--- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@...> wrote:
>
> Hi Michael,
>  
> I don't know you will need to do some diagnostic work.  What does the Plot show?  Anything printed on the Console Screen?  What does the Axis Screen show?  What is the final condition?  Did you test your Amplifier using DAC commands?
>  
> Regards
> TK
>
> From: mrosenfield2 <mrosenfield@...>
> To: DynoMotion@yahoogroups.com
> Sent: Thursday, March 1, 2012 9:10 AM
> Subject: [DynoMotion] Re: servo tuning
>
>
>  
> OK, things are moving, sort of.
> I am playing with P = 1.0 - 1.5, I = 0, D =0.
> V = 33,333; A=333,333; J=3,333,333.
> Time is set to 3, and size is 4000, which is 0.2"
> When I click move, the axis moves, but it never moves back. It is also not moving anywhere near 0.2"; more like 0.02".
> What am I missing?
> Thanks,
> Michael
>
> --- In DynoMotion@yahoogroups.com, "mrosenfield2" <mrosenfield@> wrote:
> >
> > I will try the new settings.
> > The servo amps are AC powered, and have an 8-amp current-limit set, so there is no way their power draw is affecting the KFlop.
> >
> > Michael
> >
> > --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> > >
> > >
> > >
> > > Hi Michael,
> > >  
> > > I don't know why KMotion would be crashing, obviously that shouldn't or doesn't normally happen.  The parameters are normally saved when you exit KMotion.exe, but unfortunately on a crash that doesn't happen.  You can force a save using File | Save All.
> > >  
> > > Your P gain is too high.   The Units are in DAC counts per Encoder Error Count.   So only ~2 error counts would be required to put a Max 2047 DAC counts (full 10V).
> > >  
> > > A number more like P=1 would be a starting point.
> > >  
> > > There is a default motor configuration for Kanalog called KanalogInitialPID.mot.  Load that as a starting point.  It will set the Config Screen, Filter Screen, and Step Response Screens to reasonable starting values.  Of course change the InputChan0 and OutputChan0 for the encoder and DAC you are using.
> > >  
> > > It is better to use the "Move" button rather than the "Step" button.  A sudden step of 4000 counts generates an instant error of 4000 counts which combined with your P gain of 1000 would attempt to command 4,000,000 DAC counts (heh heh). 
> > >  
> > > What might be happening is that the instant full power command to the amplifier is drawing so much current that KFLOP is somehow being effected.  With proper ground wiring and so forth this should happen.  But I'm surprised you don't hear a bang from the motor. 
> > >  
> > > Regards
> > > TK
> > >  
> > > From: Michael Rosenfield <mrosenfield@>
> > > To: dynomotion@yahoogroups.com
> > > Sent: Monday, February 27, 2012 6:25 PM
> > > Subject: [DynoMotion] servo tuning
> > >
> > >
> > >  
> > > Tom,
> > > I have started servo tuning; so far things aren't working.
> > > I am using the KFlop/Kanalog combination, with AMC +/-10V amps, running in velocity mode, and 1000-count encoders.
> > > That is 33,333 counts/sec at the max speed of 100 IPM
> > > I have tested the analog input to the amps, and they are working correctly.
> > > 1. I set up the axis configuration and downloaded the channel. I saved the axis to a file, too. I assume that each axis gets its own file?
> > > 2. And I assume I must download each axis each time I power up the Kflop (until I get the C init program done)?
> > > 3. Then I went to the step response screen, and set the following settings:
> > >  
> > > P=1000 I = 0 D = 0
> > > 33,333V
> > > 33.333 A
> > > 33.0000J
> > > dead band was left at default
> > > step was 3 secs, 4000 counts. That is a 0.2" move
> > > I clicked on Servo enable, and the axis locked.
> > > Then I clicked on Step, and nothing happened. No graph, either.
> > > Then I clicked on step again, and the program locked up, and I got the windows error box, and had to restart KMotion.
