Wire numbering systems?

[DeeKay]

I've run across a few different kinds of numbering systems in my career. Just looking for some opinions and perspective on the different systems out there. What do you prefer and why?
I've used/seen the three below, they all seem to have their pros and cons.

Point to Point: both 'ends' or terminations of the wire share the same wire number. So lets say the wire comes from a PLC input and is called I:02/11, the point that that wire terminates at (push button, level switch etc) would also be numbered I:02/11 even if the wire passes through a terminal block. I've seen these systems also have a generic wire number for any power wires, neutrals or commons no matter where they go; a 120vac control power wire would be labelled something like 'X1' no matter how many 120vac control power wires are on the machine or in the building.

PROS: schematics/ladder diagrams are very simple. Easy to follow for novices. OK for small equipment with only a few wires (<100) Easy to readdress IO or use spares.

CONS: fault tracing is difficult in the field because multiple wire numbers might be the same wire or they might not be. You need to have the schematics to find where a wire ends up. If you loose a wire number and don't have a schematic, you're screwed unless you feel like ringing out a hundred different wires.

To and From/Source-destination: Each wire connection has a unique number for the entire run, the number at the termination says where that wire is coming from (or going to). Let’s say I:02/11 passes through a terminal block called TB2- and then to a level switch called LSH203. The wires would be numbered like this: The wire on the Level switch(LSH203) would be called TB2-2, the wire on that side of the terminal block would be called LSH203, the wire on the other side of the terminal block would be called I:02/11 and the wire at the PLC input would be called TB2-2. All control power has a unique number as well, usually something like X1-1, X2-1, X3-3 etc. A lot of these systems will also have a modifier on the end of the wire number that tells you what page and line number in the schematic to find the wire on. Super helpful in schematics that are 100 pages long.

PROS: Easy to trace and troubleshoot in the field, all you have to do is look at the number on whatever you’re troubleshooting and you automatically know where the other end of that wire ends up, even without a schematic. This is especially helpful on more complicated equipment with 100s or 1000s of IO points, devices, instruments etc.

CONS: complicated schematics, complicated numbering system. Can be confusing for novices or people unfamiliar with the site. Makes readdressing or moving wires to spares a more complicated process. It can get confusing trying to keep track of multiple wire numbers for the same wire.

Combination: Each wire has a common number at both ends along with the to and from numbers.
I’ve only seen this used once in the field which is unfortunate because it’s the best system out there IMO.

 

[Jim greengo]

We used to number conduits and junction boxes on bigger commercial jobs.

 

[RPH]

I’ve used what we called USA style, one wire number on the wire from start to end. If you have a two on the wire then that two was common to all two’s. Easy and very basic from prints to machine. But like you said < 100 wires and expect to ring things out. Nothing worse than looking at an open box with a bunch of wire tags laying in the bottom.
The other system you mention is a European style, yes prints become hundreds a pages and it takes a matrix of pages to follow the circuit. The advantage of the euro is one device per page, that’s where hundreds of pages come from. But this means that a single wire will have different numbers on there respective ends and at any terminal block it passes through. Also means no common wires. This style allows for quick device change out as the device terminal is used at that device. It takes time to use the euro prints effectively.

 

[theoldwizard1]

Your wires are labeled ?

Ever been in a computer server room ?

 

[Bad Habit]

Plenty, everything is, or at least should be.

http://innovave.com/wp-content/uplo...ns-Infrastructure-of-Commercial-Buildings.pdf

Sent from my SM-G975U using Tapatalk

 

[ducatithunder]

Im a euro fan. 80% of our drawings are ABB on my class of ship. Problem is the other 20% are combined between different manufactures. Funny I was talking with a friend a few days ago. He made the transfer to shoreside and was telling me how much easier US drawings were then Euros. Once you get it you can figure out anything.

