Flickering PLC Inputs - Cause? Plausibility?

B

Thread Starter

Brad

We have a machine that has initiated movement inexplicably, without operator input. The only way in the PLC program for that to happen is if the PLC is seeing two pushbutton inputs, and a prox input. The process that's initiates will latch on until it's complete, so it's possible that a momentary flicker on these inputs would initiate the machine.

How likely is it for the PLC to see a false positive on the inputs? I can think of a few possibilities, but maybe some of you with more experience than myself could give me an idea how plausible? Or other possible causes?

- Noise on the input wires, possibly the wire itself (1-10 feet stranded) acting as an antenna

- Ground fault, possibly the potential of the PLC and all inputs could go low enough to cause a false positive on all inputs simultaneously, just a brief flicker of ~1ms

- Glitch in the PLC firmware? Our logic is robust, but the process does latch on with the presence of 3 inputs until the process completes. Is it possible that a glitch in the PLC firmware could cause a 1-cycle glitch on an input or a bit somewhere in memory?

The PLC is a GE VersaMax Micro
 
Hi,

Common method to overcome "false positive" triggered by noise (or bouncing effect on mechanical contact) is to use timer act as a debouncing filter on each DI. Place On Delay timer with at least 1 s preset time to each DI so trigger signal need to be present continously for at least 1 s to make any logic triggered. Means operator need to push the PBs for at least 1 s. I Believe this would be enough to solve false positive caused by noise or mechanical bouncing.
Here is example:
https://encrypted-tbn0.gstatic.com/...uiB5KtVdkC_76wMGeCH0X6KUcpTurZYrLchjC0SYO6TT8

Hope this help
 
C

Curt Wuollet

The likelyhood of all 3 independently glitching at the same time is pretty small. It would make me very suspicious of what they share in common. Power, grounding, especially grounding. It would be a good time to hook up a scope and take a look at power quality and noise. Or, you can guess. If the power supply looks good and you don't have any ground loops, you could actually set a DSO to watch a signal and trigger when the machine initiates. If there is a glitch on the input you're watching (remember, it takes all 3) you go that way. If not you watch power, again triggering when the machine initiates. If neither is the issue, the plc must be at fault either software or hardware. I would imagine you've already spend enough time to afford a DSO, I have used both a Hantek and a Rigol at various jobs and my employer thought them a fantastic bargain after solving a really difficult intermittent problem or two in short order. $400 doesn't buy much downtime anymore.

Regards
cww
 
Thanks for your replies; I will look into a digital oscilloscope, that would be a good idea to monitor power and inputs for noise or spikes. This is definitely one of those intermittent issues that's difficult to troubleshoot, so I think that is definitely a wise investment.

I am wondering if maybe some kind of static discharge or a surge could send the ground reference potential of the entire PLC to a negative voltage, and trigger false positives on ALL the inputs for a brief moment?
 
D

Derrickstableford

From what I understand there are 2 push button, probably so the operator doesn't become part of the product.

So you may have an NPN system which would short the input and give a true input, and initiate movement.

With safety as a priority, I would recommend a safety monitoring relay, ab guard master series, amoungst many others equally suitable, to give a true fingers out status.

Also I would check the backplane of the PlC, as the connections to the cards can become fatigued.
 
Do you by any chance have Output to solenoids in the same cable?

I have seen the solenoid turn on inputs as it is turned off which can cause a high Voltage spike on nearby inputs. The solution is to wire a Metal Oxide Varistor across the coil.
 
Top