FX5 PLC Energizing all inputs with power OFF?!

[Bullet0597 0]

Hello all! I am new here but think this will be a great resource and hopefully I can also be a help to some people. 

Recently, I had a project at work that has me confused. The machine is being run on an A series PLC. We needed to have ethernet capability for some processes so due to cost, added an FX5 PLC. It is a transistor model and is configured for NPN. 

This is where it gets weird. I have inputs for the finger swipe inputs and light curtains tied off of the A series inputs at a terminal strip and going to the FX5 (Just added the FX5 wire to the existing wire on the terminal strip) the breaker for the FX5 is only on the positive 24VDC wire. 

 

When I trip this breaker, the FX5 inputs all go "true" or, provide a neutral to the inputs which sets the A series inputs true at that point, disregarding the light curtains being blocked and energizing the "start" inputs as well...basically creating a death trap. 

I am only assuming that somehow the neutral from the output side COM terminals is somehow being connected to the FX5 inputs when the power is killed? 

 

I put a double pole breaker in to break BOTH positive and neutral in the event it tripped and this solved the issue but, trying to understand how this was possible. Any help would be greatly appreciated!

Thank you!

Edited 22 Oct 2020 by Bullet0597
Miss spelled word

[panic mode 246]

well... show your circuit and exact part numbers for PLC and light curtain (or its controller).

 

in general i can foresee issues to be of following nature:

problem #1

connecting light curtain OSSDs to PLC inputs. this is wrong. standard inputs on PLC can be used for monitoring of safety devices but they should connect to output used for monitoring, not safety outputs.

 

problem #2

mixing signal polarity. this is very common mistake when using Japanese products. most Japanese PLCs are using NPN, often even outside of Japan. they make special "world-series" products for others but people don't bother to learn the difference or don't really know what they are getting into, or maybe buyers choose to import product that is tiny bit cheaper without considering norms or standards in market where they are distributing them. in Japan you can get light curtains with NPN OSSDs but if you are buying them in Europe or North America, it is unlikely that you would find such product. i have done hundreds of projects for Japanese carmakers in North America (Honda etc.) where all PLCs are using NPN but safety devices including light curtains are western style (positive OSSDs on solid state outputs). and they insist on grounding everything - AC and DC (which is fine). solution is to use two power supplies. one is for safety devices and negative side is grounded. the other is for standard I/O etc and positive side is grounded. live terminal in both cases is "24V" but one is positive 24V (safety) and other is negative 24V (I/O), both of which are referenced to same common - 0VDC. this way all switching devices, all sensors, fusing etc are always on the "live side" regardless of polarity - never on common or neutral. signals from one circuit to another needs interposing relay (electromechanical or solid state or optocoupler etc.). in your case it seems that disabling one device causes its IO to go low which is signal to another device. this of course can happen but as you noted such behavior is not proper and design need to change...