Colin Carpenter

How do PLCs fade away?

4 posts in this topic

In your experience, do old Mitsubishi PLCs just give up the ghost and fail catastrophically , or do they start doing one or two odd things that you can't explain?

I just wondered, as an "old faithful" Q2AS-S1 did something the other day that I just can't explain after tens of thousands of hours without issue and a good few hours since without issue.

I normally find that it's me that's wrong and not the PLC, but I just can't explain this one without reverting to my "quantum fluctuation" reasoning when all else fails.

Share this post


Link to post
Share on other sites

What did it do?

I was only thinking the other day that in 25 years i have only had one fail (an FXon)

A few outputs maybe, but that's it.

Share this post


Link to post
Share on other sites

In our facility we have a lot of Mitsi PLCs, some with 15 years up time. No PLC fail so far.

In my experience strange behavior is always caused by peripheral electrical equipment failure or poor programming.

Exception was one Fx3U... one of the relay outputs was reached the cycle life limit. We change the relay and the unit is back on run.

Once another 3U stuck in error state and the diagnotic tool cough it right away. Overload on PLS's 24V service power supply. Turns out to bi a smashed sensor cable.

No program failure or loss... Mitsis are rock solid. We have had some S7 and AB failures, but never Mitsubishi. That includes VFDs and servos.

 

Share this post


Link to post
Share on other sites

Like everyone else, my experience of Mitsubishi PLCs is generally one of absolute reliability, with units "just working" for years and years. 

The only issues I've had are several output relays failing after tens of thousands of operations, occasional individual analogue inputs failing for some reason on 8 way modules and heat related issues in enclosures that are closed in the height of summer, when temperatures can get pretty high in some locations.

The issue on this PLC (which has been running for about 12 years without issue) possibly relates to my programming style, but maybe not:

I tend to use a lot of SET and RESET commands for internal flags ( M coils) and then group those flags together on the Y outputs of the various valves and pumps etc to cause things to happen in the real world. This means that if I have a valve or pump that's not running, it's easy to have a look at the flags and determine where the problem is.

Without exception, I always use a PLS command to operate the SET or RESET function so that it only happens once  on the rising edge of the relevant rung going TRUE.

This is very easy in IEC Developer as the software compiles using a system variable M coil as the pulse address without me having to specify it.

This works fine and the software has been "rock solid" for many years, but sometimes (maybe once a year)  I get something I can't explain and my suspicions always fall on whether or not the PLS command M coil was actually active on that one cycle scan when when it's supposed to be, and of course there's no way of monitoring that because it's been and gone in an instant.

This can work either way - sometimes Flags don't SET ON when they're supposed to and in the latest instance, it seemed like a Flag SET ON when it was "impossible" for it to do so without a lot of other flags being in the right state. This last instance has never happened before in the 12 year lifetime of the code and is puzzling me because I can't recreate the issue or even see how it happened.

Let's say that one of these issues happens once a year and that the cycle scan is 100 ms (it's a big programme), then that means that I'm suspicious once in every 315,360,000 cycles  or so. In most cases I just leave it to see if it happens again and it never does, but it can be hard explaining that to a production manager who has just lost some product.

They say that all quantum theory depends on probability, and it almost seems like every so often, things happen that can't be explained ..... or is that just a "get out of jail free" excuse.

Puzzled.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now