> > > When I got back to the Step response screen, half the parameters I had to reset; after I did, I got the same software lockup as before.
> > >  
> > > 4.Is there a way to have the step response screen remember all the values?
> > > 5. Any idea why KMotion is crashing? I assume I'm doing something wrong.
> > > 6. I got no plots at all - does this make sense?
> > >  
> > > Thanks,
> > > Michael Rosenfield
> > >  
> > >  
> > >
> >
>

 

TK,
I tested the amp using a pot for +/-10V input.
Commanded motion looks essentially ok; commanded position does not.
Is there a way to check encoder counts from the console?

Thanks,
Michael

--- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@...> wrote:
>
> Hi Michael,
>  
> I don't know you will need to do some diagnostic work.  What does the Plot show?  Anything printed on the Console Screen?  What does the Axis Screen show?  What is the final condition?  Did you test your Amplifier using DAC commands?
>  
> Regards
> TK
>
> From: mrosenfield2 <mrosenfield@...>
> To: DynoMotion@yahoogroups.com
> Sent: Thursday, March 1, 2012 9:10 AM
> Subject: [DynoMotion] Re: servo tuning
>
>
>  
> OK, things are moving, sort of.
> I am playing with P = 1.0 - 1.5, I = 0, D =0.
> V = 33,333; A=333,333; J=3,333,333.
> Time is set to 3, and size is 4000, which is 0.2"
> When I click move, the axis moves, but it never moves back. It is also not moving anywhere near 0.2"; more like 0.02".
> What am I missing?
> Thanks,
> Michael
>
> --- In DynoMotion@yahoogroups.com, "mrosenfield2" <mrosenfield@> wrote:
> >
> > I will try the new settings.
> > The servo amps are AC powered, and have an 8-amp current-limit set, so there is no way their power draw is affecting the KFlop.
> >
> > Michael
> >
> > --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> > >
> > >
> > >
> > > Hi Michael,
> > >  
> > > I don't know why KMotion would be crashing, obviously that shouldn't or doesn't normally happen.  The parameters are normally saved when you exit KMotion.exe, but unfortunately on a crash that doesn't happen.  You can force a save using File | Save All.
> > >  
> > > Your P gain is too high.   The Units are in DAC counts per Encoder Error Count.   So only ~2 error counts would be required to put a Max 2047 DAC counts (full 10V).
> > >  
> > > A number more like P=1 would be a starting point.
> > >  
> > > There is a default motor configuration for Kanalog called KanalogInitialPID.mot.  Load that as a starting point.  It will set the Config Screen, Filter Screen, and Step Response Screens to reasonable starting values.  Of course change the InputChan0 and OutputChan0 for the encoder and DAC you are using.
> > >  
> > > It is better to use the "Move" button rather than the "Step" button.  A sudden step of 4000 counts generates an instant error of 4000 counts which combined with your P gain of 1000 would attempt to command 4,000,000 DAC counts (heh heh). 
> > >  
> > > What might be happening is that the instant full power command to the amplifier is drawing so much current that KFLOP is somehow being effected.  With proper ground wiring and so forth this should happen.  But I'm surprised you don't hear a bang from the motor. 
> > >  
> > > Regards
> > > TK
> > >  
> > > From: Michael Rosenfield <mrosenfield@>
> > > To: dynomotion@yahoogroups.com
> > > Sent: Monday, February 27, 2012 6:25 PM
> > > Subject: [DynoMotion] servo tuning
> > >
> > >
> > >  
> > > Tom,
> > > I have started servo tuning; so far things aren't working.
> > > I am using the KFlop/Kanalog combination, with AMC +/-10V amps, running in velocity mode, and 1000-count encoders.
> > > That is 33,333 counts/sec at the max speed of 100 IPM
> > > I have tested the analog input to the amps, and they are working correctly.
> > > 1. I set up the axis configuration and downloaded the channel. I saved the axis to a file, too. I assume that each axis gets its own file?
