Radio Frequency Interference (RFI)

[gtacs]

I am an electronics technician, and I work on various types of electronic equipment including radio communications devices which is the reason for asking this question. I was recently troubleshooting a radio communications malfunction, basically when squelch was broken, there was a “noise” (like the sound of frying eggs in the background). After troubleshooting for awhile a fellow technician noticed the trouble seemed to affect our truck radio the closer we got to the power transformer feeding the communications facility.
We called the local power company and as soon as the electric company technician got close to the transformer he said it was definitely bad. He explained there was a “gap” of some sort that needed to be maintained, if not, the transformer would arc and cause radio frequency interference. I did not understand what he meant by this but, at the time we were glad to have the issue resolved.
Since the issue I have asked several technicians and polled the all knowing “Google” I keep hearing the same sort of answers like “arcs encompass all frequencies, that's why arc welders cause RFI”, I am still unclear exactly what is happening to cause the interference.
When someone says “arcing encompasses all frequencies” are they implying that when current flows from the electrode of a welding rod to the metal that there is a wave of electromagnetic energy released that encompasses all frequencies? If that is what they are implying is it correct? If it is correct how can it encompass all frequencies? I am sure there must be a basic concept I am missing, but thank you for your time in helping me to understand this.

 

[davenn]

Hi there :)

When someone says “arcing encompasses all frequencies”

interesting definition, personally I prefer to see the arc as a broadband RF source.

The RF arc noise you are hearing isn't really any different to the basic spark gap transmitters that Marconi started with. Tho he did make some effort to have them roughly tuned, from memory in the 400MHz region. But on a huge scale think of lightning strokes. They are just massive arc (spark gap transmitters) with RF spread from at least ELF 10KHz right through to at least 1GHz

The thing is to remember is that an arc like a welder, lightning and its ensuing RF emission isn't controlled by any form of oscillator, tuned circuit or filtering. Hence it is rich in harmonics right across a large portion of the spectrum

cheers
Dave

 

[sophiecentaur]

It is not surprising that 'spark gap' transmitters are, and have been for a long time, Illegal!
Great things if you are a pioneer in Radio - when there's no one to interfere with and you don't have valves or transistors. But deadly , otherwise.

 

[Averagesupernova]

It is not surprising that 'spark gap' transmitters are, and have been for a long time, Illegal!
Great things if you are a pioneer in Radio - when there's no one to interfere with and you don't have valves or transistors. But deadly , otherwise.

I would change that to: Using a spark gap for the sole purpose of radio transmission is illegal. Obviously arc welders are not illegal.

 

[sophiecentaur]

Arc welders will still have to follow some code, surely? I admit that you couldn't forbid welding on RF resonant structures because no one could identify them. I wouldn't mind betting that there's some specification about HF interference levels though and although the local fields may be quite high / detectable, they will drop off fairly quickly without a properly designed radiating system attached.

But the OP asks about : " I keep hearing the same sort of answers like “arcs encompass all frequencies".
An arc consists of a random string of high current impulses. Each impulse would look like a spike if you draw a graph of current against time. If you analyse this in terms of frequency (and any signal can be looked at both ways) it looks like a huge range of frequencies, from audible (you can hear it) to several tens of MHz and more. All the impulses add together to produce a wide 'spectrum' of random sounding hiss. This spectrum is constantly changing and there will always be some frequencies that fall inside the receiving band of an mf or vhf radio receiver. This raises the level of the general 'noise floor', which consists of all the other forms of noise and interference that you have to deal with. This sort of noise can often be more noticeable because it may also carry with it an audible buzz at the mains frequency. Structured noise is always more objectionable than truly random noise. Hence, the shash you get from a computer can be particularly annoying - even though it may satisfy all regulations about interference levels.

 

[gtacs]

Thank you for the replies!
Let me see if I understand. By "random string of high current impulses" you are referring to the number of electrons "jumping" the gap between the electrode and return path. The reason it is random is because the arc is constantly changing due to the environment in which it is "arcing" (I have watched "arcs get larger, smaller, and rotate around, even though the distance between the two electrodes stays the same). Because the "arc" is changing the number of electrons "jumping" the gap cannot be consistent/controlled.
As far as analyzing the signal, in order for there to interference at, let's say, 100MHz, there would have to be a "random high current impulse" every 10pS as freq=1/time?
You are correct there was a definite audio buzz at the main freq.
Do you think you would notice any significant noise increase at the harmonic frequencies of the originating source of the "arc"?
I am also going to assume the range/level of the interference is is going to increase as the current increases.
So, lightning would cause the same type of interference?
Thanks again for the replies, it has helped me alot.

 

[davenn]

So, lightning would cause the same type of interference?

As I said in my first post ... yes it does :)

Dave

 

[sophiecentaur]

Thank you for the replies!
Let me see if I understand. By "random string of high current impulses" you are referring to the number of electrons "jumping" the gap between the electrode and return path. The reason it is random is because the arc is constantly changing due to the environment in which it is "arcing" (I have watched "arcs get larger, smaller, and rotate around, even though the distance between the two electrodes stays the same). Because the "arc" is changing the number of electrons "jumping" the gap cannot be consistent/controlled.
As far as analyzing the signal, in order for there to interference at, let's say, 100MHz, there would have to be a "random high current impulse" every 10pS as freq=1/time?
You are correct there was a definite audio buzz at the main freq.
Do you think you would notice any significant noise increase at the harmonic frequencies of the originating source of the "arc"?
I am also going to assume the range/level of the interference is is going to increase as the current increases.
So, lightning would cause the same type of interference?
Thanks again for the replies, it has helped me alot.

Personally, I don't think that introducing electrons into the explanation is helpful. It's a layer of complexity (full of misconceptions) that is not necessary. An arc is, essentially, a 'bulk effect'. The way the current flows through an ark is very irregular and non-linear but it basically consists of many spikes of high current, at random, closely spaced, intervals.
The frequency content of such a signal is very much like a familiar noise signal which is modulated by the supply voltage and by short term, localised,variations of ion density within the arc (like the weather and a candle flame).

You have answered your own question here. Yes, the variations of the supply (e.g. mains hum) will modulate the noise and, hence, the sound of what gets into your radio receiver. These are more 'sidebands' of the basic spectral components of the noise than harmonics of the supply, though. It boils down to the same thing.

A high current spark will, of course produce more interference than a low current spark under the same conditions.

Lightning has a different characteristic - with shorter spikes than an arc welder. But that's obvious. The noise-like signal it produces (in short bursts) has a lot of RF content and the length (and height) of some lightning tracks would probably mean that they would be more efficient radiators of RF energy than a small arc welding unit, on the ground and with a proper mains filter included.

 

[gtacs]

Excellent! Thank you very much for the replies.