Help..pneumatic valve with pid

[sakis]

hi there..
i am using siemens plc s7-313c..what's my problem??
i'm working on a project trying to control pneumatic valve with pid and levitate some objects...(airflow levitation)
i have a 4 sensors, each one helps me know exactly where my object is, so the valve should give more or less air...(object goes up or down )

analog inputs: PIW752-->sensor1 PIW754-->sensor2 PIW756-->sensor3 PIW758-->sensor4

analog output: PQW752-->valve

i try to use pid FB41 but...always something is wrong and nothing works..
please can u help me with all these parameters at FB41..

thanks ...

 

[leitmotif]

hi there..
i am using siemens plc s7-313c..what's my problem??
i'm working on a project trying to control pneumatic valve with pid and levitate some objects...(airflow levitation)
i have a 4 sensors, each one helps me know exactly where my object is, so the valve should give more or less air...(object goes up or down )

analog inputs: PIW752-->sensor1 PIW754-->sensor2 PIW756-->sensor3 PIW758-->sensor4

analog output: PQW752-->valve

i try to use pid FB41 but...always something is wrong and nothing works..
please can u help me with all these parameters at FB41..
thanks ...

what is total weight of the load you are moving?

First I see 4 sensors ie one per air flotation pad.
BUT I see only one control valve.
What are you sensing ie pressure or flow?
At what point are you measuring?

How are you going to individually control pressure or flow to each of the 4 pads with only one valve.

How is PID algorythm supposed to know which of the four sensors as an input
OR
are you averaging inpts and PID then uses average?

Seems to me you need 3 more control valves.

I also wonder if you should be controlling pressure or flow. As I understand it these units are sensitive to flow to keep suspended.
Dan Bentler

 

[Peter Nachtwey]

First you must understand what you are trying to control

I am sure you can levitate the object by just blowing air at it, no PLC is required. Controlling the elevation is another matter.

I also wonder if you should be controlling pressure or flow.

Hint, Newton didn't mention either pressure or flow in his laws of motion.

 

[leitmotif]

Peter

I am sure if we got 4 politicians we could surely save money and energy by not even needing a compressor. If we collected their male bovine excrement dave Emmerich could put it in his digestor.

Dan Bentler

 

[sakis]

PID uses average of 4 sensors..
if there is only one sensor..what shall i do???

this project is about controlling the elevation that is why i am using plc..

controlling pressure of the valve i levitate objects and then try to stable them in a certain point using the sensors..that is all..

 

[Peter Nachtwey]

PID uses average of 4 sensors..
if there is only one sensor..what shall i do???

Duh, use only the one sensor. If you have only two sensors you use the average of the two sensors. If you have three sensors then use the average of three sensors.

this project is about controlling the elevation that is why i am using plc.

So what is the relationship between air flow and the elevation? The PLC is just a tool.

controlling pressure of the valve i levitate objects and then try to stable them in a certain point using the sensors..that is all..

Yes, I understand but you must understand that the PLC part is the simple part. What if the elevation is linear? What are you really trying to control? I know you say elevation but what conditions must exist for an object to be stable at an elevation? When changing elevations you should ramp the elevation set point slowly. WHY?

Perhaps you instructor doesn't realize there is a lot more to this problem than just programming a PID.

 

[jawolthuis]

I think what you need to do is add the four sensor readings together. If we assume that your sensors are numbered 1-4 from bottom to top, then as the object is raised into the range of sensor 1 your reading will increase to its full scale as you pass the sensor , as the object reaches sensor 2 you set a bit that automatically adds the full value of sensor 1 to the value of sensor 2. Do the same thing with all four sensors and you will get a signal similar to a linear transducer. use this signal as your PV in your PID and set your SP accordingly.

 

[ndzied1]

Peter

I am sure if we got 4 politicians we could surely save money and energy by not even needing a compressor. If we collected their male bovine excrement dave Emmerich could put it in his digestor.

Dan Bentler

This reminds me so much of one of my favorite Dilbert cartoons:
http://www.flubu.com/comics/dilbert-thinklikeamanager.jpg

 

[Peter Nachtwey]

There is a lot more to this than using a PID

I am assuming that sakis trying to control the position of a sphere. I shouldn't have to assume that.