> > > 2. And I assume I must download each axis each time I power up the Kflop (until I get the C init program done)?
> > > 3. Then I went to the step response screen, and set the following settings:
> > >  
> > > P=1000 I = 0 D = 0
> > > 33,333V
> > > 33.333 A
> > > 33.0000J
> > > dead band was left at default
> > > step was 3 secs, 4000 counts. That is a 0.2" move
> > > I clicked on Servo enable, and the axis locked.
> > > Then I clicked on Step, and nothing happened. No graph, either.
> > > Then I clicked on step again, and the program locked up, and I got the windows error box, and had to restart KMotion.
> > > When I got back to the Step response screen, half the parameters I had to reset; after I did, I got the same software lockup as before.
> > >  
> > > 4.Is there a way to have the step response screen remember all the values?
> > > 5. Any idea why KMotion is crashing? I assume I'm doing something wrong.
> > > 6. I got no plots at all - does this make sense?
> > >  
> > > Thanks,
> > > Michael Rosenfield
> > >  
> > >  
> > >
> >
>

--- In DynoMotion@yahoogroups.com, TK <tk@...> wrote:
>
> Michael,
>
> Look at the Axis Screen under Position.
>
> TK
>
> On Mar 4, 2012, at 1:10 PM, "mrosenfield2" <mrosenfield@...> wrote:
>
> > TK,
> > I tested the amp using a pot for +/-10V input.
> > Commanded motion looks essentially ok; commanded position does not.
> > Is there a way to check encoder counts from the console?
> >
> > Thanks,
> > Michael
> >
> > --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> > >
> > > Hi Michael,
> > > Â
> > > I don't know you will need to do some diagnostic work. What does the Plot show? Anything printed on the Console Screen? What does the Axis Screen show? What is the final condition? Did you test your Amplifier using DAC commands?
> > > Â
> > > Regards
> > > TK
> > >
> > > From: mrosenfield2 <mrosenfield@>
> > > To: DynoMotion@yahoogroups.com
> > > Sent: Thursday, March 1, 2012 9:10 AM
> > > Subject: [DynoMotion] Re: servo tuning
> > >
> > >
> > > Â
> > > OK, things are moving, sort of.
> > > I am playing with P = 1.0 - 1.5, I = 0, D =0.
> > > V = 33,333; A=333,333; J=3,333,333.
> > > Time is set to 3, and size is 4000, which is 0.2"
> > > When I click move, the axis moves, but it never moves back. It is also not moving anywhere near 0.2"; more like 0.02".
> > > What am I missing?
> > > Thanks,
> > > Michael
> > >
> > > --- In DynoMotion@yahoogroups.com, "mrosenfield2" <mrosenfield@> wrote:
> > > >
> > > > I will try the new settings.
> > > > The servo amps are AC powered, and have an 8-amp current-limit set, so there is no way their power draw is affecting the KFlop.
> > > >
> > > > Michael
> > > >
> > > > --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> > > > >
> > > > >
> > > > >
> > > > > Hi Michael,
> > > > > ÂÂ
> > > > > I don't know why KMotion would be crashing, obviously that shouldn't or doesn't normally happen. The parameters are normally saved when you exit KMotion.exe, but unfortunately on a crash that doesn't happen. You can force a save using File | Save All.
> > > > > ÂÂ
> > > > > Your P gain is too high.  The Units are in DAC counts per Encoder Error Count.  So only ~2 error counts would be required to put a Max 2047 DAC counts (full 10V).
> > > > > ÂÂ
> > > > > A number more like P=1 would be a starting point.
> > > > > ÂÂ
> > > > > There is a default motor configuration for Kanalog called KanalogInitialPID.mot. Load that as a starting point. It will set the Config Screen, Filter Screen, and Step Response Screens to reasonable starting values. Of course change the InputChan0 and OutputChan0 for the encoder and DAC you are using.