I do agree when you open that IO panel and there are 12 new modules with no labels or wire identification,

 

[mark#3]

The old military vehicles all had numbers on the electric wires

 

[ddawg16]

I’ve used what we called USA style, one wire number on the wire from start to end. If you have a two on the wire then that two was common to all two’s. Easy and very basic from prints to machine. But like you said < 100 wires and expect to ring things out. Nothing worse than looking at an open box with a bunch of wire tags laying in the bottom.
The other system you mention is a European style, yes prints become hundreds a pages and it takes a matrix of pages to follow the circuit. The advantage of the euro is one device per page, that’s where hundreds of pages come from. But this means that a single wire will have different numbers on there respective ends and at any terminal block it passes through. Also means no common wires. This style allows for quick device change out as the device terminal is used at that device. It takes time to use the euro prints effectively.

I ******* HATE that method. Without good prints, you're kinda fucked

The other method I hate is line numbers....which is each page of a schematic has a incremental line number down the page. If a wire is on line 5523....that is the wire label you use. Makes it easy to find it on the schematic....***** if you don't have the schematics.

Both of the above methods were common because it was not easy to attach wire labels. In the old days it was a book of strips of numbers. You tore out a strip of say 5's....and then a strip of 3's...then 8's. to give you 538. A real PIA.

With current hand held heat shrink label printers, you use tags a bit more descriptive....like the actual PLC Input/output...say O:15/03. On a field device 500' from the PLC, you don't have to guess where it goes.

On wires not connected to the PLC, you can just type what it is...say "Vess Temp 2"...or, "THIS WIRE GOES TO THE ******* MCC"

 

[kaffine]

I used to do everything I could to avoid needing labels. I had several racks of equipment that originally had long labels on each end of the wire with rack number, rack unit number, port number, signal type then the destination which had rack, rack unit, and port numbers. Most of the cables went from 1 rack to another. I redid all of the racks and made most of the cables less than 18 inches and used 10 colors to eliminate the need for labels.

A lot depends on type of equipment I am working on and if we are allowed to have easy to understand wire labels. I have worked in places that was actually against policy to have labels that can be understood without a lookup table.

I prefer simple labels that can be understood without a chart or schematic as I know I will never have the chart or schematic when I need it. I would use sensor name, signal type, then panel point number. So I would have a label PIR 3, ALM, 01-16. That would be on the panel side the other end normally wouldn't get a label we had standardized color for function only if we had multiple sensors in 1 point would they get labeled with sensor number. I also tried to eliminate any splices or terminal blocks between panel and device. Granted this worked as most of the system was simple even though we had hundreds of points.


For wires in a panel we would use heat shrink labels but not shrink them down. So if we needed to move a wire we could slip the label off and a new one on.

 

[cvairwerks]

Ours are individually labeled with harness number, wire number and size, all lasered on each wire. Harness gets overbraided and labeled with harness number and a systems code. Connectors and splices get labeled as well.

Works pretty easy if you need to reterminate a wire at a connector and there might be 50-100 22 gauge wires in it.

 

[TRWham]

Numbering is also useful during construction. One of my jobs as a co-op was to administer the cable routing system at a large steam power plant under construction. Each wire or cable had 3 cards: PULL, FROM and TO terminations. The cards were issued to the field by schedule and returned when the work was done. The pull card identified the wire type and routing listing the various raceways it would run through and the termination cards identified where it was going from and to.

It might be months or longer between phases of a given pull. That is, the wire might be pulled, but the terminals were not yet installed so nothing was connected, or sometimes the pull itself was incomplete. The cards were left attached to the wire until the work for that card was complete. Without a robust numbering system it would have been chaos but I don't recall the details of the numbering.

There were 14,000 identified pulls in Unit 1, with 12,000 in each of the other three (additional instrumentation in unit 1 required more connections).

 

[DeeKay]

I’ve used what we called USA style, one wire number on the wire from start to end. If you have a two on the wire then that two was common to all two’s. Easy and very basic from prints to machine. But like you said < 100 wires and expect to ring things out. Nothing worse than looking at an open box with a bunch of wire tags laying in the bottom.
The other system you mention is a European style, yes prints become hundreds a pages and it takes a matrix of pages to follow the circuit. The advantage of the euro is one device per page, that’s where hundreds of pages come from. But this means that a single wire will have different numbers on there respective ends and at any terminal block it passes through. Also means no common wires. This style allows for quick device change out as the device terminal is used at that device. It takes time to use the euro prints effectively.

