Electronics > Projects, Designs, and Technical Stuff
wire tracer for live 24V control system tracing
(1/1)
Cicada:
Hi
I would be really thankful if you look at this post carefully and provide me with advice.
First off. Let me set the tone for the rest of this post.
I work at a facility where we handle radioactive materials. I mention this to emphasize the importance of not interfering with the normal safe operations of the facility while I conduct the measurements that I describe below.
First a bit of background:
The facility is fairly old. We have long cable runs between different geographic locations on site. The cables are landed in terminal boxes (sometimes as big as wendy houses) on both sides of these long cables runs. The majority of the cables are multi-core twisted pair cables with overall braid or foil shields. The pairs on these cables interconnect electronic control units in these different geographical locations and are used to convey control and status signals. The majority of these control and status signals are on/off type signals. The transmit side will have an isolated dry contact (relay) or transistor output (optoisolator) connected to a twisted pair in a multi-core cable. This pair runs off to a receiver that has a 24V power supply and opto receiver circuit. See the attached pictures for the inputs and output circuits.
(See note 1 later. Not important to this conversation)
There will, of course, be different variants of this type of design implemented over the years with varying levels of quality in design i.e. safety margins against inadvertent switching due to noise pickup. I mention this lack of immunity against "inadvertent switching due to noise pickup" because I think it will have a big influence on the choice of test equipment or measurement method that I will have to use to conduct the tests that I want to perform.
The problem and test:
In some instances, over the years proper labeling of cables (and documentation)has not been done as the system was repaired and upgraded. We now sit with a problem that we have to manually trace cables on cable racks with between 10 to 100 tightly packed cables per rack. Indeed a nasty job. I want an easier method to trace the cables. Maybe with equipment such as the type that telephone technicians use to trace cables in facilities. As mentioned earlier the general control and safety control systems cannot be switched off so I have to trace cables live i.e. with the power to these control units on and the system operational. My first thought is that one should use equipment with a transmitter on the one side that does NOT connect galvanically to a conductor in the multi-core cable. There must be capacitive or inductive coupling to the conductor from the transmitter. The conductor should then act as an antenna along its whole length. The receiver might be able to detect this induced signal along the length of the cable depending on the effectiveness of the cable shield. The receiver should be able to detect the signal on the other side in the terminal box. The voltage/current induced on/in the wire by the transmitter should not be so big that it triggers the receiving opto circuit to change state; even for poor designs with low margin. The test method should also work whether the control signal circuit is open or closed circuit, or rapidly changing state i.e. 10Hz to 1kHz.
Is there such test equipment available for control circuit type applications. Preferably from reputable test equipment suppliers. All the solutions that I see are either for tracing power cables to circuit breaker panels or identify wires in telephone circuits.
I mentioned the galvanically isolated type of test equipment above purely because it might be safer than directly connecting to a wire with high voltages without checking first. In a hurry, you don't always check things. Of course for the type of opto circuit that I describe above a direct connection to one wire might also work well if the test method and equipment are properly chosen and well designed.
Please share your thoughts and advice with me.
Cicada
Note 1: There are of course many other types of cables and signals. For example, thick 100A+ 10V cable, pico to nano amp level sensor signals on triax cables and also 200V to 2000V signals on coax cables. These cables are not as much of a problem for me.
jbb:
Blimey!
That’s a difficult ask. I can’t speak to what odd the shelf stuff might be available, but I’m sure the physics allows a clamp on exciter.
However, if there’s a variety of poorly documented stuff in the installations, I would be concerned about accidentally disturbing something important on the far end of your probe signal.
I suspect you can’t tell us much about your site operations, but I expect there some processes where disturbances are a Bad Thing?
David Hess:
For coupling on the input and output sides, I would use current clamps however for this to work, only one wire of each pair must go through the clamp. Then apply a low level signal to one clamp and use a tuned receiver on the other, or if you want something more sophisticated, apply a spread spectrum signal and use a correlator on the receive side but that is a lot more complicated.
Navigation
[0] Message Index
Go to full version