> > > > > ÂÂ
> > > > > It is better to use the "Move" button rather than the "Step" button. A sudden step of 4000 counts generates an instant error of 4000 counts which combined with your P gain of 1000 would attempt to command 4,000,000 DAC counts (heh heh).ÂÂ
> > > > > ÂÂ
> > > > > What might be happening is that the instant full power command to the amplifier is drawing so much current that KFLOP is somehow being effected. With proper ground wiring and so forth this should happen. But I'm surprised you don't hear a bang from the motor.ÂÂ
> > > > > ÂÂ
> > > > > Regards
> > > > > TK
> > > > > ÂÂ
> > > > > From: Michael Rosenfield <mrosenfield@>
> > > > > To: dynomotion@yahoogroups.com
> > > > > Sent: Monday, February 27, 2012 6:25 PM
> > > > > Subject: [DynoMotion] servo tuning
> > > > >
> > > > >
> > > > > ÂÂ
> > > > > Tom,
> > > > > I have started servo tuning; so far things aren't working.
> > > > > I am using the KFlop/Kanalog combination, with AMC +/-10V amps, running in velocity mode, and 1000-count encoders.
> > > > > That is 33,333 counts/sec at the max speed of 100 IPM
> > > > > I have tested the analog input to the amps, and they are working correctly.
> > > > > 1. I set up the axis configuration and downloaded the channel. I saved the axis to a file, too. I assume that each axis gets its own file?
> > > > > 2. And I assume I must download each axis each time I power up the Kflop (until I get the C init program done)?
> > > > > 3. Then I went to the step response screen, and set the following settings:
> > > > > ÂÂ
> > > > > P=1000 I = 0 D = 0
> > > > > 33,333V
> > > > > 33.333 A
> > > > > 33.0000J
> > > > > dead band was left at default
> > > > > step was 3 secs, 4000 counts. That is a 0.2" move
> > > > > I clicked on Servo enable, and the axis locked.
> > > > > Then I clicked on Step, and nothing happened. No graph, either.
> > > > > Then I clicked on step again, and the program locked up, and I got the windows error box, and had to restart KMotion.
> > > > > When I got back to the Step response screen, half the parameters I had to reset; after I did, I got the same software lockup as before.
> > > > > ÂÂ
> > > > > 4.Is there a way to have the step response screen remember all the values?
> > > > > 5. Any idea why KMotion is crashing? I assume I'm doing something wrong.
> > > > > 6. I got no plots at all - does this make sense?
> > > > > ÂÂ
> > > > > Thanks,
> > > > > Michael Rosenfield
> > > > > ÂÂ
> > > > > ÂÂ
> > > > >
> > > >
> > >
> >
> >
>

--- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@...> wrote:
>
> Hi Michael,
>  
> The initial gain settings and max output is probably very low which is why the servo is very loose and weak.  But it it should have feedback.
>  
> There are some servo tuning threads on cnczone.com.
>  
> http://www.cnczone.com/forums/dynomotion_kflop_kanalog/142000-optimizing_servo_tuning.html
>  
> http://www.cnczone.com/forums/dynomotion_kflop_kanalog/142770-servo_tuning_snapamp.html
>  
> Regards
> TK
>
>
> ________________________________
> From: mrosenfield2 <mrosenfield@...>
> To: DynoMotion@yahoogroups.com
> Sent: Sunday, March 4, 2012 3:25 PM
> Subject: [DynoMotion] Re: servo tuning
>
>
>
>  
>
> Found it! Also found I was getting following errors, which shut the axis down. That explains the motion!
>
> I increased the FE to a large value, and now I'm getting wild axis oscillations at something like 0.3 Hz. They do damp out, so I guess I'm headed in the right direction. I know I've seen threads on axis tuning here somewhere, though, so I'm looking.
>
> When no move is commanded, running KMotion, are the axes supposed to be locked in position? Mine can be moved pretty easily by hand.
