When is opto-isolation necessary?

[coalhound]

On a proposed project I have multiple 4-20 mA process control signals interfacing to an AB PLC system via remote I/O. There are two options to choose from in the remote I/O, isolated and non-isolated. There is a big price difference between the options. I normally think of optical isolation for voltage spikes, ie. lightning strikes, will I gain possible noise suppression?

 

[johnday]

Opto isolation is perhaps most ocmmonly used to eliminate potential problems of earth loops, noise loops and potentail exposure to high voltage. In a hostile indiustrial environment or where high voltage difference spotentially exist is is near ly essential. In other circumstances, especially in a lab, it is probably overkill.

 

[bleach]

Optical Isolation will "suppress" common mode noise which _might_ not be suppressed using the "non-isolated" inputs.

 

[ghasem]

A comprehensive designer's guide and application of "signal isolation" can be found in the following papers:

http://www.sensormag.com/articles/0400/70/main.html

http://www.sensormag.com/articles/0199/iso0199/main.html

 

[Acajun]

Another reason would be if you had a mix of transmitters with full wave and half wave rectified power supplies on the same AC line. You would want to isolate the full wave from the half wave units. This doesn't really apply to your application but thought I would mention it from a past experience.

Jp

 

[jimgraves]

Another common issue where 4-20 ma loops is concerned is the type of transmitter in use. If the loop is powered by the control system no isolation is required.

If the transmitter is 4-wire (field powered) then isolation is required to prevent tying the field power to the control system power. At a minimum the signal will drift and be inacurate. At worst a fault could bring the system down.

 

[bmsg]

Noise will only be suppressed if there is channel-to-channel isolation as well as field-to-system isolation in the input card.