Measuring induced noise on low voltage cables

[PhaseShift]

I have been experiencing several problems lately with induced noise on low voltage DC systems. These problems have been occuring on 24V power systems and 4-20mA signal wires. Usually when these problems occur we just guess there is induced noise and end up having to change cable, move cable etc... I wanted to see if there was a way to acutally view the noise with an O-scope or something to see what was exactly causing problems.

If I were to hook an O-Scope up to these signals to look for this noise would I want to look at the voltage on these signals or the current or both?

For instance on case we have is were a 24V system is causing a false ground fault alarm on a processor when we know there is no ground fault present. We suspect noise. Would we want to look at this 24V voltage in this case to see if there was any evidence of noise and where it was coming from?

Another instance is where we have erratic readings on a 4-20mA signal If I wanted to look at this signal would I want to look at the voltage or the actual 4-20mA current signal on the cable with the scope? What would I want to look at to see what was causing the noise?

 

[ELA]

You can use an Oscilloscope to look at voltage at the receiver end as a 0-5 volt =normal signal.
You could also use a low amplitude current probe on the cable (connected to an O'scope) to look directly at the current.

The amplitude that you would be looking for would of course depend upon what the normal expected voltage and current were. You would then look for significant deviations (possibly a higher frequrency constant wave source of noise).

Often times the problem with diagnosing noise issues is that they may be intermittent in nature. In that case you need to set the scope up to trigger on "out of normal" deviations.

I would say that what is most important is to define what constitutes a ground fault error in your system? Is it anything less than 4 ma, greater than 18 ma ? , or is there no real current loop level associated with this error?

Often times noise is common mode and gets into the electronics and corrupts internal signals that may have nothing to do with the actual 4-20 ma signal level.
When this is the case then the levels of noise you are looking to find are totally undefined until you know the level of susceptibility of the equipment.
In these cases a Spectrum Analyzer and current probe are helpful tools.

Are your problems intermittent in nature?
Can you readily duplicate the error?
It can be helpful to attempt to correlate the error occurence with some other event occuring. Things such as other electrical equipment turning on/off, RF radios being using nearby etc.

It is said that it is always best to cure an EMC issue by attacking the issue at the source of the noise if possible.

There are EMC consultants that can be hired to help with these types of issues if it becomes an ongoing issue and you do not have access to all the required test equipment.

 

[PhaseShift]

I actually have two issues that I may be making sound confusing.

The first issue is with the ground fault on a processor. The processor has several 24V field switches wired into inputs on the processor. Occassionally the processor will read a ground fault. We traced circuit and find no fault. We start isolating sections of the circuit and can find that by removing section we can make ground fault go away. All the cable in this section of circuit is ran with other 24V cable and I find it hard to believe there is induction or sorts. I want to look at scope to be able to indeed see that we are seeing some sort of induced noise? Should we look at voltgage in this case.

The second issue is with a 4-20mA signal. This singal goes erratic when certain pieces of equipment are started. We think we have the pieces of equipment narrowed down but I again would like to prove that we are seeing noise by looking at this signal on a scope. In this case should we look at the mA current signal? Or you are saying that we can read voltage across terminating device resistor?

 

[nawao]

noise on analog sigmal wires

noise on analog sigmal wires

To avoid all kind of problems with induced noise use shielded pairs grounded at one end preferably at the receiving end

 

[ELA]

The first issue is with the ground fault on a processor. The processor has several 24V field switches wired into inputs on the processor.

So is one of the functions of these inputs called "ground fault" or is it implying that the signal line is grounding out when it should be isolated from ground?

Occassionally the processor will read a ground fault. We traced circuit and find no fault. We start isolating sections of the circuit and can find that by removing section we can make ground fault go away. All the cable in this section of circuit is ran with other 24V cable and I find it hard to believe there is induction or sorts.

How does the system define a ground fault?

I want to look at scope to be able to indeed see that we are seeing some sort of induced noise? Should we look at voltgage in this case.

You can look at voltage but how will you know what level deviations are significant until you can define what a ground fault level detection is set at (or how defined in the system)?

The second issue is with a 4-20mA signal. This singal goes erratic when certain pieces of equipment are started. We think we have the pieces of equipment narrowed down but I again would like to prove that we are seeing noise by looking at this signal on a scope. In this case should we look at the mA current signal? Or you are saying that we can read voltage across terminating device resistor?

You can look at either in this case.

 

[glene77is]

To avoid all kind of problems with induced noise use shielded pairs grounded at one end preferably at the receiving end

Are you refering to a faraday shield
connected in the 'inductive' mode ?
That tends to block high frequency signals.

This has come up in wiring VFD control signals to a motor, wherein the engineers specified the 'capacitive' mode shielded cable, which passes the VFD high frequency control signals, but rejected external induced signals.

In connecting power panels (with circuit breakers) the Grounding Electrode Conductor can be sent through a metal conduit to the ground rod. The conduit must be 'grounded' (bonded) at both ends in order for the transient voltages to pass on to the ground rod. This is called the 'capacitive' mode which will pass high frequency signals, such as lightning.

Hope I got that right.