How to detect faulted or broken wires for PLC inputs 2

[CGSmith]

I have a PLC with several inputs to various field devices with open/closed contacts such as limit switches and proximity switches. For some time now, I've been trying to figure out a way to detect a faulted or broken control wire to one of these devices. To me, the PLC would see would see faulted control wiring as a closed contact and a broken wire as an open contact. This could be bad for my system and I'm looking for ways to protect against it. Any ideas?

 

[tygerdawg]

Redundant signals? It would add cost & complexity, but could be made to serve the purpose. A lot of robotic ESTOP schemes require dual-channel ESTOP signals, with signals of opposite polarity. Used in case of wire-flex failures or crash-failures giving false positive or false negative signals.

TygerDawg
Blue Technik LLC
Virtuoso Robotics Engineering

http://www.bluetechnik.com

 

[controlsdude]

You create fault logic with the signal in mind. Example below

If your prox was the forward position prox, then make logic when the prox is made with the cylinder in the forward positon.

xio prox xic ForwCylinderSolv timer on delay 2sec = latch fault

In the above logic when the cylinder is in the forward position the prox is not made in 2 secs then latch fault the machine.

Create same logic for the retracted position of the cylinder

I have been to so many places that do not do this makes me ashamed to be a controls engineer.

 

[CGSmith]

Interesting. In your example, the cylinder is being controlled by the PLC too, right? In many cases, the I/O is triggered by external events (not something the PLC commanded). For example, over temp switch on a motor or limit switches on manual type equipment. So, the PLC would not be able to make any comparison. What am I missing?

 

[IRstuff]

One approach that does not require additional wiring, trades that for complexity. For example, a typical home security system must be able to detect the difference between a magnetic sensor's normal open signal and a cut cable. This can be done by placing a resistive load at the contact. Therefore, a normal open contact would read, say, 1 kohm, while a cut connection would read a true open.

Thus, a circuit that sources a current into the switch contact can produce, say 5V for a complete open, 2.5V for a normal open, and 0V for a normal close.

TTFN

FAQ731-376
Chinese prisoner wins Nobel Peace Prize

 

[controlsdude]

If your saying you cant monitor it by faults then you need to buy an IO card that has broken/short circuit wire detection build in. Most PLC manufactures have this as an optional card or sometimes a feature on the card. Just need to enable it. Then create some program that looks at these fault bits and display on whatever alarming HMI display that you have selected.

 

[panelman]

to expand on IRstuff's post, add a series resistor and you'll be able to detect the full set, fault open, fault closed, normal open and normal closed.

You will need an AI instead of a DI for each circuit which might get expensive depending on numbers and/or PLC used

 

[CGSmith]

IRstuff/panelman, for some reason I didn't consider switching to analog inputs but it sounds like that would work. The resistance over the cable run is so negligible, I think a basic resistor would do the trick.

Controlsdude, I am looking around for modules that have this built in feature and think I have a couple of options.

Before I posted this topic, I came across a company that sells separate modules specifically to detect these conditions. They work with the PLC I/O rather than replace it so, its a lot of added hardware. They use zener diodes rather than resistors but function off of the same idea, I think.

 

[AstroEngineer]

Food for thought. Depending on the application of the input you can monitor the input state with code to make sure it cycles (discrete control) if it's process control it's a little harder. Some discrete manufacturers even force a "bad master" to force the opposite state of a switch to make sure it's working.

It's good practice to always make sure a switch fails to the safer state. More often than not a switch will fail to the open not closed.