[solved ] Cable type PPOP 20 pairs 0.35mm2

[Grossel]

Hi forum.

Have two cabinets with busbars, one of the cables between those cabinets are of type PPOP 20 pairs. The conductors have colored insulation, but several colors repeats (at least 5 strands colored red).
the cable type is printed in drawings (cable list) from '94 papers.

Problem is - for this plant, the original drawings are from '94, some of the components inside are being replaced around 2001 - and sadly only parts of the drawings are being redrawn/upgraded but others are expired, and where not redrawn when this facility was upgraded.

Therefore I have the need of tracking some of the single conductors. But I face these problems:
1. Since there are 40 conductors, the color insulation repeats (i cannot just follow the red one, because there are 5 or 6 of same color)
2. The plant is operative 24/7 - so it is very undesirable to cut power in order to track every conductor.
3. The conductors seem to be pretty much being choosen by random, there are no color order nor use of pairs (is next to impossible to locate pairs anyway because the connectors in the middle of the cable split is so tight that it is not fasible to try to track)

Only reason I post this issue is that I hope somebody might have a good idea if there is clever ways to track the cable conductors while operative.


Thanks in advance

 

[JimB]

There are a couple of possibilities:

Look carefully at the individual cores, they may be identified by text printed on to the insulation.

If the wires are in "pairs", one of the reds will be twisted with (say) a white, another with a blue, another with a green,... etc.

JimB

 

[Grossel]

Ok - I have photos of the cables and busbars, but those aren't good enough to spot a stripe of a second color.

Thanks for the input, I'll get out to the plant and have a look - hopefully I get time tomorrow.

 

[JimB]

I have photos of the cables and busbars, but those aren't good enough to spot a stripe of a second color.

I was not thinking about wires with two colour insulation.

I was thinking about the type of cable where wires were twisted together in pairs, and each wire of the pair is a different colour.
Like this one:
https://uk.rs-online.com/web/p/twisted-multipair-industrial-cable/0823229/

JimB

 

[Grossel]

Well - I have to get to the plant in some near future and make corrections to the drawings anyway, and I assume there is no other real choices than have to disconnect the plant communication anyway in order to find the correct wiring . . .

 

[KeepItSimpleStupid]

What types of signals are on these wires?

I don't know if this will work, but it;s still an idea.

Say you have 5 reds and you don't know which red it really is. Say it's a 4-20 mA signal.

You can use signal type.

there are meters with detachable heads, but ai don;t know how far they work. Maybe you could do something like an IEE-488 mater on WIFI to look at a signal where your not.

4-20 mA is non intrusive. Monitor voltages/currents at both ends. Maybe even graph them.

I'd think about graphing them at both ends. Use labview and an IEEE-488 meter or even USB.

4-20 mA signals can be tapped without opening the circuit. There might be a slight offset, but the end to end signal should be recognizeable.

RS-485 might be a little harder.

How about something even wierder. Two scopes with a way of triggering them at the same clock time. e.g 10:00:01 sec. Capture at \both ends and compare.

A scanner might be able to remotely connect to the 5 red wires.

I'm giving you ideas, but you haven't speced a lot of stuff.

Type of signals? voltage, current, serial
How far away?
Access?
Are we talking cabinet to cabinet?

These are just off the wall ideas.

 

[Grossel]

Thanks for input.

The signal cables is for remote control of the plant. It is a mix of commands and binary messages.
There are relays in both end, both for sending commands and for receivving messages (or error status) - aka 24V voltage readings.

The communication unit is an ABB RTU 210 - contains 4 physically units, where 2 of them contains the modules of sending commands and receiving messages (24V on or off).

The cable goes between two cabinets, Acess to both.
In both cabinets, the cables is connected to named bus bars.

There are also analog readings (0-10mA according to some of the new drawings), but all those are sorted out and is assigned to another cable (one that have number labeled onto conductors insulation) unrelated to this one.
The cable that have all conductors numbered is really super easy to see where it goes, I really wish that was true for all multi-conductor cables (where the sheer numbers of conductors requires same color to repeat).

I'm not sure what you mean with how far away, the length of the cable is only about 3 meters (10 feet) in length and the cabinets are so close to each other that if the voltage to the plant remote control system is shut, It would be fairly simple to just conduct a simple diode test to check the actual connection and just draw up a table for it. But in this case, it's not desired to sut it down - as said before, somebody in the past failed to take ownership of the job, and never updated the drawings due to an upgrade of the plant

 

[KeepItSimpleStupid]

I'm not sure what you mean with how far away, the length of the cable is only about 3 meters (10 feet) in length and the cabinets are so close to each other that if the voltage to the plant remote control system is shut,

That's exactly what I mean 10' vs 200 feet.

So, I was looking at this: https://www.fluke-direct.com/product/fluke-233-remote-display-multimeter

as an example for signals within 30'.

Within 10', you could do it with a single dual channel storage scope.

If you get the same capture both sides, yippee.

When the plant actually does have to go down, you can do a quick verify without chasing everything.

Two dumb examples:
4-20 mA signals.

probe both with a non-contact 4-20 mA probe
Use the headless meter or wit 10' just use another DVM with 10' of probe.
Both the same value, good chance of being the same cable.

If they were 0-10V signals, same thing.

if you could disturb and perturb even better.

Serial type protocols could be harder.
Probing with a scope with a serial analyzer if it was RS232 or RS485. You can monitor RS485.

There's too many variables.

But like I said, if you just had to identify "which red" it could be easier.

So, I'm suggesting some type of signature analysis where you determine the type of signal first then use a technique that would differentiate the information.

4-20 is closely related to 0-10 probably, but you know it's analog.

With a scope at one end, you MIGHT be able to use a toner close to the wire and see the result at the other end,

When your all done, you have this "probable map". You just have to verify when you disconnect stuff.

But I think for serial protocals, you can capture both ends with a single trigger and compare the results.

10' away does make your job easier.

Digital 24 VDC logic signals would be very difficult to find. Even if it's 24V, you may be able to find it by current.

From what you said, you might have 5 red wires ad you know it's one of them. This may make it easy.

Ths https://tmi.yokogawa.com/us/solutio...amp-on-process-meter-for-dc-ma-current-cl420/ has analog output available for current.

 

[Grossel]

UPDATE

Lucky for me, it turns out that the cable conductors are twisted into pairs, but not like a cat5. For this particular, cable each full turn (twist) takes at least 10 inches, and therefore it is somewhat tricky to identify the pairs, but doable as they tend to hang together.

In cabinet one, there was enough length for the strands for me to be ale to identify pairs. It also turned out that none pairs had similar color combination, but all strands have solid colored insulation (no stripes or other markings).
So I was able to make a table of pairs.

In cabinet 2, that's another story. Those dumbasses (sorry for saying so but I waste a lot of time on this) that installed this back in the time have splittet many of the pairs, and therefore made me waste more time because of need to guess and then rule out the impossible posibilities, so to speak.
But in the end I'm now able to tell what busbar ID and connection number each strand is connected to, in both ends of the cable.

 

[KeepItSimpleStupid]

yaah!!!

 

[JimB]

Grossel, a bit of terminology...

You are using the expression "Busbar".
This usually refers to a thick conductor made from solid copper.

In your situation I think you are referring to a series of individual terminals which are mounted together on a metal bar (usually steel).
This is usually called a terminal block or a terminal rail.

JimB

 

[Grossel]

Thank you JimB for correcting my english

terminal block it is.