ultra5000 cam window

mike_dr · Apr 18, 2013

Hi,I'm trying to use athe cam window(ultraware software) with the Ultra 5000 servo drive. I editted the cam table(specified input from 0 - 360, position and its velocity from 0 - 360),loaded it and enabled the cam. what would be the next step in simulating an encoder going through 0 to 360 to be followed by the servo? Is it going to be executed via C program or is there a way to easily see the servo following the cam profile? thank you

mike_dr · Apr 19, 2013

Additional question: anyone have experience in (clx processor) with controlling servo's torque positive and negative limits in a certain range (ex. Within 80 -120 deg) while MAPC is in process? There's a problem sometimes with drifting motors which I doubt is the case but I can't confirm.

mike_dr · Apr 22, 2013

anyone have experience using the loop configuration in the axis servo drive with Torque option? It's currently position servo in my program and changing it to torque will surely require a new cam profile if it is possible. What I am trying to solve is how to avoid physical drifting which causes the clamp(controlled by servo) to close too tightly. The servo apparently drifts while it is in the dwell region of the profile. Will changing the loop config help in avoiding the drifting?

MichaelG · Apr 22, 2013

what exactly are you trying to do?

If there are two issues then you will be better off with two threads.

It is easy to get the clx 5000 to maintain position with reduced torque
Use a GSV and change "Continuous Torque Limit"
don't forget to restore the value when you are in position control

mike_dr · Apr 22, 2013

MichaelG said:
what exactly are you trying to do?

If there are two issues then you will be better off with two threads.

It is easy to get the clx 5000 to maintain position with reduced torque
Use a GSV and change "Continuous Torque Limit"
don't forget to restore the value when you are in position control

thank you for replying. The machine is like a press which does the same thing every cycle based on the profile specified in MAPC. If it is in auto mode, for example from the 0-90 degrees encoder position, the servo should have torque limit of X but from 90-200 degrees it should dwell but for some reason it drifts or doesn't dwell so there is too much tightening in the clamp that the servo controls. that is the issue. So initially I thought of changing the torque limits when position is 90-200 more or less then change it back when it is BUT this is a continuous operation and I think Clx can't handle it and I dont think
that is the solution. So here I am, asking for your experiences regarding this situation

MichaelG · Apr 23, 2013

mike_dr said:
thank you for replying. The machine is like a press which does the same thing every cycle based on the profile specified in MAPC. If it is in auto mode, for example from the 0-90 degrees encoder position, the servo should have torque limit of X but from 90-200 degrees it should dwell but for some reason it drifts or doesn't dwell so there is too much tightening in the clamp that the servo controls. that is the issue. So initially I thought of changing the torque limits when position is 90-200 more or less then change it back when it is BUT this is a continuous operation and I think Clx can't handle it and I dont think
that is the solution. So here I am, asking for your experiences regarding this situation

I am still confused about how the machine operates
It appears that you have a Master Axis that goes 0 to 360 Deg (1 press cycle I guess)
and a slave axis that has ?? units and tightens a clamp in the first 90 deg of the press cycle and then releases in the last 45 deg of the cycle???

Can we see the CAM that you are using in the MAPC?

Have you a trend of one cycle with pens Master axis position, Slave axis position, slave torque, slave position error (need to set this up in the axis and group properties)?

How did you tune the Slave axis?

mike_dr · Apr 23, 2013

so the master axis is a virtual axis and it's just commanded with MAM. the servo is the slave axis and follows the profile supposedly.

the machine is already running for a year, so there is a possibility that the servo needs tuning right?

MichaelG · Apr 23, 2013

mike_dr said:
anyone have experience using the loop configuration in the axis servo drive with Torque option? It's currently position servo in my program and changing it to torque will surely require a new cam profile if it is possible.

When you change the servo loop configuration to torque you are taking responsibility for the motor velocity loop and position loop. Ie you must "roll your own" velocity and position control

So far I have not had the need to do this.

MichaelG · Apr 23, 2013

mike_dr said:
the machine is already running for a year, so there is a possibility that the servo needs tuning right?

If the initial install worked correctly and you have had no issues then something has changed. usually mechanically - eg bearings, gearbox, linkages.
Sometimes I have found that the maintenance department has replaced something from the original install with a similar part than is heaver or has a different inertia (eg greater radius).
Before re-tuning I try to find out what has happened first.
Other wise I tune to the degraded system and hide the real reason
This is not good if an imminent gearbox collapse is what is being covered up.

mike_dr · Apr 24, 2013

MichaelG said:
If the initial install worked correctly and you have had no issues then something has changed. usually mechanically - eg bearings, gearbox, linkages.
Sometimes I have found that the maintenance department has replaced something from the original install with a similar part than is heaver or has a different inertia (eg greater radius).
Before re-tuning I try to find out what has happened first.
Other wise I tune to the degraded system and hide the real reason
This is not good if an imminent gearbox collapse is what is being covered up.