I’ve never seen schematics having one device per page, that sounds like a nightmare! Our prints for our plants are laid out just like normal ANSI ladder diagrams but with the “euro” numbering system with as many things as they can fit on an 11x17 sheet, there’s just a **** ton of IO so it takes 100s of pages haha.
I have noticed that I saw the point to point numbering system more when I worked on standalone machines like the box machines, conveyors and packaging line stuff.
It seems like I see more of the ‘To and From/Source-destination’ or ‘Euro’ system now that I work on more process control systems on air separation and chemical plant type stuff.
That being said I’ve seen a mix of both, everywhere, on all types of equipment from the US, Europe, Asian etc. I guess it probably depends on when the stuff was built, if it was to be exported, imported, domestic, custom blah blah blah. I know there are standards from IEEE, ANSI, NEMA, IEC etc. but there’s nothing to say what gets built to what standard from what I can tell.

Your wires are labeled ?Ever been in a computer server room ?

haha I have seen some nasty server rooms, thankfully they've never been my responsibility though. Wires are 'supposed' to be labelled in industrial controls but I've definitely had to resort to the old tug and find out where it goes method. Or cut, ground and ring out method too, sometimes there's just no way around it!

 

[RPH]

Years ago I the pleasure of installing multiple bonding tools at the Ford plant in Saarlouis, Germany. They took 11 page of 11x17” prints and informed me that they did not meet specifications. They reworked the prints and returned a package that was now 60+ pages. This was their improvement. My set included pneumatic and cooling prints. Thankfully I still had my original set to work from.

 

[DeeKay]

That's nuts, If they tried to do that with some of our sites the prints would end up being as thick as an NEC code book!
converting some of our stuff on PDFs has proven to be helpful because you can just ctrl+F and find what you're looking for pretty quick.

 

[RPH]

I have prints that are 11 x17” and 800 pages for the complete cell.

 

[Jagmandave]

Whew! Need my glasses....I though the title was WIFE numbering systems! Which led to the next questions....were they talking about numbers or making her not feel anything?

 

[Dustball]

We do our numbering based on column + row numbers in our schematics. Column 12, row 13 would be wire 1213 or cable CBL1213. For wires and cables that begin and end on different columns, the number it'd have would be where the signal originates from.

Every wire has it's own number unless it's on the same circuit where it's split to more than one place.

 

[ddawg16]

We do our numbering based on column + row numbers in our schematics. Column 12, row 13 would be wire 1213 or cable CBL1213. For wires and cables that begin and end on different columns, the number it'd have would be where the signal originates from.

Every wire has it's own number unless it's on the same circuit where it's split to more than one place.

I HATE that method......

 

[Dustball]

I HATE that method......

It's no different than the other methods you list.

With current hand held heat shrink label printers, you use tags a bit more descriptive....like the actual PLC Input/output...say O:15/03. On a field device 500' from the PLC, you don't have to guess where it goes.

Ok, you're at the PLC and you have a wire that says O:15/03 and you don't have the schematic with you. Where's the other end of that wire go to?

 

[ddawg16]

It's no different than the other methods you list.

With current hand held heat shrink label printers, you use tags a bit more descriptive....like the actual PLC Input/output...say O:15/03. On a field device 500' from the PLC, you don't have to guess where it goes.

Ok, you're at the PLC and you have a wire that says O:15/03 and you don't have the schematic with you. Where's the other end of that wire go to?

I go into the PLC and see what the description of that wire is. Otherwise, if I don't have drawings, I shouldn't be working on it. But if that wire says 5395, and you don't have schematics, what do you do?

But, in most cases, you're looking at a field device. If it has wire label O:15/03, then I know where it goes. If it says 5395? Well, ****.

