Possibility to ignore signal disturbance?

[Migel 0]

Hi there I have a small problem with faulty signal distubance ( sorry for bad english in advance ) I have 2x CS1G-H connected to each other in the same closet. One simulates a process the other is connected to it as a controller plc ( there is also a DCS system connectable in the closet next to it ) I have used the manual to ground both plcs. Unfortuantly there is a slight disturbance still on the signal, I have no idea where it's coming from since the DCS system is connectable without any disturbance. While the DCS ground is not even connected to PLC ground. Though the signal disturbance is small the problem in the program arises when the signal gets below 4mA or above 20mA because of this error. The cards only handle 16 bit so when the signal gets below 0000(hex) or higher than 1068(hex) the signal transfers to #FFFF. For the controller PLC this will make the PID go nuts and from the simulation plc this results in really bad calculations. Since both cards use 4 times anolog in and 4 times analog out it would be to impossible ( or rather unpleasant ) to manually script all signals to stay between these 2 values so I was hoping that I lack the knowledge of some setting. Which I hope you guys could help me with. Thanks in advance. Mr.Feddema

[Michael Walsh 268]

The only way to do this that I am aware of is to do it the "impossible ( or rather unpleasant )" manner that you mentioned above. It is not that difficult however. I have created a simple function block to do what you need and I have done the first 8 of your input points in the sample code below (it took about 10 minutes, it would take another 5 to finish, I will leave that to you). Simply put this section of code at the beginning of your code and then use the outputs from the Function blocks as inputs into your PID function. I used a distinct Function block instance for each input, but in truth, you could use the same instance 16 times since the code inside the function block executes in its entirety in one scan. Analog limiting example.cxp Note that I have assumed that you are using a CS1W-AD041-V1 and that you have the resolution set up to 8000. If the resolution is not 8000, you will need to modify the inside of the function block. Good luck. Edited 26 Jan 2012 by Michael Walsh

[BobB 58]

If the inputs are loop powered I would suggest you put isolators on each input - I do this everytime with loop powered signals as they will interfere with each other and drive you nuts. Powered inputs from the device are generally OK as they are not sharing a common voltage reference.

[Migel 0]

I dont exactly have the right to mess with the cables myself, but I will share this knowledge with the guy who has.

[BobB 58]

If you can determine if the problem is loop powered signals interfering with each other the filter will not work - been caught out in the past. You can get variation within the limits and get wrong readings that way. Sometimes the readings do not even move.

[Migel 0]

But to my knowledge the DCS system also uses loop powered signals and the signal is isolated in the same way as the controller PLC, so I find it rather odd that when I connect DCS the fluctuations and the below zero value problem is no longer a problem.

[Michael Walsh 268]

I agree with this statement. If the signal is noisy, the input can be worthless. You need to address the noise issue. However, even with a perfect signal, the interpreted value can dip below 4mA and a very small negative number can look like a very large positive number if not interpreted as a signed integer.

[BobB 58]

Not really surprised - it depends on the configuration of the analogue input. The Omron analogue inputs are not completely isolated from the field and use a common ground also meaning they are not completely isolated from each other. You need an Opto isolator in the circuit to get complete isolation to the best of my knowledge. It is quite possible that the DCS has fully isolated analogue inputs and therefore would not have the fluctuation symptoms. If the Omron analogue card had total isolation from the field and between each analogue input I would expect you would not be seeing the fluctuations but as it does not you need to address the issue. It may not even be noise but may be just interference between analogue inputs. Have a look in the analogue card manual at the input circuit and you will see what I mean with respect to the input not having complete isolation - there are resistors and capacitors between the + & - inputs and ground and then there is a common ground. I will give you an example - I have recently installed a job where there is one loop powered input into an analogue card and 3 driven, isolated inputs - no problem at all. I had another job recently where there were 2 driven, isolated inputs and 3 loop powered inputs. The 2 driven inputs were OK but the loop powered inputs played merry hell with each other. I put isolators in the circuit for the 3 loop powered inputs and the problem went away.

[Migel 0]

Thanks after implementing the limiter and setting the value between &0 en &4200 to accomodate the 16 bit signal, it worked. The PID worked well with the incoming signals. Unfortuantly I was neither allowed to install an isolator on the signals or make changes to the program on the simulation PLC. So the calculation values remained incoherent, I did stay after work for a while to test it anyway, it did work then. Oh well co-workers :D