Skip to Content

Register Login

Industry Online Support

Technical Forum

Navigation

Syncronize two induction motors

Created by: SebOber at: 12/18/2012 6:10 PM (14 Replies)

Rating (0)

Thanks 0

Result pages:: |; 1 | 2; |

15 Entries

Entries per page: 10 | all

12/18/2012 6:10 PM	Rate (0)
SebOber Posts: 34 Rating: (3)	Hello all, I'm having a little difficulty with programming the controls of two induction motors on my machine. I have s7-1200, and two CU305's communicating via profibus. The motors connected to the CU's are the same motors with the same gearbox and have encoders on the motor side with a 45:1 gear ratio. I need someway of synchronizing the two motors. My best guess as how to do it would be to use a PID loop or a series of PID's. However, on top of syncronizing speed, I need to syncronize position also seeing as the two motors need to start, stop, and run at the same position. I have a prox sensor on each load for a "home" position. Any insight would be very appreciated! Thanks, -Seb
SebOber Posts: 34 Rating: (3)	Last edited by: SebOber at: 12/18/2012 6:24 PM
Suggestion To thank Quote Answer

12/19/2012 5:28 PM	Rate (0)
Marko Bursic Platinum Member Joined: 9/8/2009 Last visit: 7/17/2023 Posts: 1410 Rating: (153)	Not realy a nice job for s7-1200, but it worth to try. I would go with a simple trajectory planner outputing velocity and position like a master drive, then for each axis a regulator for position error compensation. I can give you a SCL code from Step 7 project I made (it was done for hydraulics). One important detail: are encoders mounted on inverters? Or you can realize a closed loop velocity control on inverters?

Suggestion To thank Quote Answer

12/19/2012 5:38 PM	Rate (0)
SebOber Posts: 34 Rating: (3)	Marko Bursic Not realy a nice job for s7-1200, but it worth to try. I would go with a simple trajectory planner outputing velocity and position like a master drive, then for each axis a regulator for position error compensation. I can give you a SCL code from Step 7 project I made (it was done for hydraulics). One important detail: are encoders mounted on inverters? Or you can realize a closed loop velocity control on inverters? Yes, the encoders are mounted on the motor side. So far, I have it worked out so that I will take the speed of the Main drive as setpoint in PID loop and Speed of the slave as the process value into a PID loop and take the output to the speed control of the slave. Then I have position monitoring which I have an equation worked out to get the positioning error, but need to somehow feed that into the output of the slave for that extra position correction. This is rather frustrating, I've never had to do something like this with this set-up. The product selection came straight from siemens and was made clear that these two drives need to be sync'ed. But there is no GOOD way to do it. But maybe there is a way to sync the drives in the Starter software? Also I'm worried that all this communication with the position and speed, and corrections, equations, is going to be to much for the s7-1200 to handle.
SebOber Posts: 34 Rating: (3)
Suggestion To thank Quote Answer

12/19/2012 5:53 PM	Rate (0)
0_Einstein Diamond Expert Joined: 7/7/2010 Last visit: 11/8/2024 Posts: 15388 Rating: (2443)	Can you use the power of the profibus network to communicate using slave-to-slave profibus cyclic communications and have the slave drive read the needed information directly from the master drive? That would make more sense than slowing things down with the plc in the middle.
	science guy
Suggestion To thank Quote Answer

12/19/2012 6:00 PM	Rate (0)
SebOber Posts: 34 Rating: (3)	huggy_d1 Can you use the power of the profibus network to communicate using slave-to-slave profibus cyclic communications and have the slave drive read the needed information directly from the master drive? That would make more sense than slowing things down with the plc in the middle. I would have no idea how to do that. This is the first time I've used profibus. I would usually use a D4xx system for this application, but it wasn't my choice here. Can you perhaps give me an example, or PDF to read into it. Thanks, -Seb
SebOber Posts: 34 Rating: (3)
Suggestion To thank Quote Answer

