Skip to Content

Register Login

Industry Online Support

Technical Forum

Navigation

?> (S7-1200) Closing the loop on a PTO

Created by: William_B at: 11/12/2011 4:43 PM (34 Replies)

Rating (3)

Thanks 3

Result pages:: |; 1 | 2 | 3 | 4; |

35 Entries

Entries per page: 10 | 20 | all

11/12/2011 4:43 PM	Rate (0)
William_B Posts: 5225 Rating: (1196)	Hi all, It is a nagging question I am yet to resolve, maybe some help please? See also this LINK which inspired me to put his to paper. The PLC generates a pulse train (PTO) that is used as a speed setpoint (ramp up, constant, ramp down). The PLC receives a pulse from an encoder that measure the actual position of the moving object. The motion generating device gives it speed as an analog (0..10V) signal. The accuracy of the system to follow the speed setpoint (PTO signal) can't be assured. The accurate following of the speed is not at issue. The system should reach an internal setpoint for position - again the accuracy does not need to be 100%. Examples of such systems include.. 1. Stepper motor that gives slip at ramp up. 2. Cylinder positioning to open/close valves. 3. AC motor driven by an inverter (VSD). 4. DC motor driven by a (old generation) DC drive. 5. Upgrade of control on old generation motion control. In the above cases the open loop control is just not good enough. "Put a servo in there" is not the answer I'm looking for. Any suggestions on the control strategy? Greetings. -- William B.
William_B Posts: 5225 Rating: (1196)
Suggestion To thank Quote Answer

11/14/2011 6:55 PM	Rate (0)
Denilson Pegaia Platinum Member Joined: 9/23/2005 Last visit: 11/14/2024 Posts: 4352 Rating: (1452)	Hi, once you recive a position feedback from the encoder, the garanty of the driver follow the speed setpoint is secundary. If you receve position feedback from a encoder and send setpoints as PTO, you can use technological motion funcions (OPEN PLC Motion blocks).
	Denilson Pegaia Siemens Industry Sector Technical Support
Suggestion To thank Quote Answer

11/15/2011 8:33 AM	Rate (0)
William_B Posts: 5225 Rating: (1196)	Pegaia Hi, once you recive a position feedback from the encoder, the garanty of the driver follow the speed setpoint is secundary. If you receve position feedback from a encoder and send setpoints as PTO, you can use technological motion funcions (OPEN PLC Motion blocks). Pegaia, thank you for your reply. Please forgive me if I have difficulty to understand how this reply answers my query. The PLC motion blocks are open loop control. What is required is closed loop control. The best I can think of is using the MC_Velocity block to initiate motion. There is an internal position setpoint which is evaluated against the measured position. The error between setpoint and measurement should control MC_Velocity. In my mind a PI control block should be used to give MC_Velocity a speed setpoint. In theory this looks sound, but I am not sure how to ensure predictable ramp up and ramp down. The control loop should also not have overshoot. Am I on the right track or am I blowing up smoke?
William_B Posts: 5225 Rating: (1196)
Suggestion To thank Quote Answer

11/15/2011 7:56 PM	Rate (0)
Denilson Pegaia Platinum Member Joined: 9/23/2005 Last visit: 11/14/2024 Posts: 4352 Rating: (1452)	William B. Pegaia Hi, once you recive a position feedback from the encoder, the garanty of the driver follow the speed setpoint is secundary. If you receve position feedback from a encoder and send setpoints as PTO, you can use technological motion funcions (OPEN PLC Motion blocks). Pegaia, thank you for your reply. Please forgive me if I have difficulty to understand how this reply answers my query. The PLC motion blocks are open loop control. What is required is closed loop control. The best I can think of is using the MC_Velocity block to initiate motion. There is an internal position setpoint which is evaluated against the measured position. The error between setpoint and measurement should control MC_Velocity. In my mind a PI control block should be used to give MC_Velocity a speed setpoint. In theory this looks sound, but I am not sure how to ensure predictable ramp up and ramp down. The control loop should also not have overshoot. Am I on the right track or am I blowing up smoke? Hi, I think that you are in the rigth way. I suggest you: position value-> PI (or just P controler) -> limit (if necessary) -> speed setpoint forMC_Velocity ... maybe you should also place a turn off (a hysteresis)the control, if it is near enought.
	Last edited by: Pegaia at: 11/15/2011 8:21 PM Denilson Pegaia Siemens Industry Sector Technical Support
Suggestion To thank Quote Answer

11/16/2011 1:47 PM	Rate (0)
Marko Bursic Platinum Member Joined: 9/8/2009 Last visit: 7/17/2023 Posts: 1410 Rating: (153)	hi, I was thinking about the same approach for one my project. the problem is that you would needa PI regulator with a ramp limiting. I was thinking about the use of closed loop stepper motor control, but i have found that stepper motor is very weird to control it, when it stalls it stalls, so it needs to start again from zero. I read much literature also on other forums for CNC, etc.. ther are a lot of aproaches how to put a stepper motor in a closed loop, but it seems that no one is 100% except some custom controllers with dsp procesors that are able to detect stalling,....this is for previous posts when i was saying that stepper motor is working in open loop only, and if it stalls you can do much. Another discussion is how to write your PI controller with ramp limiting, because if you put the output of regulator on MC_Velocity input, it is probable that regulator will windup, since MC_Velocity has a ramp generator. Maybe it is better to use P-regulator only in order to avoid windup?

