Dynomotion

Group: DynoMotion Message: 6211 From: superstepo Date: 12/3/2012
Subject: Backrunning after Jog
What is wrong, when the axis, after pressing and releasing a jogbutton, runs backwards a few millimeters? Its a driven backrunning, ADC changes the sign and gets a value about 200.

Thanks for any idea,

Stefan
Group: DynoMotion Message: 6214 From: TK Date: 12/3/2012
Subject: Re: Backrunning after Jog
Hi Stefan,

Do you see the same effect when doing a similar move on the Step Response Screen?  If so post a plot. What are all your axis settings?

Regards
TK

On Dec 3, 2012, at 11:34 AM, "superstepo" <waxmatch@...> wrote:

 

What is wrong, when the axis, after pressing and releasing a jogbutton, runs backwards a few millimeters? Its a driven backrunning, ADC changes the sign and gets a value about 200.

Thanks for any idea,

Stefan
Group: DynoMotion Message: 6216 From: superstepo Date: 12/4/2012
Subject: Re: Backrunning after Jog
Hello Tom,

The Problem is gone, but i don´t know why. On the StepResponse Screen it worked well everytime! To give you an overview, you will find the axis settings and a plot below. If the Problem comes again, I will tell you.



ch1->InputMode=ENCODER_MODE;
ch1->OutputMode=DAC_SERVO_MODE;
ch1->Vel=500000;
ch1->Accel=900000;
ch1->Jerk=7e+006;
ch1->P=1.5;
ch1->I=0.2;
ch1->D=80;
ch1->FFAccel=0;
ch1->FFVel=0;
ch1->MaxI=50;
ch1->MaxErr=1500;
ch1->MaxOutput=2000;
ch1->DeadBandGain=1;
ch1->DeadBandRange=0;
ch1->InputChan0=1;
ch1->InputChan1=0;
ch1->OutputChan0=1;
ch1->OutputChan1=0;
ch1->MasterAxis=-1;
ch1->LimitSwitchOptions=0x0;
ch1->InputGain0=1;
ch1->InputGain1=1;
ch1->InputOffset0=0;
ch1->InputOffset1=0;
ch1->OutputGain=-1;
ch1->OutputOffset=0;
ch1->SlaveGain=1;
ch1->BacklashMode=BACKLASH_OFF;
ch1->BacklashAmount=10;
ch1->BacklashRate=1000;
ch1->invDistPerCycle=0;
ch1->Lead=0;
ch1->MaxFollowingError=1500;
ch1->StepperAmplitude=20;

ch1->iir[0].B0=1;
ch1->iir[0].B1=0;
ch1->iir[0].B2=0;
ch1->iir[0].A1=0;
ch1->iir[0].A2=0;

ch1->iir[1].B0=1;
ch1->iir[1].B1=0;
ch1->iir[1].B2=0;
ch1->iir[1].A1=0;
ch1->iir[1].A2=0;

ch1->iir[2].B0=0.000769;
ch1->iir[2].B1=0.001538;
ch1->iir[2].B2=0.000769;
ch1->iir[2].A1=1.92076;
ch1->iir[2].A2=-0.923833;







--- In DynoMotion@yahoogroups.com, TK <tk@...> wrote:
>
> Hi Stefan,
>
> Do you see the same effect when doing a similar move on the Step Response Screen? If so post a plot. What are all your axis settings?
>
> Regards
> TK
>
> On Dec 3, 2012, at 11:34 AM, "superstepo" <waxmatch@...> wrote:
>
> > What is wrong, when the axis, after pressing and releasing a jogbutton, runs backwards a few millimeters? Its a driven backrunning, ADC changes the sign and gets a value about 200.
> >
> > Thanks for any idea,
> >
> > Stefan
> >
> >
>
Group: DynoMotion Message: 6217 From: Stefan Polatzky Date: 12/4/2012
Subject: Re: Backrunning after Jog
Hi Tom,

the Plot is added here!
Sorry, but i took the wrong Button at Yahoo.
I copied the whole mail.