Schematic line #'s are just as bad. They really **** if you're the one doing the drawings. ***** even more when you need to insert pages.

 

[RPH]

Every customer seems to have their own way too. Some make sense and others are just plain difficult. In today’s world of distributable i/o systems it can be more confusing.

 

[manwithtools]

We do sheet and line number, so if the wire number is 10025 or CBL10025, it's from sheet 100 line 25. We then break the sheet numbers into groups, 100's = one line diagrams, 200's = there phase and single phase power distribution, 300's = 24VDC power distribution, etc... This makes it easy to add sheets if needed without affecting the entire drawing set. We use AutoCad Electrical for all drawings.

This makes it super simple to find the correct drawing where the wire originated, the wire always has the same number from beginning to end. With a complicated machine, you are screwed without drawings, I don't care what your wire ID system is.

 

[DeeKay]

I go into the PLC and see what the description of that wire is. Otherwise, if I don't have drawings, I shouldn't be working on it. But if that wire says 5395, and you don't have schematics, what do you do? But, in most cases, you're looking at a field device. If it has wire label O:15/03, then I know where it goes. If it says 5395? Well, ****. Schematic line #'s are just as bad. They really **** if you're the one doing the drawings. ***** even more when you need to insert pages.

I hear you man, I hate when wire numbers appear to just be arbitrary numbers, even if they actually do represent something , they should be fairly intuitive. Like if I see wire number LSH500-NO I know that wire is going to level switch “high” 500 which is on the air compressor and the normally open contact. If the drawings have a good legend and sequence page, it’s pretty easy to look up the page and line #. Like you said If I find wire 5395 then I’m fucked unless I have the drawings memorized and know what 5395 is going to.
Yeah you should always have drawings to work on stuff, but in the real world that's not always the case, or you have the drawings but they're not accurate. You work with what ya got sometimes...or what ya don't got.

We do sheet and line number, so if the wire number is 10025 or CBL10025, it's from sheet 100 line 25. We then break the sheet numbers into groups, 100's = one line diagrams, 200's = there phase and single phase power distribution, 300's = 24VDC power distribution, etc... This makes it easy to add sheets if needed without affecting the entire drawing set. We use AutoCad Electrical for all drawings.
This makes it super simple to find the correct drawing where the wire originated, the wire always has the same number form beginning to end. With a complicated machine, you are screwed with drawings, I don't care what your wire ID system is.

The company I work for uses a similar numbering system, but they’re numbered so that the system works for both the electrical drawings and the P&ID; 50-59 air compressor lube sytem, 60-69 O2 compressor lube system, 500-599 air compressor process, 600-699 O2 compressor process etc…
I’m not a huge fan of point to point wiring in large systems, but if the drawings I have are fairly accurate and have a good legend I’m a happy camper with any of them.

 

[ddawg16]

I hear you man, I hate when wire numbers appear to just be arbitrary numbers, even if they actually do represent something , they should be fairly intuitive. Like if I see wire number LSH500-NO I know that wire is going to level switch “high” 500 which is on the air compressor and the normally open contact. If the drawings have a good legend and sequence page, it’s pretty easy to look up the page and line #. Like you said If I find wire 5395 then I’m fucked unless I have the drawings memorized and know what 5395 is going to.
Yeah you should always have drawings to work on stuff, but in the real world that's not always the case, or you have the drawings but they're not accurate. You work with what ya got sometimes...or what ya don't got.


The company I work for uses a similar numbering system, but they’re numbered so that the system works for both the electrical drawings and the P&ID; 50-59 air compressor lube sytem, 60-69 O2 compressor lube system, 500-599 air compressor process, 600-699 O2 compressor process etc…
I’m not a huge fan of point to point wiring in large systems, but if the drawings I have are fairly accurate and have a good legend I’m a happy camper with any of them.

I 'really' love working on systems retrofitted by someone else.....and wire colors change mid stream.....and the wire nut or tape is inside a conduit somewhere.