Suggestion To thank Quote Answer

11/16/2011 6:32 PM	Rate (0)
simone.m Posts: 4 Rating: (0)	Hi Marko, there are some manufacturer tha can supply full digital stepping motor drive which can handle the stepper as servomotor with a true closed loop control. With these drives motor never stalls because it works like a brushless servo motor and reacts to load changes. I've tried a couple of these drive.... It's a very cool approach to stepping motor.
simone.m Posts: 4 Rating: (0)
Suggestion To thank Quote Answer

11/16/2011 9:17 PM	Rate (0)
Marko Bursic Platinum Member Joined: 9/8/2009 Last visit: 7/17/2023 Posts: 1410 Rating: (153)	Yes, I know, that's why I abandoned the idea of making a closed loop control of stepper motor by means of PLC. However if it possible to make it with simatic it would be great - I mean the whole application on robust an easy to find replace parts makes a proffesional end product. I see that William has the same ideas as I havehad, I gave up, but who knows maybe he has right and I am ready to share my experience, maybe it would help him.

Suggestion To thank Quote Answer

11/17/2011 8:17 AM	Rate (0)
William_B Posts: 5225 Rating: (1196)	Thank you Marko Bursic and simone.m for your interest and taking part in this discusion. There is a group that developed coding for Siemens using the S7-200 that does positioning control for 3-phase AC motor and variable speed drive. You can download the locked code blocks from the web or buy a system as a kit from Siemens. It would have been great to have had such control for the S7-1200. Personally I would utilise digital control all the way; not an analog output for speed control as the S7-200 code does. But then again the number output of the block could be an input to the pulse block. The control concept goes much deeper as stepper motors. With the right programming you can control anything which accept a pulse for speed control and where you can measure its position. In most cases micrometer precision is not required. There are cases now that I move my PLC code to the position controller to try and avoid having a position controller and PLC. It would be nice if the reverse could happen; moving the position control back into the PLC. It would better the servicability of the system. Now you know the motivation. Any theories on how to put positioning into the hands of the S7-1200? Greetings, -- William
William_B Posts: 5225 Rating: (1196)
Suggestion To thank Quote Answer

11/17/2011 11:20 AM	Rate (0)
Marko Bursic Platinum Member Joined: 9/8/2009 Last visit: 7/17/2023 Posts: 1410 Rating: (153)	OK. In my opinion there is a big difference for controlling stepper motor and AC drive, because AC drive adds slip when it is loaded, but still the system is linear, so linear algebra (regulation loops) is the approach for such problems. Meanwhile stepper motor is a nonlinear system - it follows pulses until it can hold a torque, then suddenly stalls, if you rise the frequency (that would happen with AC drive controller) you did even worth thing, because you should lower the frequency at thestarting frequency in order that magnetic field hooks a rotor again. Another thing is using PTO pulses, for AC drive is not applicable, since it needs analog speed or torque command value. Both systems if you need positioning starts with trajectory planner. You want to move from pointA to point Bwithcertain velocity and acceleration/deceleration.This is simple if you can write in some high level language, but in PLC is quite difficult. So lets use system functions that are available on PLC and with some trick get working in your way. For example lets take this stepper motor with closed loop control. My approach would be like this: diff_position:=setpoint_position-target_position; // error_pos:=output_pos_from_Technology_funct - encoder_position if (very_small_error<abs(error_pos)<small_error) then //some lost step diff_pos:=encoder_pos-target_pos; retriger MC_Move_relative(diff_pos); //bumpless transfer look in manual Exe,Exe2 elsif abs(errorpos)>small_error // big error, motor stalled abort motion; //set frequency to start frequency, not possible- no such command MC_move_relative(diff_pos); // start againmanual else // very small error mc_move_relative(diff_pos); end_if;

Suggestion To thank Quote Answer

11/17/2011 1:04 PM	Rate (0)
William_B Posts: 5225 Rating: (1196)	Marko Bursic, Thank you for your response. There are things I like and things I do not like -- using relative motion is inspired but can give problems. I am going to chew on what you have written and respond in more detail on a later post. General notes now. Marko Bursic Meanwhile stepper motor is a nonlinear system - it follows pulses until it can hold a torque, then suddenly stalls, if you rise the frequency (that would happen with AC drive controller) you did even worse thing, because you should lower the frequency at thestarting frequency in order that magnetic field hooks a rotor again. There is a critical point I did not consider too much -- stalling of a stepper motor. But then this is true even for AC motors if the (start-up) torque needed overwhelms the torque given. If it is at start-up, in most cases one can overcome the problem by (a) stop (b) increase the ramping time (c) restart. Stalling while running constant speed is more difficult to get rid of. I think however we should assume that the motor is driven correctly and does not stall - as to simplify the model as to add the exceptions later. Marko Bursic Another thing is using PTO pulses, for AC drive is not applicable, since it needs analog speed or torque command value. The AC drive I use can take in a frequency pulse or analog signal for the speed command. But maybe the best general solution would be to have a number output (example 0..10,000). To take this number and convert it to pulses or analog output is not too difficult. The solution is then general and mathematical. Marko Bursic Both systems if you need positioning starts with trajectory planner. You want to move from pointA to point Bwithcertain velocity and acceleration/deceleration. The idea I toyed with was taking over the task of the deceleration too. This needs the MC_Velocity block to work. But I need more info on how the MC_Velocity works inside as to predict if my idea will succeed or not.. it's got to do with the transfer from one motion to another Marko Bursic This is simple if you can write in some high level language, but in PLC is quite difficult. Agreed that PLC code will be difficult to make discussion with. Pseudo high-level language will do just fine. I'll comment on your code piece later on with some of my thoughts. ..unless someone else feels like participating (which you are welcome to). Greetings.
William_B Posts: 5225 Rating: (1196)	Last edited by: William B. at: 11/17/2011 1:08 PM
Suggestion To thank Quote Answer

Result pages:: |; 1 | 2 | 3 | 4; |

35 Entries

Entries per page: 10 | 20 | all

|

Share this page:

Share this page on...

Related topics by keywords:

Follow us on