PLC vs Motion Controller

Join Date
Apr 2002
Location
No income tax, no capital gains tax. Freedom!
Posts
8,351
As you guys know I sell hydraulic motion controllers. We keep getting the question "why should I spend all that money on a hydraulic motion controller"
At the same time Alan Hitchcock, the Hydraulics and Pneumatics editor, thought that this would make for a good article. So here it is. I tried not to be biased.

https://www.hydraulicspneumatics.com/controls-instrumentation/plc-vs-motion-controllers

Any comments? There is one sentence that might grab peoples attention.

BTW, I am not done yet. There will be a part 2 where I go into using PI, PD and PID control and the need for a target generator.

I also play to have a part 3 where we actually implement a hydraulic motion controller in a PLC. We have already done this on a S7-1500. I hope to also compare with a CompactLogix.

BTW, thanks to Norm Dziedzic I got an early warning the Fig 2 and Fig 3 were exchanged so we got that corrected.
 
Last edited:
I know the sentence that grabbed my attention. I thought it was funny.

I generally like it. You definitely don't take the leap of saying EVERY hydraulic application needs a motion controller. You pretty clearly state the limitations of the easy proportional-only control method and you set up well for the next installment. As you have said before, writing something like this for a mixed audience is always difficult because of the widely varying degree of experience in the readers. I think you cut a good compromise with this.

One thing I would say is the equations need some parentheses. I had a pretty good idea of what I was looking at and I still got a little lost. Also, the text seems to indicate the simulations are at the end of the article but the simulation graphs are scattered through the article and they don't necessarily appear in the areas of the text that refer to what they show. But that could just be peculiar to the online version.

Keith
 
I like the idea of having a dedicated Hydraulic controller. I've used the Bosch Rexroth HACD controllers over the years but my issue is always integration and connectivity. If someone has a product that you can connect to and get data in and out of Then all the better.
 
I like the idea of having a dedicated Hydraulic controller. I've used the Bosch Rexroth HACD controllers over the years but my issue is always integration and connectivity. If someone has a product that you can connect to and get data in and out of Then all the better.

Delta systems has a hydraulic controller that interfaces well with Allen Bradley products over Ethernet IP.

https://deltamotion.com/
 
One thing I would say is the equations need some parentheses.
This will be fixed soon. The problem is that I used the Microsoft equation editor to write the equation in a .doc file. What I am realizing is that one the equation is in the .doc file it is useless to the type setter or web page designer at H&P because the pictures and equations are converted to pixels. A person at H&P usually remakes the pictures but this takes time and is obviously error proone. My python hydraulic cylinder simulator generates and .svg file. I am looking at using an external equation editor that will generate .svg files.

I had a pretty good idea of what I was looking at and I still got a little lost. Also, the text seems to indicate the simulations are at the end of the article but the simulation graphs are scattered through the article and they don't necessarily appear in the areas of the text that refer to what they show. But that could just be peculiar to the online version.

Keith
Yes, I sent the .doc file with the graphs interspersed with the text.
 
I like the idea of having a dedicated Hydraulic controller. I've used the Bosch Rexroth HACD controllers over the years but my issue is always integration and connectivity. If someone has a product that you can connect to and get data in and out of Then all the better.
I have an HACD attached to a 19 inch rack by my desk. I know the limitations.

We support many different PLC protocols. Our controller is really a multi protocol Ethernet controller that just happens to do hydraulic motion control. The amount of code dedicated to Ethernet is huge! You can even use our controller like a Redlion Data Station. You can write to the controllers variable table using one protocol and read the data up using a different Ethernet protocol.

I will be curious to read this and how bias you are. Whether IRT is used and a TO... Bus cycle times etc...
Siemens tech support told us that we should be using a Simotion controller to control a hydraulic cylinder. The article is about doing all in the PLC vs using a motion controller.
The S7-1500 by itself did not do well.
We are going to try the CompactLogix. If using a ControlLogix is it smart to use a HYD02 or M02AS module.
I have lots of different PLCs here for compatibilty testing.