>
> I also discovered that after a Move command, it takes a second or three for the graphs to appear. If another Move is issued before the graphs appear, KMotion hangs, and I have to close it, sometimes with Windows Task Manager, and restart the app. Comments?
>
> Thanks,
> Michael
>
> --- In DynoMotion@yahoogroups.com, TK <tk@> wrote:
> >
> > Michael,
> >
> > Look at the Axis Screen under Position.
> >
> > TK
> >
> > On Mar 4, 2012, at 1:10 PM, "mrosenfield2" <mrosenfield@> wrote:
> >
> > > TK,
> > > I tested the amp using a pot for +/-10V input.
> > > Commanded motion looks essentially ok; commanded position does not.
> > > Is there a way to check encoder counts from the console?
> > >
> > > Thanks,
> > > Michael
> > >
> > > --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> > > >
> > > > Hi Michael,
> > > > Â
> > > > I don't know you will need to do some diagnostic work. What does the Plot show? Anything printed on the Console Screen? What does the Axis Screen show? What is the final condition? Did you test your Amplifier using DAC commands?
> > > > Â
> > > > Regards
> > > > TK
> > > >
> > > > From: mrosenfield2 <mrosenfield@>
> > > > To: DynoMotion@yahoogroups.com
> > > > Sent: Thursday, March 1, 2012 9:10 AM
> > > > Subject: [DynoMotion] Re: servo tuning
> > > >
> > > >
> > > > Â
> > > > OK, things are moving, sort of.
> > > > I am playing with P = 1.0 - 1.5, I = 0, D =0.
> > > > V = 33,333; A=333,333; J=3,333,333.
> > > > Time is set to 3, and size is 4000, which is 0.2"
> > > > When I click move, the axis moves, but it never moves back. It is also not moving anywhere near 0.2"; more like 0.02".
> > > > What am I missing?
> > > > Thanks,
> > > > Michael
> > > >
> > > > --- In DynoMotion@yahoogroups.com, "mrosenfield2" <mrosenfield@> wrote:
> > > > >
> > > > > I will try the new settings.
> > > > > The servo amps are AC powered, and have an 8-amp current-limit set, so there is no way their power draw is affecting the KFlop.
> > > > >
> > > > > Michael
> > > > >
> > > > > --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> > > > > >
> > > > > >
> > > > > >
> > > > > > Hi Michael,
> > > > > > ÃÆ'‚Â
> > > > > > I don't know why KMotion would be crashing, obviously that shouldn't or doesn't normally happen.ÃÆ'‚Â The parameters are normally saved when you exit KMotion.exe, but unfortunately on a crash that doesn't happen.ÃÆ'‚Â You can force a save using File | Save All.
> > > > > > ÃÆ'‚Â
> > > > > > Your P gain is too high.ÃÆ'‚Â ÃÆ'‚Â The Units are inÃÆ'‚Â DAC counts per Encoder Error Count.ÃÆ'‚Â ÃÆ'‚Â So onlyÃÆ'‚Â ~2 error counts would be requiredÃÆ'‚Â to put a MaxÃÆ'‚Â 2047 DAC counts (full 10V).
> > > > > > ÃÆ'‚Â
> > > > > > A number more like P=1 would be a starting point.
> > > > > > ÃÆ'‚Â
> > > > > > There is a default motor configuration for Kanalog called KanalogInitialPID.mot.ÃÆ'‚Â Load that as a starting point.ÃÆ'‚Â It will set the Config Screen, Filter Screen, and Step Response Screens to reasonable starting values.ÃÆ'‚Â Of course change the InputChan0 and OutputChan0 for the encoder and DAC you are using.
