I want to configure a closed loop with three inputs and one output, have you got any suggestions?
This output interfeers all of three inputs.

 

A closed loop system can be built with as little as one input and one output.

The output is the controlled signal and the input is the feedback. These could be simple static in and outs, or PWM I/Os or Frequency Modulated or analog.

What does "interfeers" mean? Are you saying a change in output may cause changes in some or all of the inputs?

All that happens in the closed loop is that a change in one or more of your feedback inputs will (might) precipitate a change in your controlled output.

Is this a software solution or are you designing a circuit?

Rufus

Rufus V. Smith
Software/Hardware Design (esp. Automation)
Recently available for full, part-time, or contract work.
Home Page: http://members.aol.com/rufusvs
Looking for work? Check my Job Search Links.

 

Could you provide a little more information about how the process wants the three inputs to be related? For example, should the total of all three be held to some setpoint? Or Input 1 - Input 2 + (2 * Input 3) be controlled?

If you follow me, I'm trying to say that you must mathematically define the desired relationship between the three process variables so that you can dynamically calculate a single number from the three which will be the PV for your control loop. Only then can the loop know if the process is at the desired operating condition, and if not, which way to adjust the output.

 

A change in the output can affect one or more of the 3 inputs. I am using a Siemens PLC S7-400 and my firt idea is to configure it with a PID and take the average value of the 3 inputs.

 

The 3 inputs have the same setpoint but they will not reach the setpoint at the same time. The sequence of inputs that will reach the setpoint it is different every time.

Gfleg
[email protected]

 

If you won't take the effort to specify the problem completely, you should not expect any useful responses. This happens way too often on this board.

That said,
Are the inputs using identical instruments?
Are they calibrated the same way?
Are they located in the same location?
What really are the process / control objectives?

If the 3 are not measuring the same thing, you need to prioritize the 3 somehow; if priorities change, you may need to schedule the control structure you use or use constraint control.

Just because one may have redundant measurements, doesn't mean you have to use them all... unless things are critical. Often better than averaging is to use the middle value (consider input failure).

 

Sounds like your idea of taking the average of the three might work alright. Sorry about my previous post which seemed to ask the question you had just answered, your post was not up when I wrote mine.

Steve

 

I have a power measurement instrument and i want to control the current of the three phases(3inputs).
The variation of the current per phase is different every time beacause it depends of chemical reactions every time in different locations of the product so it affects its phase different.
I have one chemical material(output) that i will throw symmetricaly to control the 3 currents.

 

If this is a 2 out of 3, or even a three out of 3, situation why aren't you using the setpoint establshed by the average signal?

Regards,
Phil Corso, PE
Boca Raton, FL
[[email protected]] ([email protected]) {[email protected]}

 

Is this similar to a three-phase electrical annealing pocess that is controlled via SCRs or Triacs?

Regards,
Phil Corso, PE
Boca Raton, FL
[[email protected]] ([email protected]) {[email protected]}

 

Sounds like you might be trying to do some kind of endpoint control based on "average" amps on the mixer motor. As others suggest, you probably want some kind of average or check out http://www.loadcontrols.com/index.html especially if motor is lightly loaded.