I have yet to get down and dirty and compare control algorithms and target generator programs.
 
Peter
What does “did not do well” mean?
How was it set up?
Back it up with some details
We have what we call our basic hydraulic system. It is easy to control. Disconnected our controller. We changed out the SSI MDT rod for an analog MDT rod so we could use analog inputs and output from the PLC.
Here is a video of our basic hydraulic system
You can see the load moving in the picture in picture on the left.
https://deltamotion.com/peter/Videos/AutoTuning3.mp4
I was attempted speak some Chinese in parts of it.
Many forum members have been trained on this system.
You should notice that we have auto tuning for hydraulic cylinders and that the following error between the target and actual position is extremely small after I use our tuning procedure.

There are three ways to use an S7-1500 to controller a hydraulic cylinder.
One is to use the motion blocks. The fastest loop time was 4 milliseconds. The motion block only allowed for a proportional only control but it generated a position target. This results in lots of following error. Again, we were told the S7-1500 and the motion blocks were not meant to be used the way were were using it which is understandable. We were suppose to be using the S7-1500 with Simotion controllers. Load the motion block up and see for yourself. If you know something we don't then tell us.

The second way is to use the built in PID. Honestly, that isn't very good either. The built in PID is meant for temperature control systems or slow stuff. Then a target generator still must be used even if it is as crude as the one in the article.

I personally would write the closed loop control from scratch instead of using the built in motion blocks if I was only using the S7-1500.

I/we are not done yet. That article will not come out for another two months or so.

It is not my intention to bash PLCs. It is just that few people really know what it takes to write a motion control program and to do it for only one or even a hand full of cylinders is not worth the time. It is cheaper to buy the controller and get better speed and precision for the life of the machine. That alone will pay for the motion controller many times over.
 
Why can't you do motion control with a standard plc?
It's not fast enough.

Can you do motion control with a fast plc? Yes, but this would make this article obsolete.

Another example:
http://elm-chan.org/works/ltc/report.html
Can you program a small line follower robot with a standard plc? No, it's not fast enough. But a 8 Bit Microcontroller can do this.
 
Originally posted by Peter Nachtwey:

We are going to try the CompactLogix.

Good luck with that. As you said with the S7-1500 case, you would be better off writing your own control loop than using anything canned in the CompactLogix. The one benefit you get with the CompactLogix is you get all the motion commands so you get the profile generator and virtual axes. You just don't get the loop closure stuff.

And then there is getting position information into the plc on a timely basis. That isn't a guarantee at the rates you will be looking for. And if you come up with a decent way to get a stable digital differential out of that position information (much less a double differential) let me know.


Originally posted by Peter Nachtwey:

It is not my intention to bash PLCs. It is just that few people really know what it takes to write a motion control program and to do it for only one or even a hand full of cylinders is not worth the time.

We all need to take off our favorite plc sweaters for a little bit and let the information take us where it leads us. Some tools are better than others at specific tasks. I'm just as stubborn as the next guy and will try to jam that square peg unto a round hole for weeks on end. But, like most, I also apply too small a value on my time; I see it as a sunk cost. This is really where using a specific tool for a given job really pays for itself. The are many applications you CAN do with a plc...well enough...given enough time. It may just make more sense using a dedicated tool instead of expending the effort.

Keith
 
Originally posted by Thomas_v2:

Can you do motion control with a fast plc? Yes, but this would make this article obsolete.

Are you a Beckhoff or a B&R guy? Those are the only two mainstream "PLC" platforms I know of with the combination of processor and peripherals that can really pull off high bandwidth motion without a dedicated coprocessor. But that isn't necessarily the whole story either. I imagine there is probably a Joe's Garage brand plc out there that can also keep up that i just don't know about. But if i can't get it where i am it doesn't help me much I suppose.