> > > > > > ÃÆ'‚Â
> > > > > > It is better to use the "Move" button rather than the "Step" button.ÃÆ'‚Â A sudden step of 4000 counts generates an instant error of 4000 counts which combined with your P gain of 1000 would attempt to command 4,000,000 DAC counts (heh heh).ÃÆ'‚Â
> > > > > > ÃÆ'‚Â
> > > > > > What might be happening is that the instant full power command to the amplifier is drawing so much current that KFLOP is somehow being effected.ÃÆ'‚Â With proper ground wiring and so forth this should happen.ÃÆ'‚Â But I'm surprised youÃÆ'‚Â don't hear aÃÆ'‚Â bang from the motor.ÃÆ'‚Â
> > > > > > ÃÆ'‚Â
> > > > > > Regards
> > > > > > TK
> > > > > > ÃÆ'‚Â
> > > > > > From: Michael Rosenfield <mrosenfield@>
> > > > > > To: dynomotion@yahoogroups.com
> > > > > > Sent: Monday, February 27, 2012 6:25 PM
> > > > > > Subject: [DynoMotion] servo tuning
> > > > > >
> > > > > >
> > > > > > ÃÆ'‚Â
> > > > > > Tom,
> > > > > > I have started servo tuning; so far things aren't working.
> > > > > > I am using the KFlop/Kanalog combination, with AMC +/-10V amps, running in velocity mode, and 1000-count encoders.
> > > > > > That is 33,333 counts/sec at the max speed of 100 IPM
> > > > > > I have tested the analog input to the amps, and they are working correctly.
> > > > > > 1. I set up the axis configuration and downloaded the channel. I saved the axis to a file, too. I assume that each axis gets its own file?
> > > > > > 2. And I assume I must download each axis each time I power up the Kflop (until I get the C init program done)?
> > > > > > 3. Then I went to the step response screen, and set the following settings:
> > > > > > ÃÆ'‚Â
> > > > > > P=1000 I = 0 D = 0
> > > > > > 33,333V
> > > > > > 33.333 A
> > > > > > 33.0000J
> > > > > > dead band was left at default
> > > > > > step was 3 secs, 4000 counts. That is a 0.2" move
> > > > > > I clicked on Servo enable, and the axis locked.
> > > > > > Then I clicked on Step, and nothing happened. No graph, either.
> > > > > > Then I clicked on step again, and the program locked up, and I got the windows error box, and had to restart KMotion.
> > > > > > When I got back to the Step response screen, half the parameters I had to reset; after I did, I got the same software lockup as before.
> > > > > > ÃÆ'‚Â
> > > > > > 4.Is there a way to have the step response screen remember all the values?
> > > > > > 5. Any idea why KMotion is crashing? I assume I'm doing something wrong.
> > > > > > 6. I got no plots at all - does this make sense?
> > > > > > ÃÆ'‚Â
> > > > > > Thanks,
> > > > > > Michael Rosenfield
> > > > > > ÃÆ'‚Â
> > > > > > ÃÆ'‚Â
> > > > > >
> > > > >
> > > >
> > >
> > >
> >
>

--- In DynoMotion@yahoogroups.com, "mrosenfield2" <mrosenfield@...> wrote:
>
> TK,
> Do you know of any threads that deal with tuning a velocity servo?
> I have made some progress, but still aways to go.
> Current gain: P - 14.5, I - 0.0001 D - 1650. The speed is 40000, and the accel is 60000. Jerk is 1e6. The axis is stiffer now, but I'm getting FE of 200 counts with each edge of the move. If I add any filtering, the axis tends to oscillate with the second move.
> I must be missing something with the Bode plot - the command plot is mostly off the screen (negative) and there is no listing of gain or phase margin at the bottom of the plot.
> Any suggestions?
> Thanks,
> Michael
>
> --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> >
> > Hi Michael,
> >  
> > The initial gain settings and max output is probably very low which is why the servo is very loose and weak.  But it it should have feedback.
> >  
> > There are some servo tuning threads on cnczone.com.
