INPUT & OUTPUT wire numbers

[JeffKiper]

Do you use the PLC input or output address as the wire number or do you use the sheet-line number?

The system that I am currently working on is a Logix 5000. I have a customer that is use to the PLC5 world and will not listen to the idea that there are new ways of doing things. I didn't say better I said new. He is going to get what he asked for. I just wonder what other guys do.

I can maybe see on a small machine with a few slots of I/O using the I/O address but on the RS5K world we have local addressing of L:3:O.D.09 and remote I/O addressing of R1:1:I.Pt00D. Add 5 or 6 racks of Point I/O with 10 ~30 slices of I/O. 60 pages of prints etc...

OK enlighten me with the wisdom. The only argument I can see is that if they loose the prints they have something to halfway go by.

 

[Kick-It-Up]

Sheet#-Line#,

Outputs/Inputs like you say are fine for a small machine you can trouble shoot without having to work through the print to any great degree but for something of any scale not having a quick reference back to a print page is a nightmare.

 

[Robobob]

We do it two different ways. First is to mimic SLC addressing ie: I3.5 for local rack addressing and the remote racks are R1:1:I.0. This will fit on a wire label. Second is 3:I.0 for local racks. Leaving out some of the characters helps.

 

[JeffKiper]

We do it two different ways. First is to mimic SLC addressing ie: I3.5 for local rack addressing and the remote racks are R1:1:I.0. This will fit on a wire label. Second is 3:I.0 for local racks. Leaving out some of the characters helps.

Robobob How does this work on large systems? Does it work OK?

 

[mellis]

The first ControlLogix project I did way back when it first came out, we came up with a standard that was based on the PLC-5 and SLC systems we had done before with a little twist.

PLC-5 wire number
I:0104/12 I:rrgg/cc rr-rack gg-group cc-channel

ControlLogix wire number
010412 rrggcc rr-rack gg-group cc-channel

We purposely left out the ":" and the "/" as not providing any information and taking up space on the wire label. Leaving the "I" or "O" off was OK for ControlLogix since there is no problem with duplicate input and output addresses like the PLC-5 had. The rack number was the controlnet node number. The group was the slot. Channel works fine for digital or analog. All I/O wire numbers were the same length.

I've since seen lots of variations, but none simpler or shorter. If you have multiple processors and I/O networks on the same system you can add couple letters to the front of the string. Ab010314 for processor A, network b, rack/node 01, slot 03, channel 14.

As to the issue of I/O number vs. sheet/line. The I/O number is useful information when you are looking at the field end of the wire. It's pretty obvious at the panel end. An arbitrary number based on the drawing/line doesn't do anything for me. It's one way to generate a unique number, but that's about all it has going for it. I don't usually have trouble finding the right drawing, the system name, processor, network, rack and slot should be in the drawing title.

 

[DamianInRochester]

Since I generally do everything EthernetIP I usually use SLC style addressing but use the fourth octet of the IP to designate the remote IO. I4th octet).Slot.Point O4th octet).Slot.Point. I use the IP of the main controller for the local IO.

Doing it that way for many years with no complaints.

 

[James Mcquade]

i have seen it and have done it both ways.
lately i have used the symbol description to indicate the device sheet and line number.
a limit switch on sheet 10 and line 23 would be labeled as
ls1023.

i had a 250 page print once for a slc 500 system.
i had the owner of the company find the input for a device and after about 30 minutes he found it. i explained to him that system and his next plc program revision had those symbol notations.

i am still learning the 5k system.

regards,
james

 

[dcooper33]

I prefer to see line number on non-IO wiring, but the IO# on IO wiring. I figure if I'm troubleshooting an IO problem, then I'm probably going to be online troubleshooting before I look at a print. Often times, I've seen the print line # as the symbol, or in the description for IO points in the PLC program.

Every bit of information helps, IMO, but I guess this one is really a matter of preference if there is no written standard where one works.

As far as what convention to use when using the IO number on wiring, I prefer to keep the AB-PLC5/SLC style, even on 5k systems. So "Local:2:O.Data.08" is just "O:02/09". For Analog, "Local:5:I.Ch0" is "AI:5.0". All (well, most) of our techs no how to translate this into 5k terminology.