With that said....there is a lot to be said for distributed control.

We do a lot of field mounted boxes with a single Ethernet and power going out there. I'm also really liking the Armor Block stuff. Kind of expensive up front, but pays for itself on installation.

 

[DeeKay]

Is that the distributed IO stuff with the little micro M12 connectors?

 

[Muzzy]

As long as both ends of the wire have the same label I'm fine.

Having more than one label is cool too...its nice to know that "124532233-a17-niner-gamma" goes to that LAHH-02 and it's 120 VAC not 24 VDC despite the wire color. Loop diagrams are amazing when you can find the right one. Finding the right one in the 800 pages can be challenging.

 

[RPH]

We use what ever the customer requires. So from Siemens CNC systems with Murr blocks to AB systems. And others, the Murr blocks have two large connectors for power and communication and then 6-8 individual smaller connector going to each device. They have to addressed and ports enabled for function. They have a **** load of blocks. Saves on a lot of wires but you still need a **** full of wiring.

 

[RPH]

If you think the control wiring is tough try getting into components on the boards. All the logic symbols I was taught in school went out the window with the European stuff. Especially the Norwegian and German equipment. Ah, the fun.

 

[ericm]

Ever been in a computer server room ?

I've set them up and ran them! I even had a room with ~500 ethernet terminations, on 110 punch down blocks no less (which tell you how long ago that was).

Wires get labeled when it's no longer easy to trace them. You bet the terminations got labeled.

I recently tore down a small server/network room of just a couple racks. It hadn't been labeled and it was hard to figure out what went where so I could duplicate it in it's new home.

 

[ddawg16]

Is that the distributed IO stuff with the little micro M12 connectors?

Yup

I'm switching everything over to 24 Vdc and I like using prox's vs limit switches.

All you need for an ArmorBlock is 24 Vdc and Ethernet. Because the M12 cords are pretty much like SO cord, you need less conduit and sealtight. The cordsets are easy to wire up....you get it with one M12 connector already on it. Takes 2 min to cut to length and wire the M12 on the other end.

A cool thing too about the ArmorBlock....if you put it in the right place, it's a convenient spot to see the status of the IO. Works best when you have a cluster of IO points to monitor or control.

Cost wise? Up front, a bit more, but saves you time in the field.

All too often I've had managers in the past try to save money on the front end by cheaping out on hardware, to only pay more on the backside during install.

Some think it's cheaper to have ALL of your PLC IO in one cabinet and run a few hundrid feet of conduit and pull a few hundred wires vs using dist IO

 

[Metal-Marc]

Your wires are labeled ?

Ever been in a computer server room ?

A server room located in a Central Office.

 

[kaffine]

Yup


Some think it's cheaper to have ALL of your PLC IO in one cabinet and run a few hundrid feet of conduit and pull a few hundred wires vs using dist IO

The issue I have seen with distributed IO is the location of the remote IO panels. With centralized IO the panels were easy to access and had decent room around them to work. Often times with the distributed IO we had to put the panels in hard to reach areas either requiring ladders for access, having to climb into areas I would rather not or the panel was just hidden somewhere.

I would rather run extra conduit and pull more wires and have easy access to the IO panels. I liked the distributed IO when the panels are convenient though.

Now I work on circuit boards and the traces are mostly labeled NET00129 ect sequentially numbered based on when the trace was added to the schematic. Only major nets get names based on the signal.

 

[DeeKay]

Yup

I'm switching everything over to 24 Vdc and I like using prox's vs limit switches.

All you need for an ArmorBlock is 24 Vdc and Ethernet. Because the M12 cords are pretty much like SO cord, you need less conduit and sealtight. The cordsets are easy to wire up....you get it with one M12 connector already on it. Takes 2 min to cut to length and wire the M12 on the other end.

Some think it's cheaper to have ALL of your PLC IO in one cabinet and run a few hundrid feet of conduit and pull a few hundred wires vs using dist IO

My last job had armor block on a palletizer, I actually really liked it. Super easy to replace stuff like photo eyes and prox.
We also had a system called ASi on our brewhouse tied into Siemens PCS7, Also pretty nice because everything was pretty much plug and play with a little hand held programmer.

