Hung pushbutton on HMI

[ControlSystems]

Stuc HMI push button solution

I paid alot to learn that lesson. The HMI or most SCADA packaged momentary push buttons write a 1 followed a moment later by a zero. Or vice versa... The network if TCP is not deterministic which bit arrives first. If tthose two bits arrive in the wrong order, you get a stuck in pressed button on your display. Never use those HMI momentary PBs without adding in the PLC logic to unlatch the bit. I prefer an off delay timer, but there are lots of options. A simple network may never suffer the issue, explaining why some people never experience the issue.

 

[Jim G.]

Thanks everyone for all the feedback. I will look into programming an unlatch on the bit from the HMI.


Question: My momentary PB has a 2 second delay set on it to prevent an operator from inadvertently pushing this button. It makes it more of a "deliberate" act. The 2 second delay is set in the C-More Software.


Does an unlatch in the logic override this 2 second delay or is the delay still in effect and the "0" is not written until the PB is released?

 

[drbitboy]

... The network if TCP is not deterministic which bit arrives first. ...

I had heard that too; has anyone every used tcpdump or Wireshark to see it happen?

 

[OkiePC]

Thanks everyone for all the feedback. I will look into programming an unlatch on the bit from the HMI.


Question: My momentary PB has a 2 second delay set on it to prevent an operator from inadvertently pushing this button. It makes it more of a "deliberate" act. The 2 second delay is set in the C-More Software.


Does an unlatch in the logic override this 2 second delay or is the delay still in effect and the "0" is not written until the PB is released?

The unlatch logic should not occur until after the PLC has found a "1" in that address which should not happen until after the HMI delay.

I typically program this unlatch with a timer in the PLC (unless there are hundreds of buttons and I don't want hundreds of timers)... Doing this helps in two ways.

1) Monitoring the logic you can actually see the state of the bit change briefly.

2) The order of operations is not as critical, since the "button" will stay "on" or a second or two, you can make use of it anywhere in the code where as with an immediate unlatch, you need to ensure that all uses of that button have already occurred.