Cheers,
Dustin

 

[OkiePC]

The first Logix system I developed (around '02 i think), we settled on

Rr:T ss(/pp or .word)

Where rr was the rack, T was the slot type, ss was the slot, and pp was the point. This was a single controller system, with one local rack and two remotes, replacing a 3 chassis PLC-5, so it was easy to make the transition from what was wire diagram line numbers mixed with PLC-5 address labels added over the years to this:

R0:I 06/14 (discrete input, local chassis, slot 6 point 14)
or
R1:O 10.1 (analog output, remote rack 1, slot 10 channel 1)

Later we realized that every wire number began with R, which seemed wasteful, but they were easy "orient" when reading them at any angle, and short enough to fit a decent size font on the labels.

It does make going from wire to prints more difficult (you have to "know" where the I/O begins), but trades that loss for the advantage of instantly identifying the physical location (and usually the path) of any real I/O.

EDIT: We could have started all the analogs with "A" instead of "R"
A1:O 10.1 (analog remote rack 1, output slot 10 channel 1)...and maybe "D" for digital, skip the R altogether.

 

[seppoalanen]

Right profession?

If I can't find wiring from circuit diagrams without wire marking, I have wrong profession.

 

[RheinhardtP]

We have used mostly 2 methods in the past.

1. Use Address of point in module. What i dont like about this is that addresses are defined in Hardware design stage already. ex 'I 113.1'.

2. RackNo/SlotNo/PointNo similar to above mentioned posts i think this is a better way to go. ex 'Rx/Sx/Ix Where Ix is digital input.

On bigger systems i would prefer the latter.

I have seen people use the actual instrument Tags but not very fond of this as it again relies on Software associations.

Rheinhardt

 

[The Plc Kid]

On slc and PLC 5 systems I like the IO number on IO and line ref numbers on the rest but in contrologix systems I just use line ref numbers to make it simple.

Get the prints go to the page and linme ref and it has all you need. I also put the local and alias tag name on the print and in the program I have the local tag and alias displayed and the print line ref # in the description so things in the program are the same as on the print.

I use eplan so all my pdf drawings are searchable and have cross ref searching so that makes it a breeze.

 

[OkiePC]

Putting the I or the O first I think I like better than toward the end. I want main classification toward the beginning and details toward the right as you go along.

Line references are just fine when you have modern straight foward design. Having the old PLC address was highly useful for systems with PLC bypass circuits or relay logic circuits shoehorned through a PLC.

The no labels thing is fine and dandy until you need to replace a section of the system...now what...you label the affected wires, no? You replace the full length of all of them? No, I like labels. Panel fires are another reason. They happen. When the relay base or contactor is left in a smolder pile, and you have no idea which burnt end was where, now you hand trace it point to point puzzling over those puny little euro symbols drawn sideways. "Is that a relay contact or a foot switch?"... Gimme labels...any system...over no labels.

EDIT: I take that back, those "dial-a-down" snap on (and off) spinning single digit plastic numbers, you can keep those. I once chase wire 9 for ten minutes only to realize some one else had tied it to wire 39, oh, okay this is wire 39, no later we found that it was 239 every where outside the main panel...you shouldn't have to dial a combination to decode the wire number. The peel and stick single digit gumballs? Yeah, keep those too..

One old machine I had to maintain had all the field devices as wire numbers. This meant that wire LS1A came from limit switch 1 terminal A, but across from it in the local panel would be a different wire number, like L1B or I:02/00.

I never liked the same electrical path to have two different names by passing only through a terminal block, but it was very easy to modify and follow without prints, you just read the terminal strip to see what devices went where. this made it easy to move things, add things, I don't think we even had a drawing of that section.

 

[Marc_U]

I started off on the shop-floor, wiring and building control panels, and then progressed to designing & doing PLC programming.

This has left me with a hatred with excessively long cable numbers, i.e. anything over 4 characters is too long, and that the wire numbers may result in being different at each end of the same wire.

Therefore we operate the policy that we number our wiring as in sequence of where they appear on the schematics, e.g. 1, 2, 3, etc.

Our arguement being, that you should not be working on a control system, without a set of panel schematics available to you.

 

[Winklehoffen]

Man, you people are obviously all lucky enough to have never worked on machinery where you don't have the diagrams to refer back to. A lot of the stuff I have worked on does not have schematics (lost or non-existant) so I'm all in favour of labeling wiring with the IO number.

Of course there's also always the stuff with no wiring diagrams and none of those fancy wire labels that needs to be up and running 3 hours ago...