Check the CAM
first slave entry = 0
First Master Entry = 0
Last Slave entry = 0 (so no incremental move)
Last Master move = 360 (one cycle)

based on the profile there should be no incremental move since it returns to 0 but in reality it is not happening.

MichaelG said:
When you change the servo loop configuration to torque you are taking responsibility for the motor velocity loop and position loop. Ie you must "roll your own" velocity and position control

So far I have not had the need to do this.

the cam profile(upon further configuration) already determines the position,velocity per cycle so that would suffice for the velocity and position control, am I right?

MichaelG · Apr 24, 2013

mike_dr said:
based on the profile there should be no incremental move since it returns to 0 but in reality it is not happening.

Does the servo return to position 0 or slowly creep up each cycle?
If the servo creeps up in position each cycle then it is in your programming

If the servo returns to position 0 every time but the mechanics shows that there is incremental move then mechanically there is something wrong
To test if this is occurring
1)Home the axis
2) Run the machine for many cycles
3) Rehome the machine - Note: how much the feedback changes. if the feedback changes a long way then you have mechanical slip

the cam profile(upon further configuration) already determines the position,velocity per cycle so that would suffice for the velocity and position control, am I right?

Not really - The cam profile determines the Position, velocity and accel REFERENCE you need to close the position and speed control loops using a servo in torque mode
look at the servo loop block diagrams in "MOTION-UM001" for more information

mike_dr · Apr 24, 2013

MichaelG said:
Does the servo return to position 0 or slowly creep up each cycle?
If the servo creeps up in position each cycle then it is in your programming

If the servo returns to position 0 every time but the mechanics shows that there is incremental move then mechanically there is something wrong
To test if this is occurring
1)Home the axis
2) Run the machine for many cycles
3) Rehome the machine - Note: how much the feedback changes. if the feedback changes a long way then you have mechanical slip

Not really - The cam profile determines the Position, velocity and accel REFERENCE you need to close the position and speed control loops using a servo in torque mode
look at the servo loop block diagrams in "MOTION-UM001" for more information

I will follow your advice and check whether this is mechanical slip or programming. I think I should've posted this in a mech's thread since it would be dealing with the mechanical slip but the responsibility is being passed onto software which is "easier to solve".

thank you for the advise and experience sharing MichaelG

Peter Nachtwey · Apr 24, 2013

This doesn't sound right. The slave executes a motion postion profile as the master goes from 0 to 360 degrees. The slave speed profile should be determined by the derivative of the position/degree and the degree per second of the master. This last part is the chain rule if you remember calculus. There should be an easy way to wrap from 0 to 359 and back to 0. Does the slave position the same at 0 and 360 degrees? Is the derivative the same? There should be a way of showing not only the slave position but the rate of change of the slave position on the cam window.

mike_dr · Apr 24, 2013

Peter Nachtwey said:
This doesn't sound right. The slave executes a motion postion profile as the master goes from 0 to 360 degrees. The slave speed profile should be determined by the derivative of the position/degree and the degree per second of the master. This last part is the chain rule if you remember calculus. There should be an easy way to wrap from 0 to 359 and back to 0. Does the slave position the same at 0 and 360 degrees? Is the derivative the same? There should be a way of showing not only the slave position but the rate of change of the slave position on the cam window.

if this is referring to the cam window using ultraware, I am still lost after finding no resource on how to simulate the master(virtual axis or just an int moving 0 to 360) or slave(servo) movement using cam table. I just copied values from the logix program to the table in ultraware. and yes both master and slave have 0-360 degree units

mike_dr · Apr 25, 2013

Does the slave position the same at 0 and 360 degrees?

yes, it goes back to 0 degrees when master hits 360 degrees in the cam profile.

I dropped the ultraware cam window simulation because I realized that the cam profile wouldn't be the same for both application since I have to convert the degree values in the rslogix into count units which is what ultraware cam window reads. It was just a sidequest on figuring how to simulate cam using ultraware which is possible.

ultra5000 cam window

mike_dr

Member

mike_dr

Member

mike_dr

Member

MichaelG

Member

mike_dr

Member

MichaelG

Member

mike_dr

Member

MichaelG

Member

MichaelG

Member

mike_dr

Member

MichaelG

Member

mike_dr

Member

Peter Nachtwey

Member

mike_dr

Member

mike_dr

Member

Similar Topics