Fault id subroutine needs a little tweaking...

[Toot Uncommon 0]

One of our machines would only do a single alarm for every fault. There are dozens of things that can go wrong and troubleshooting can be very difficult when your only clue is "machine fault." 'Replace stuff til it's fixed' had sometimes been the troubleshooting method used, especially for those less familiar with the machine and with ladder troubleshooting. So, I made a subroutine, installed 3 lights and hooked them into 3 of the available outputs on the plc (SLC 5/03). There are 3 stations on the machine and 6 or 7 things that can go wrong at each station. My logic takes each fault and makes one of the 3 lights blink some number of times according to a fault indication reference sheet I posted at the machine. I know my logic is over done and sloppy, but it works. The only trouble I've found is that every once in a while--maybe 10% of all fails--I get half blinks at the beginning and end of each blink sequence. I put a note on the reference sheet to treat these two half blinks as one. Yet, it is best if I can perfect it and avoid any potential confusion. I would appreciate anyone who would take a look at my logic and help me understand how I can eliminate the half blinks. I could not attach the xps document (ladder), but you can view it here: www.4micah.net/blinks2.xps I was able to attach the fault reference sheet. faultreference.doc Edited 17 Aug 2010 by Toot Uncommon

[Armadillo852 25]

Can you send me the rss file? What about descriptions for the other bits so we can see what is going on? Yes, it can be cleaned up.

[Toot Uncommon 0]

Sent you a message with the goods. Does this (your request for the entire ladder) mean the subroutine looks ok by itself? I thought for sure I had about 5-7 times as much logic as is really necessary.

[Toot Uncommon 0]

Nobody wants to give this one a shot?

[AJZ 0]

You need to synchronize the blinking routine with the clock bit. Only start and stop when the clock bit is off.

[kaiser_will 65]

From your Word file, you are heading in the right direction. I work for an OEM equipment builder and we have been refining our fault logic for many years to the point that we can easily generate a fault coil and a fault number for just about any conceivable fault. The flashing alarm sequence is a good idea, but might be cumbersome with the more faults you add or integrating into another machine. Have you considered a simple digital display, driven by a PLC output? I was thinking of a Red Lion CUB digital display, driven by an analog output from your PLC; push a number into the display corresponding to the highest or lowest priority alarm/warning. This could be a feature to integrate into other machines in your facility.

[Armadillo852 25]

Sorry I have been busy here. I will see what I can do in the next few days. 14 hour days, and now bronchitis, so I will get back with you when I can. Sorry

[Toot Uncommon 0]

Armadillo: Summer bronchitis. That stinks. Take your time. No problem. Kaiser: I think my initial thought for not doing any numbers was the I/O I assumed it would take up. There is no analog output on the plc. But, we do have other machines that are limited in how much fail info they give, and they may have more i/o available. I may look into that in the future if any of them become as aggravating as this one was. Thanks. AJZ: Can you give detail on how to synchronize with the clock bit?

[OkiePC 14]

I think what he means is that your flasher timers should only start when a fault is detected, rather than running freely all the time. The way it is, when a fault occurs, your flasher timer might be in the middle of its timing already... All you should have to do is condition the flasher timer rungs with a bit (or bits) that make the rung false when there are no faults. Edited 18 Aug 2010 by OkiePC