Hello Tom,

The Problem is gone overnight, but i don´t know why. On the StepResponse Screen it worked well everytime! To give you an overview, you will find the axis settings and a plot below. If the Problem comes again, I will tell you.


ch1->InputMode=ENCODER_MODE;
ch1->OutputMode=DAC_SERVO_MODE;
ch1->Vel=500000;
ch1->Accel=900000;
ch1->Jerk=7e+006;
ch1->P=1.5;
ch1->I=0.2;
ch1->D=80;
ch1->FFAccel=0;
ch1->FFVel=0;
ch1->MaxI=50;
ch1->MaxErr=1500;
ch1->MaxOutput=2000;
ch1->DeadBandGain=1;
ch1->DeadBandRange=0;
ch1->InputChan0=1;
ch1->InputChan1=0;
ch1->OutputChan0=1;
ch1->OutputChan1=0;
ch1->MasterAxis=-1;
ch1->LimitSwitchOptions=0x0;
ch1->InputGain0=1;
ch1->InputGain1=1;
ch1->InputOffset0=0;
ch1->InputOffset1=0;
ch1->OutputGain=-1;
ch1->OutputOffset=0;
ch1->SlaveGain=1;
ch1->BacklashMode=BACKLASH_OFF;
ch1->BacklashAmount=10;
ch1->BacklashRate=1000;
ch1->invDistPerCycle=0;
ch1->Lead=0;
ch1->MaxFollowingError=1500;
ch1->StepperAmplitude=20;

ch1->iir[0].B0=1;
ch1->iir[0].B1=0;
ch1->iir[0].B2=0;
ch1->iir[0].A1=0;
ch1->iir[0].A2=0;

ch1->iir[1].B0=1;
ch1->iir[1].B1=0;
ch1->iir[1].B2=0;
ch1->iir[1].A1=0;
ch1->iir[1].A2=0;

ch1->iir[2].B0=0.000769;
ch1->iir[2].B1=0.001538;
ch1->iir[2].B2=0.000769;
ch1->iir[2].A1=1.92076;
ch1->iir[2].A2=-0.923833;




Thanks a lot,

Stefan


Stefan Polatzky, Mechanische Systeme, Leitenweg 8, 83543 Rott am Inn, Mobil: 0172/8477872, Tel/Fax: 08039/908297, waxmatch@...
Am 04.12.2012 08:37, schrieb TK:
 
Hi Stefan,

Do you see the same effect when doing a similar move on the Step Response Screen?  If so post a plot. What are all your axis settings?

Regards
TK

On Dec 3, 2012, at 11:34 AM, "superstepo" <waxmatch@...> wrote:

 

What is wrong, when the axis, after pressing and releasing a jogbutton, runs backwards a few millimeters? Its a driven backrunning, ADC changes the sign and gets a value about 200.

Thanks for any idea,

Stefan
Group: DynoMotion Message: 6219 From: Tom Kerekes Date: 12/4/2012
Subject: Re: Backrunning after Jog
Hi Stephan,

Here are a few comments on your plot:

Following error seems fairly high at ~ 100 counts (although you haven't included the resolution of your system).

Friction/Stiction seems very high.  At the end of a move the output remains high (100 counts) but there is no motion.  It appears to take an output of over 150 before any motion occurs.

The I gain of 0.2 seems very high.  Typical values are usually 0.001 ~ 0.1.   But you have a Max Integrator Limit of 50 counts.  The result is that the integrator quickly applies 50 counts and then saturates without enough range to achieve any motion and drive the error to zero.

Regards
TK

Group: DynoMotion Message: 6221 From: Stefan Polatzky Date: 12/4/2012
Subject: Re: Backrunning after Jog
Hi Tom,

thanks a lot for your comments! You are completly right! The friction is very high, thats because a large linear guide (i hope this the right name) is driven by a small motor. But for testing and learning its good enough, i think.

What do you mean with "although you haven't included the resolution of your system" ?
I suppose i have overlooked one detail!

Thanks a lot!