The issue I have seen with distributed IO is the location of the remote IO panels. With centralized IO the panels were easy to access and had decent room around them to work. Often times with the distributed IO we had to put the panels in hard to reach areas either requiring ladders for access, having to climb into areas I would rather not or the panel was just hidden somewhere.

Distributed and remote IO is great as long as it's done right. That comes down to how they engineer it in the first place though.

 

[Muzzy]

We also had a system called ASi on our brewhouse tied into Siemens PCS7, Also pretty nice because everything was pretty much plug and play with a little hand held programmer.

AS-i is alright, but has less capability than most other bus protocols. I've used it a lot for valves and proximity switches, but it is more or less obsolete technology.

 

[DeeKay]

AS-i is alright, but has less capability than most other bus protocols. I've used it a lot for valves and proximity switches, but it is more or less obsolete technology.

Yeah, that's all we used it for, relatively small brewhouse with maybe 50 valves if I remember correctly so we never needed to have more than one master. I never explored any of its other capabilities if there are any.

 

[ddawg16]

The issue I have seen with distributed IO is the location of the remote IO panels. With centralized IO the panels were easy to access and had decent room around them to work. Often times with the distributed IO we had to put the panels in hard to reach areas either requiring ladders for access, having to climb into areas I would rather not or the panel was just hidden somewhere.

I would rather run extra conduit and pull more wires and have easy access to the IO panels. I liked the distributed IO when the panels are convenient though.

Now I work on circuit boards and the traces are mostly labeled NET00129 ect sequentially numbered based on when the trace was added to the schematic. Only major nets get names based on the signal.

Wait until you have conduit runs of over 100'.

One of the reasons 4-20ma was invented for sensors. In refineries where you could be 1000' or more from a pressure sensor, the 4-20ma loop would compensate for the voltage drop.

 

[RPH]

Current loops are also less immune to noise. With CNC based equipment noise can cause some crazy reaction in systems. But once the length gets long enough you need amplifiers to pump it up. Some of the tube welders 300’ or so minimum to over a thousand foot long. The Germans are great at o-10 vdc analog card that goes to a current loop converter then to its final destination cabinet. It’s then converted back to 0-10 vdc. The only rule in German engineering is you must add more complexities. The Angolan card could output the 2-20 mA and the receiving device could too. But that decreases it by two devices on the complexity count. This was on heat treat cell that make your electric steering assemblies down in Kentucky. Had 8 machining cells there.

 

[kaffine]

Wait until you have conduit runs of over 100'.

One of the reasons 4-20ma was invented for sensors. In refineries where you could be 1000' or more from a pressure sensor, the 4-20ma loop would compensate for the voltage drop.

I have done a few that were over 1000 feet. I think the longest I did was 1/4 mile and that was caused by stupidity of design. We actually pulled the signal wires though a different control panel 10 feet from the sensor and could have been used. The design engineer wanted it to go to the other panel because he didn't know how to program it to use the closer panel.

We had more issues with power for the distributed panels as they normally sourced the DC from the main panel due to battery backup requirements. So now instead of low power sensor I have the controller to power as well.

 

[toplessHO]

I have done a few that were over 1000 feet. I think the longest I did was 1/4 mile and that was caused by stupidity of design. We actually pulled the signal wires though a different control panel 10 feet from the sensor and could have been used. The design engineer wanted it to go to the other panel because he didn't know how to program it to use the closer panel.

We had more issues with power for the distributed panels as they normally sourced the DC from the main panel due to battery backup requirements. So now instead of low power sensor I have the controller to power as well.

Go work on the Cape sometime
there are points MILES apart

Im so glad that fiber has come of age.

 

[ddawg16]

Go work on the Cape sometime
there are points MILES apart

Im so glad that fiber has come of age.

Yeah....because when **** goes wrong....you want to be miles away