I adjust it to say 11 I can match the command and the position closely. But at 50 there is no way.
>
> Any help would be greatly appreciated. Maybe a step by step on tuning somewhere?
>
> Thanks in advance
>
> Rick
>
Group: DynoMotion Message: 2605 From: Tom Kerekes Date: 12/5/2011
Subject: Re: Motor tuning blues
Hi Rick,
 
InputGain0 should have no effect on the motion when in Step/Dir mode which doesn't make use of the Encoder as an input device (except for plotting).  So I dont really understand what you are saying.  Maybe you could explain a bit more what you are doing.
 
It looks like your uStep resolution is:
 
2000 Counts/rev * 2.6 / 0.2 Inch/rev = 26000 counts per inch.
 
With your OutputGain at 1.0 and in Step/Dir mode if you move 260,000 counts does the axis move 10 inches?
 
If this is true then because your scales have a resolution of only 2000 counts/inch and with 26000 uSteps/inch then you should use an InputGain0 of 13.0 to match them.
 
Regards
TK
Regards
TK 
 
 
 

Group: DynoMotion Message: 2684 From: skyluvr Date: 12/11/2011
Subject: Re: Motor tuning blues
HI Tom,

Yes, you are right. With a OUTPUTGAIN of 1.0 the axis does in fact move 10 inches when commanded 260,000 counts.

I am going reconnect the encoder and set the InputGain0 to 13 and see how it behaves.

Rick

--- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@...> wrote:
>
> Hi Rick,
>  
> InputGain0 should have no effect on the motion when in Step/Dir mode which doesn't make use of the Encoder as an input device (except for plotting).  So I dont really understand what you are saying.  Maybe you could explain a bit more what you are doing.
>  
> It looks like your uStep resolution is:
>  
> 2000 Counts/rev * 2.6 / 0.2 Inch/rev = 26000 counts per inch.
>  
> With your OutputGain at 1.0 and in Step/Dir mode if you move 260,000 counts does the axis move 10 inches?
>  
> If this is true then because your scales have a resolution of only 2000 counts/inch and with 26000 uSteps/inch then you should use an InputGain0 of 13.0 to match them.
>  
> Regards
> TK
> Regards
> TK 
>  
>  
>  
>
>
> ________________________________
> From: skyluvr <lovethebuzz69@...>
> To: DynoMotion@yahoogroups.com
> Sent: Monday, December 5, 2011 8:41 AM
> Subject: [DynoMotion] Re: Motor tuning blues
>
>
>  
> Hi Tom,
>
> I was able to get the axis to move the proper distance when commanded in Step/Dir mode. Using a Input0 gain of 2.65.
>
> The motor is coupled to the axis with a toothed belt drive a gear reduction of 2.6:1 to a ball screw of .2/rev.
>
> Rick
>
> --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> >
> > Hi Rick,
> >  
> > You didn't provide enough information for us to determine your uStep resolution.  We also need your leadscrew pitch and any other mechanics.
> >  
> > Its not clear what "gain" you are referring to.
> >  
> > But the first step is to set the InputGain0 parameter to match your encoder position measurement to your commanded uStep destinations.  Test this by running in open loop Step/Dir output mode and verify that the plots match within the mechanical accuracy of your lead screw and mechanics.  This is described here:
> >  
> > http://www.dynomotion.com/Help/ClosedLoopStep/ClosedLoopStepper.htm
> >  
> > Let us know when you have that part working.
> >  
> > Regards
> > TK
> >  
> >  
> >  
> >
> >
> > ________________________________
> > From: skyluvr <lovethebuzz69@>
> > To: DynoMotion@yahoogroups.com
> > Sent: Wednesday, November 30, 2011 11:40 AM
> > Subject: [DynoMotion] Motor tuning blues
> >
> >
> >  
> > Hi Folks,
> >
> > I have the motor tuning blues. I cannot seem to get the motors tuned on the x axis of my vertical mill. I am trying to tune a closed loop stepper system that is using the linear scales for feedback. I am starting to suspect that using the linear scales for feed back my be causing my problems. The mill has ball screws and I can detect no backlash with an indicator. Have others used the scales successfully? The mill currently runs Mach3 through the parallel port and works ok. I wanted to start using the Kflop to control the machine closed loop and get away from the mach driver.
> >
> > I am using mirco-step controllers at 2000 steps/rev on 200step/rev motors and the scales are .0005 resolution. I am using a gain of (2000mircostepsX200step motor=400000/8000=50. Is this right? At a gain of 50 I can not tune the motor. If I adjust it to say 11 I can match the command and the position closely. But at 50 there is no way.
> >
> > Any help would be greatly appreciated. Maybe a step by step on tuning somewhere?
> >
> > Thanks in advance
> >
> > Rick
> >
>