AB 5-03 swap

I have a 5-03 processor that I believe may be bad. I have one on the shelf that I would like to swap out.
Does someone have an explanation or a link for this process?

What I have is a DC sinking input card that does not read the state of the input everytime. The LED on the card itself will indicate everytime, but it does not with a histogram . I added an unused input to the rung and forced the old, but the new input responds the same. The prox is fired off a cam at the rate of about twice a second and sometimes will not show on the histogram for 2 or three seconds even though the indicater on the card changes state.

It could still be the Input Card, but I don't have one on the shelf.

I am using RsLogix 500 for this.

Help, please!

From your description, it's unlikely that either the Input module or the CPU is faulty. SLC-5/xx controllers check the integrity of the backplane and module communications every scan, so there's almost no way for an input to turn on at the module's indicator array but not arrive in the data table.

The histogram or data table views are a different story; they rely on a communication network that can slow down based on the performance of the network interface and network congestion.

My method for checking I/O filtering and response is to just put in a dummy rung that connects an unused Input point to an unused Output point:

--] [--------------( )----

Use a jumper to trigger the input, watch the output.

Another mechanism I use is to run a Counter from the questionable input. If I see as many Counts as I expect, then I know the input data is getting as far as the Input Data Table so I can focus elsewhere in the program.

Since you described going online with a histogram, you likely have the right tool (RSLogix 500) to upload from your current controller and download into a newly installed one. But I wouldn't try that yet.

Ken Roach said:

From your description, it's unlikely that either the Input module or the CPU is faulty.

Click to expand...

The input module could very well be faulty; if you used an input on the same module, that might indicate a module problem. I've seen lights lit with no input/output.

Use the bit check, though... another suggestion is to use an OTL so you can make sure the input is valid.

OK, thanks guys. The input check with the counter showed the anticipated number of counts. So I guess I am back to it being a timing/pressure issue on the mech/elec side. What I have is a Alcoa bottle leak check that I can't get to do anything but accept or reject all bottles, whether they are known leakers, known good, or a combination of the two.

Another thing to check is the speed of the sensor itself. Some are opitimised for high speed operations and some aren't.

When your monitoring the histogram, make sure your update rate is set to the fasted. It's hard for me to believe that you can see the prox indicator come on and the histogram doesn't. You see these problems when using trends, but I have not seen it with histograms. They actually catch every .acc change in .01 based timers. That's pretty quick.