Stefan
Stefan Polatzky, Mechanische Systeme, Leitenweg 8, 83543 Rott am Inn, Mobil: 0172/8477872, Tel/Fax: 08039/908297, waxmatch@...
Am 04.12.2012 19:44, schrieb Tom Kerekes:
Hi Stephan,

Here are a few comments on your plot:

Following error seems fairly high at ~ 100 counts (although you haven't included the resolution of your system).

Friction/Stiction seems very high.  At the end of a move the output remains high (100 counts) but there is no motion.  It appears to take an output of over 150 before any motion occurs.

The I gain of 0.2 seems very high.  Typical values are usually 0.001 ~ 0.1.   But you have a Max Integrator Limit of 50 counts.  The result is that the integrator quickly applies 50 counts and then saturates without enough range to achieve any motion and drive the error to zero.

Regards
TK

Group: DynoMotion Message: 6222 From: Tom Kerekes Date: 12/4/2012
Subject: Re: Backrunning after Jog
Hi Stefan,

I was referring to the counts/inch of your system. Depending on your encoder resolution, gear ratios, lead screw pitch and so forth one encoder count will correspond to some linear distance on your table.  If you don't tell us this we don't know if an error of 100 counts is 0.1 inches or 0.001inches.  In KMotionCNC you set the resolution in Tool Setup | Trajectory Planner | Axis Parameters | cnts/inch

Regards
TK


Group: DynoMotion Message: 6223 From: Stefan Polatzky Date: 12/4/2012
Subject: Re: Backrunning after Jog
Hi Tom,

i thought i´ve something forgot to adjust but it was only to tell you! Sorry!
I have adjusted  a resolution of 101600 counts/inch to move the right distance
So a fault of 100 counts are roughly 1e-5 inches.

That brings me to another question:
Why the resolution i have to adjust, is 4 times higher then the real one?
Is there a 4 times evaluation per count in the controller?

For example:
I have a encoder with 4000 counts/rev and a leadscrew pitch of 0.157480 inches ( 4mm ), so i have a real resolution of 25400 counts/inch.
25400  *4  counts/inch  = 101600 counts/inch is what i have to adjust, as i described above.

Thanks a lot for your patience!

Best Regards,

Stefan



Stefan Polatzky, Mechanische Systeme, Leitenweg 8, 83543 Rott am Inn, Mobil: 0172/8477872, Tel/Fax: 08039/908297, waxmatch@...
Am 04.12.2012 20:46, schrieb Tom Kerekes:
Hi Stefan,

I was referring to the counts/inch of your system. Depending on your encoder resolution, gear ratios, lead screw pitch and so forth one encoder count will correspond to some linear distance on your table.  If you don't tell us this we don't know if an error of 100 counts is 0.1 inches or 0.001inches.  In KMotionCNC you set the resolution in Tool Setup | Trajectory Planner | Axis Parameters | cnts/inch

Regards
TK


Group: DynoMotion Message: 6224 From: Tom Kerekes Date: 12/4/2012
Subject: Re: Backrunning after Jog
Hi Stefan,

There is often confusion between encoder lines (or cycles) per rev vs encoder quadrature counts.   One line on an encoder generates 4 quadrature transitions of the A B signals.

It often amazes me that encoder manufactures often spec an encoder as 1000cpr with there being no way to be sure whether they mean cycles or counts.  Most of the time they mean cycles.

Regards
TK


Group: DynoMotion Message: 6225 From: Stefan Polatzky Date: 12/4/2012
Subject: Re: Backrunning after Jog
Hi Tom,

 your support only was it worth to buy the controller!

Thanks and have a nice Day.

Best Regards

Stefan
Stefan Polatzky, Mechanische Systeme, Leitenweg 8, 83543 Rott am Inn, Mobil: 0172/8477872, Tel/Fax: 08039/908297, waxmatch@...
Am 04.12.2012 22:27, schrieb Tom Kerekes:
Hi Stefan,

There is often confusion between encoder lines (or cycles) per rev vs encoder quadrature counts.   One line on an encoder generates 4 quadrature transitions of the A B signals.

It often amazes me that encoder manufactures often spec an encoder as 1000cpr with there being no way to be sure whether they mean cycles or counts.  Most of the time they mean cycles.

Regards
TK