Limit Switch Errors.

Hello,

I am working with 12 machines that use two microswitch. One determines the retracted position and one determines the extended position. I know the model of the limit switch should be ok as they have been used in other applications. They are EXd and IP66 rated so I don't believe that the fault that is occurring is an element condition

The operators notice a switch has failed if they get an alarm like fail to start. This comes up because the plc hasn't noticed the limit switch changing state when the machine leaves the retracted limit switch and starts to move forward.

This problem has been happening intermittently across several machines for 6 months now. I don't believe they all can be faulty! Inside the panel the 0vdc is not connected to earth. Would this be a good start in doing this?

Does anyone have any other theories in what it could be?

http://ecmweb.com/content/troubleshooting-plcs

The above article has some points to check and monitor when troubleshooting the PLC.

Hope this helps you out.
Regards,

Hard to say without more info. No I don't think grounding 0V will help. If the PLC is checking that the switch changed state to show motion, IS motion happening right away all the time? If this was on a hydraulic unit it may take time to build up pressure to make it move in differant temperatures, if they have a timer in the PLC to check if motion started within XX time it may be too short of time.

We switched to solid state limit switches on a machine that did 4000cycles per hour and greatly reduced our failure frequency. We also got better cycle times, as the solid state limit switches would tell us we had reached the limit in a more timely fashion.

Mechanical limit switches have a limited number of cycles they can do.

Could the terminals be vibrating loose?

Depending on your budget, consider sending them to the manufacturer for failure analysis.

If they all ways fail in the same manner, you could install redundant microswitches and do 1oo2 voting in the PLC. This will mean they get an alarm without stopping the machine.

How long have the 12 machines been running previously with no limit switch faults?

just the cowboy said:

if they have a timer in the PLC to check if motion started within XX time it may be too short of time.

Click to expand...

This. Get in the program and really find out what's really driving the alarm.

Loose limit switch mounting screws
loose limit switch head
loose limit switch arm connection
limit switch arm mashed onto limit switch or onto something else
dirt and grime around limit switch arm or in the way of travel.
loose or bad wiring.
bad input to plc.
Mechanical failure as already suggested.

regards,
james

James Mcquade said:

Loose limit switch mounting screws
loose limit switch head
loose limit switch arm connection
limit switch arm mashed onto limit switch or onto something else
dirt and grime around limit switch arm or in the way of travel.
loose or bad wiring.
bad input to plc.
Mechanical failure as already suggested.

regards,
james

Click to expand...

Great analysis. Short and to the point.

Only thing I will add is to check for loose terminations at PLC and any points in-between.

if low voltage used and nothing above is the failure, can the signal wires routed along the power wires of the actuator can pick up noise and false trigger the alarm ?

Jeff23spl said:

if low voltage used and nothing above is the failure, can the signal wires routed along the power wires of the actuator can pick up noise and false trigger the alarm ?

Click to expand...

Highly doubtful

CallumMacEwen said:

...Does anyone have any other theories in what it could be?

Click to expand...

The electrical and software aspect of these machines are only part of the puzzle here. It's still quite possible that something is at fault with these limit switches themselves or even the program for these machines has a glitch. But I would always recommend people to step back from a process and consider the installation as a whole. Then develop likely theories and prioritise them. Then start checking from the top of the list down.

You should also consider other parts of the process which may be common to all machines. Hydraulics have been mentioned. Pneumatics may also be a possible cause here. If these were pneumatic cylinders extending and retracting and a common air supply pressure for the Plant was fluctuating or dropping intermittently then a PLC watchdog timer may possibly be timing out as the limit switches are slower to switch at certain times. This could explain why it happens across many of the 12 machines. I've seen that many times with air pressure faults. It could be an intermittent compressor fault, etc. If applicable, do the machines have a low pressure alarm?

If there are watchdog timers in a program then I often place logic to trap the longest duration the signals took to change beyond the watchdog time so as to see what is happening in the real world. The trapped durations are not reset with the watchdog timer and the longest duration will always be there to view if there are issues with that particular alarm condition.

I don't know enough about the installation to get into more specifics than that but my best guess from experience is that this will most likely be a single fault common to all the machines and not a different individual fault per machine.

Regards,
George

Even if the old style mechanical limit switch looks in great condition, they are especially prone to contact corrosion at low voltages [particularly DC] where buildup of oxides and other nonconductive products on the contact surfaces cause them to fail.

Mechanical limit switches with solid state switching one solution.

Inductive proximity switches are the best solution.

My 2 Cents