Keith
 
I don’t have access to the article, but based on the reaction, I think I know what the hot topic is.

The idea that because you bought a four-wheel drive truck, you can’t get stuck. The reality is that the four-wheel drive truck will only get you stuck much further down the trail.

Unless the truck is fitted or designed for a particular task, it generally will only have mediocre performance in all other tasks.

The dedicated motion controller will always surpass the performance of a PLC.
I think the bigger issue abroad is figuring out if you need a stock 4x4 or a specially modified 4x4 for the trail ahead.


Are you a Beckhoff or a B&R guy? Those are the only two mainstream "PLC" platforms I know of with the combination of processor and peripherals that can really pull off high bandwidth motion without a dedicated coprocessor.
Keith

I was about to mention B&R as they do have many canned function blocks for controlling hydraulics. Along with profile generation, MIMO, and others. I personally would like to see one of their PLCs tested against a dedicated motion controller out of curiosity.
 
Another example:
http://elm-chan.org/works/ltc/report.html
Can you program a small line follower robot with a standard plc? No, it's not fast enough. But a 8 Bit Microcontroller can do this.
That is a poor example for many reasons. The main one is that the cars are only following a line. They are not generating the motion path like a motion controller must do.
I can poke more holes in that article like why don't they use an integrator. I know the real reason.

Can you specify the precise S7-1500 model?


e.g. S7-1511, s7-1517T, S7-1518....
I don't know for sure. I can find out on Monday. Assume it is the S7-1511.
I can make allowances for the speed but the motion blocks are a joke.
One thing I want to do is test the Siemens target generator. This is the most difficult part of a motion controller to write that actually does motion. The Ethernet code is a killer too.

The one benefit you get with the CompactLogix is you get all the motion commands so you get the profile generator and virtual axes. You just don't get the loop closure stuff.
I want to test the target generators too.

And then there is getting position information into the plc on a timely basis. That isn't a guarantee at the rates you will be looking for. And if you come up with a decent way to get a stable digital differential out of that position information (much less a double differential) let me know.

It may just make more sense using a dedicated tool instead of expending the effort.
You are very right but would anybody even think of that if they weren't an experienced motion control guy? Look at all the threads on this forum trying to filter a velocity.
Do you realize we/our sales force gets asked this question all the time? In China they always want to save a Yuan and will attempt writing a motion controller in a PLC.

I don’t have access to the article, but based on the reaction, I think I know what the hot topic is.
I don't think anybody does. The article was taken down to fix the parentheses problem. It will be back up Monday for sure. The good thing is that these problems were found before printing.

The dedicated motion controller will always surpass the performance of a PLC.
That is obvious but what if it is a 'simple' job?

This isn't just motion control. It is hydraulic motion control. A hydraulic cylinder can be modeled as a mass between two springs. The response is under damped and will natrually tend to oscillate a bit. A motor has the just the inertia of the motor and load. Controlling a motor is much easier.

I have a test system that was designed to be uncontrollable, unless moving very slow, using a PID with velocity and acceleration feed forwards. Think about controlling a slinky.
 

Similar Topics

Hi Where can I find B&r 2005 plc - NC154 motion controller Manual and Software? Thanks in advance.
Replies
1
Views
1,431
Hi all. It's been a while since I last posted here :-) Looking for an advice on handling command/feedback mismatch during transitions. I have...
Replies
0
Views
1,012
A coworker of mine keeps telling me that the PacDrive3 (Elau) is a PC based Motion Controller while the ControlLogix is a "real" PLC. I pointed...
Replies
18
Views
12,918
Hello all of you I have a research about (Design and implementation of a fuzzy motion controller based on PLC) But I have a confuse... How...
Replies
2
Views
4,756
Hello Experts, I'm wondering if this has been done before if possible to create an Emulate file that have motion control axis? I tried to...
Replies
0
Views
130
Back
Top Bottom