The key to controlling this is knowing how force changes as a function of height above the nozzle. The pdf is a pretty good explanation of relating the steady state flow required to maintain a relatively steady state but it doesn't say how to achieve it. Bernoulli's principle is what makes this possible.
http://www.hep.princeton.edu/~mcdonald/examples/beachball.pdf
Did you notice that the formula for the equilibrium point, equations 16 and 18, are not dependent on the air velocity?

See this
http://www.expertvillage.com/video/9860_bernoulli-theorem-ping-pong.htm

Did you notice that no matter how one blows one can't blow the ping pong ball out of the cup. So what difference would it make if the PLC told the valve to open more or less? The cup keeps the air flow around the ball that is closest to the air supply the highest. As the ball move farther from the air supply the pressure drops and the air on the opposite pushes it back. An equilibrium of forces is reached at a certain distance from the air supply.

http://www.californiasciencecenter.org/FunLab/DoItYourself/FloatingInAir/FloatingInAir.php

This is just to prove that a computer isn't required to maintain stability

I am not convinced that the PLC can control the elevation with flow. It looks like the stable distance is not dependent on flow or the velocity of the air. That would make attempting to control the elevation by PLC a futile task.

I am sure if we got 4 politicians we could surely save money and energy by not even needing a compressor. If we collected their male bovine excrement dave Emmerich could put it in his digestor.

The four politicians are irrelevant. All of congress couldn't generate enough hot air or foul gas to blow that ping pong ball out of that cup or change the stable height of the ball.

A video
http://www.youtube.com/watch?v=tUYPN6O5cCg

 

[leitmotif]

I thought he was trying to control the air skids used to move machinery.

Dan Bentler

 

[Peter Nachtwey]

That is the problem, students can't even state problems correctly

I thought he was trying to control the air skids used to move machinery.
Dan Bentler

sakis should have been more clear as to exactly what he is trying to do. We should have to guess. Normally I don't post on student questions but this one was crying out because it is a good example of why one needs to understand what they are controlling. No matter how tough you guys might think I am on the poor student, at least I am not letting sakis try, try, try and then fail to control his impossible system. Hopefully you guys can see why I wasn't interested in the sensor when I could have been beating up sakis about information on how the four sensors were arranged, their response times etc. The sensors aren't important to the basic problem.

Hopefully Derick123 sees this thread. Perhaps now he will understand what I am talking about. The PLC is just a tool.

 

[leitmotif]

sakis should have been more clear as to exactly what he is trying to do. We should have to guess. Normally I don't post on student questions but this one was crying out because it is a good example of why one needs to understand what they are controlling. No matter how tough you guys might think I am on the poor student, at least I am not letting sakis try, try, try and then fail to control his impossible system. Hopefully you guys can see why I wasn't interested in the sensor when I could have been beating up sakis about information on how the four sensors were arranged, their response times etc. The sensors aren't important to the basic problem.

Hopefully Derick123 sees this thread. Perhaps now he will understand what I am talking about. The PLC is just a tool.

Peter

I remember my chemistry instructor (German trained) who beat into our heads "understand the experiment before you walk thru the lab door". The trouble with being a student is it is hard to understand the experiment when you are trying to learn it.
I am sure there have been one or two instances where you went in blind and did not understand a thing but figured it out.

I tend to be forgiving of students.

What seems to be getting to me lately is people being paid to do the work but are not able to lay out a problem as well as some of the students do.

Dan Bentler

 

[Peter Nachtwey]

I tend to be forgiving of students.
Dan Bentler

As I said, I normally don't post to student threads. Most students are just trying to get a grade instead of really learning and the ones that do well have 'the knack.

The instructor should have known better or at least had doubts.

Honestly, I didn't know that the elevation was independent of flow but I did know that there was no chance of controlling the height without knowing the physics of the application. The two things that raised warning flags in my mind was a lack of damping and that the sysem is non-linear so I did an immediate google search and found the pdf file in about 5 minutes. Reading through that was enlightening and the videos seem to back it up. One could always get a high and low speed air dryer to see if there is a difference in the elevation.

My company gets involved with hundreds of application each year. Most applicatios work as expected but the 1% tend to be 'applications from hell'. The person setting up the motion controller is given an impossible task and it make no difference what the control engineer does because the mechanics aren't designed correctly and the system will never work.