> >  
> > http://www.cnczone.com/forums/dynomotion_kflop_kanalog/142000-optimizing_servo_tuning.html
> >  
> > http://www.cnczone.com/forums/dynomotion_kflop_kanalog/142770-servo_tuning_snapamp.html
> >  
> > Regards
> > TK
> >
> >
> > ________________________________
> > From: mrosenfield2 <mrosenfield@>
> > To: DynoMotion@yahoogroups.com
> > Sent: Sunday, March 4, 2012 3:25 PM
> > Subject: [DynoMotion] Re: servo tuning
> >
> >
> >
> >  
> >
> > Found it! Also found I was getting following errors, which shut the axis down. That explains the motion!
> >
> > I increased the FE to a large value, and now I'm getting wild axis oscillations at something like 0.3 Hz. They do damp out, so I guess I'm headed in the right direction. I know I've seen threads on axis tuning here somewhere, though, so I'm looking.
> >
> > When no move is commanded, running KMotion, are the axes supposed to be locked in position? Mine can be moved pretty easily by hand.
> >
> > I also discovered that after a Move command, it takes a second or three for the graphs to appear. If another Move is issued before the graphs appear, KMotion hangs, and I have to close it, sometimes with Windows Task Manager, and restart the app. Comments?
> >
> > Thanks,
> > Michael
> >
> > --- In DynoMotion@yahoogroups.com, TK <tk@> wrote:
> > >
> > > Michael,
> > >
> > > Look at the Axis Screen under Position.
> > >
> > > TK
> > >
> > > On Mar 4, 2012, at 1:10 PM, "mrosenfield2" <mrosenfield@> wrote:
> > >
> > > > TK,
> > > > I tested the amp using a pot for +/-10V input.
> > > > Commanded motion looks essentially ok; commanded position does not.
> > > > Is there a way to check encoder counts from the console?
> > > >
> > > > Thanks,
> > > > Michael
> > > >
> > > > --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> > > > >
> > > > > Hi Michael,
> > > > > Â
> > > > > I don't know you will need to do some diagnostic work. What does the Plot show? Anything printed on the Console Screen? What does the Axis Screen show? What is the final condition? Did you test your Amplifier using DAC commands?
> > > > > Â
> > > > > Regards
> > > > > TK
> > > > >
> > > > > From: mrosenfield2 <mrosenfield@>
> > > > > To: DynoMotion@yahoogroups.com
> > > > > Sent: Thursday, March 1, 2012 9:10 AM
> > > > > Subject: [DynoMotion] Re: servo tuning
> > > > >
> > > > >
> > > > > Â
> > > > > OK, things are moving, sort of.
> > > > > I am playing with P = 1.0 - 1.5, I = 0, D =0.
> > > > > V = 33,333; A=333,333; J=3,333,333.
> > > > > Time is set to 3, and size is 4000, which is 0.2"
> > > > > When I click move, the axis moves, but it never moves back. It is also not moving anywhere near 0.2"; more like 0.02".
> > > > > What am I missing?
> > > > > Thanks,
> > > > > Michael
> > > > >
> > > > > --- In DynoMotion@yahoogroups.com, "mrosenfield2" <mrosenfield@> wrote:
> > > > > >
> > > > > > I will try the new settings.
> > > > > > The servo amps are AC powered, and have an 8-amp current-limit set, so there is no way their power draw is affecting the KFlop.
> > > > > >
> > > > > > Michael
> > > > > >
> > > > > > --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > Hi Michael,
> > > > > > > ÃÆ'‚Â
> > > > > > > I don't know why KMotion would be crashing, obviously that shouldn't or doesn't normally happen.ÃÆ'‚Â The parameters are normally saved when you exit KMotion.exe, but unfortunately on a crash that doesn't happen.ÃÆ'‚Â You can force a save using File | Save All.
> > > > > > > ÃÆ'‚Â
> > > > > > > Your P gain is too high.ÃÆ'‚Â ÃÆ'‚Â The Units are inÃÆ'‚Â DAC counts per Encoder Error Count.ÃÆ'‚Â ÃÆ'‚Â So onlyÃÆ'‚Â ~2 error counts would be requiredÃÆ'‚Â to put a MaxÃÆ'‚Â 2047 DAC counts (full 10V).
