torque control ,kinco motor,plc

hello everyone !
i have kinco's motors ( if anyone has already worked with one of these it will be very helpfull ) and i try to apply torque control giving signals through a PLC.
My control system is like this http://arvc.umh.es/parola/lesson1_archives/five_r.png , where in points A1 ,A2 i have my motors and they move the l1 and l4 .

I have already searched in here for some similar threads but my concern is that my load as you can see isn't stable as the robot moves from right to left and reverse .

My question is : I would like to be able to control or set a certain velocity ,so that i would be able to achieve a movement from right to left and reverse the exact same way back and again ( something like oscillation ). How can i be able to do this using torque control?

I am not using velocity control since it is necessary my robot to have a stable movement .When i tried this mode the robot didn't move in a stable way (it did some ups and downs that i don't want ). My experience says that it needed more torque -to be more powerfull in torque- .

Thank you in advance

Stavrina Deve said:

hello everyone !
i have kinco's motors ( if anyone has already worked with one of these it will be very helpfull ) and i try to apply torque control giving signals through a PLC.
My control system is like this http://arvc.umh.es/parola/lesson1_archives/five_r.png , where in points A1 ,A2 i have my motors and they move the l1 and l4 .

I have already searched in here for some similar threads but my concern is that my load as you can see isn't stable as the robot moves from right to left and reverse .

Click to expand...

I would use position or angle control. The controller gains will change as the angles change. That math will be difficult to work out but just making a table may be good enough.

My question is : I would like to be able to control or set a certain velocity ,so that i would be able to achieve a movement from right to left and reverse the exact same way back and again ( something like oscillation ). How can i be able to do this using torque control?

Click to expand...

This is an inverse kinematics problem. You need to calculate the angle and rate of change of the angle for theta 1 and theta 2 as a function of X and Y.

I am not using velocity control since it is necessary my robot to have a stable movement .When i tried this mode the robot didn't move in a stable way (it did some ups and downs that i don't want ). My experience says that it needed more torque -to be more powerfull in torque- .

Thank you in advance

Click to expand...

You may need more torque if your system can't hold position for all X,Y but you still should be using position control.

This must be a student problem

The good thing is that one should be able to calculate closed form solution instead of an iterative one for this application.

thank you peter for your replay! the thing is that i already have the equations-the inverse kinematics- and already tried the position control . The problem with that was : the kinco motors i have, for the position control they use pulses.I tried to send these pulses with the PLC but the frequency i needed was impossible for the Plc to produse.

My aim is to mimic the motion of Delta robots (they are fast but they reach their position precisely, and as far as the torque they will "insist" to reach their position that can be dangerous if a hand tries to stop them ). This was the reason i thought maybe the torque control is better for my application . I hope you understand the way I am writing and your opinion is very welcome!

Stavrina Deve said:

thank you peter for your replay! the thing is that i already have the equations-the inverse kinematics- and already tried the position control . The problem with that was : the kinco motors i have, for the position control they use pulses.I tried to send these pulses with the PLC but the frequency i needed was impossible for the Plc to produse.

My aim is to mimic the motion of Delta robots (they are fast but they reach their position precisely, and as far as the torque they will "insist" to reach their position that can be dangerous if a hand tries to stop them ). This was the reason i thought maybe the torque control is better for my application . I hope you understand the way I am writing and your opinion is very welcome!

Click to expand...

You need a motion control card in your PLC or externally connected by Ethernet. Controlling your system with a PLC will be impossible to very difficult.

Are you talking about controlling the inner torque loop? How can you do that with pulses? You could do it with analog signals and it would be better than using pulses but controlling a torque loop requires a fast update time. You must do the math so the torque controls the acceleration and that is integrated to provide velocity and position. This is still a position control problem. The problem I see is that you don't have a motion controller that can do your application.

PLCs are too slow. So is their I/O
PLCs do not have a way of sampling the positions at regular intervals. Sampling during interrupts is not good enough.
Doing the inverse kinematics in a PLC must be painful. It would be OK using SCL or structured text.
PLCs do not have the trending ability necessary for these applications. PLCs cannot sample fast enough.

Peter Nachtwey said:

You need a motion control card in your PLC or externally connected by Ethernet. Controlling your system with a PLC will be impossible to very difficult.

Are you talking about controlling the inner torque loop? How can you do that with pulses? You could do it with analog signals and it would be better than using pulses but controlling a torque loop requires a fast update time. You must do the math so the torque controls the acceleration and that is integrated to provide velocity and position. This is still a position control problem. The problem I see is that you don't have a motion controller that can do your application.

PLCs are too slow. So is their I/O
PLCs do not have a way of sampling the positions at regular intervals. Sampling during interrupts is not good enough.
Doing the inverse kinematics in a PLC must be painful. It would be OK using SCL or structured text.
PLCs do not have the trending ability necessary for these applications. PLCs cannot sample fast enough.

Click to expand...

I agree that inverse kinematics with PLC will be for sure painful and that's why I want to try connecting my PLC with matlab (I have seen some interesting ideas on the internet but I haven't worked on it yet). At first i wanted to figure out which mode that my drivers-motors offer will do for my work .
1) there is the pulse mode where i can control the degrees of the motor with pulses ,as I told you before.

2)there is the speed mode where i give the driver a voltage and this is translated to speed (speed and volts are proportional for this one )

3)there is the torque mode where again volts are proportional to torque this time .And current has a role here which still i am tring to figure out ...

As i said before only with the torque mode i saw the desired motion (not the position , i mean the robust motion of the robot).

So if i understand you suggest that the problem is the limits of the PLC and that my life will be much easier if i could find the right motion controller?

I am attaching a photo of the inner driver servo system