12/19/2012 6:08 PM	Rate (0)
0_Einstein Diamond Expert Joined: 7/7/2010 Last visit: 11/8/2024 Posts: 15388 Rating: (2443)	SebOber huggy_d1 Can you use the power of the profibus network to communicate using slave-to-slave profibus cyclic communications and have the slave drive read the needed information directly from the master drive? That would make more sense than slowing things down with the plc in the middle. I would have no idea how to do that. This is the first time I've used profibus. I would usually use a D4xx system for this application, but it wasn't my choice here. Can you perhaps give me an example, or PDF to read into it. Thanks, -Seb I usually punt back to Siemens via Support Requestwhen I cannot find the "how" within 5 minutes of searching their web site/forums/FAQs/Manuals/Getting Started Guides/etc. They said the hardware you have is sufficient to the task, so get them to explain "how". I recently modified a GSD/GSE file for a 3rd party torque feedback sensor so I could retransmit to a Siemens drive as actual torque feedback. If I knew the answer to your question, I could have the drive read directly from the torque sensor via profibus, but instead, I have the plc in the middle performing the retransmit. I will have to warn you about the s7-1200 & Siemens SIMATIC drive communications -- the experts are kind of like in knowledge silos. The plc people do not generally know much about how to do drive stuff, and the drive folks tend to only know legacy step 7 stuff. Over the last year, Siemens has been improving that situation, but do not be surprised if you get bounced around a bit and ultimately someone in Siemens HQ has to find the solution you need.
	science guy
Suggestion To thank Quote Answer

12/19/2012 6:22 PM	Rate (0)
SebOber Posts: 34 Rating: (3)	Thanks huggy, i'll take the route, if I find a solution I will post for others who may be going through the same.
SebOber Posts: 34 Rating: (3)
Suggestion To thank Quote Answer

12/19/2012 8:29 PM	Rate (0)
Marko Bursic Platinum Member Joined: 9/8/2009 Last visit: 7/17/2023 Posts: 1410 Rating: (153)	OK, I think it could be done, not very performant, but it is possible to try out. IMO trying to make a master-slave drive trough profibus is not the good idea. I thought to make a virtual master drive on S7-1200 and to output speed to others two. This can be done by making a trajectory planner of trapezoidal profile that outputs speed and position at each sample time, and you feedforfard the velocity, the drives will follow this trajectory - just for example if we integrate velocity we get position, and if drive is in closed velocity loop it will folow very close this trajectory, next is to slightly compensate for deviation from this trajectory - so called position loop, which is only a P-regulator. Now let wehave a trapezoidal trajectory planner (velocity, position), velocity is feedforfarded to the output. From traj. planner we get the position and with known feedback position we feed trough P-regulator. Now the output velocity is a sum of feedforwarded velocity and output from P-regulator.

Suggestion To thank Quote Answer

12/19/2012 8:59 PM	Rate (0)
Marko Bursic Platinum Member Joined: 9/8/2009 Last visit: 7/17/2023 Posts: 1410 Rating: (153)	ok, this is a one axis planner and p-reg, it uses position from encoder as input, look if some part is useful for you. p.s. : it was made in Step7 SCL, not TIA. Attachment trajplanner.pdf (411 Downloads)

Suggestion To thank Quote Answer

12/19/2012 10:46 PM	Rate (0)
SebOber Posts: 34 Rating: (3)	Thanks Marko, I appreciate the input. However, this application is different than what i need to accomplish. My two motors will always be moving. the point of finding the position is to keep the two motors running together at the same position and speed to accomplish the tasks these two axes are ment to do. The most difficult task is to move them at the same rate from standstill and to standstill. I will also be "homing" the drives at startup of the machine so the two drives won't necessarily start with an encoder feedback of '0'. So far, I figure I'll monitor the position feedback from the encoders pass them to a variable to '0' them out at start. then monitor the difference in position and take that value to adjust the speed of the slave motor to keep it in position. so I would be a very crude P loop. This should hopefully give me enough accuracy. It doesn't need to be 100% dead on. but they need to be at the proper position during the proper process.
SebOber Posts: 34 Rating: (3)
Suggestion To thank Quote Answer

|

Share this page:

Share this page on...

Follow us on