> > > > > > > ÃÆ'‚Â
> > > > > > > A number more like P=1 would be a starting point.
> > > > > > > ÃÆ'‚Â
> > > > > > > There is a default motor configuration for Kanalog called KanalogInitialPID.mot.ÃÆ'‚Â Load that as a starting point.ÃÆ'‚Â It will set the Config Screen, Filter Screen, and Step Response Screens to reasonable starting values.ÃÆ'‚Â Of course change the InputChan0 and OutputChan0 for the encoder and DAC you are using.
> > > > > > > ÃÆ'‚Â
> > > > > > > It is better to use the "Move" button rather than the "Step" button.ÃÆ'‚Â A sudden step of 4000 counts generates an instant error of 4000 counts which combined with your P gain of 1000 would attempt to command 4,000,000 DAC counts (heh heh).ÃÆ'‚Â
> > > > > > > ÃÆ'‚Â
> > > > > > > What might be happening is that the instant full power command to the amplifier is drawing so much current that KFLOP is somehow being effected.ÃÆ'‚Â With proper ground wiring and so forth this should happen.ÃÆ'‚Â But I'm surprised youÃÆ'‚Â don't hear aÃÆ'‚Â bang from the motor.ÃÆ'‚Â
> > > > > > > ÃÆ'‚Â
> > > > > > > Regards
> > > > > > > TK
> > > > > > > ÃÆ'‚Â
> > > > > > > From: Michael Rosenfield <mrosenfield@>
> > > > > > > To: dynomotion@yahoogroups.com
> > > > > > > Sent: Monday, February 27, 2012 6:25 PM
> > > > > > > Subject: [DynoMotion] servo tuning
> > > > > > >
> > > > > > >
> > > > > > > ÃÆ'‚Â
> > > > > > > Tom,
> > > > > > > I have started servo tuning; so far things aren't working.
> > > > > > > I am using the KFlop/Kanalog combination, with AMC +/-10V amps, running in velocity mode, and 1000-count encoders.
> > > > > > > That is 33,333 counts/sec at the max speed of 100 IPM
> > > > > > > I have tested the analog input to the amps, and they are working correctly.
> > > > > > > 1. I set up the axis configuration and downloaded the channel. I saved the axis to a file, too. I assume that each axis gets its own file?
> > > > > > > 2. And I assume I must download each axis each time I power up the Kflop (until I get the C init program done)?
> > > > > > > 3. Then I went to the step response screen, and set the following settings:
> > > > > > > ÃÆ'‚Â
> > > > > > > P=1000 I = 0 D = 0
> > > > > > > 33,333V
> > > > > > > 33.333 A
> > > > > > > 33.0000J
> > > > > > > dead band was left at default
> > > > > > > step was 3 secs, 4000 counts. That is a 0.2" move
> > > > > > > I clicked on Servo enable, and the axis locked.
> > > > > > > Then I clicked on Step, and nothing happened. No graph, either.
> > > > > > > Then I clicked on step again, and the program locked up, and I got the windows error box, and had to restart KMotion.
> > > > > > > When I got back to the Step response screen, half the parameters I had to reset; after I did, I got the same software lockup as before.
> > > > > > > ÃÆ'‚Â
> > > > > > > 4.Is there a way to have the step response screen remember all the values?
> > > > > > > 5. Any idea why KMotion is crashing? I assume I'm doing something wrong.
> > > > > > > 6. I got no plots at all - does this make sense?
> > > > > > > ÃÆ'‚Â
> > > > > > > Thanks,
> > > > > > > Michael Rosenfield
> > > > > > > ÃÆ'‚Â
> > > > > > > ÃÆ'‚Â
> > > > > > >
> > > > > >
> > > > >
> > > >
> > > >
> > >
> >
>