As mentioned, this is a kiosk substation, so it's fully enclosed, similar to this:

http://www.ausgrid.com.au/~/media/Images/Network/Network%20Projects/local%20projects/kioskimage.jpg?h=197&w=197

I have in the past experienced RFI from dirty high voltage insulators on poles but, again, that tends to create a characteristic arcing type hiss, whereas what I'm experiencing seems to have a low frequency fundamental that sounds to me like a motor running (and knocking) at maybe 2000 rpm.

I have a neighbour that works in a Schneider factory that builds those things, I could ask him if he could run it by them?

We do get Farm electric fences get into the tired, run down phone cables around here when they short but that is normally a slow even pulse. Some of those bigger power units have fans in them.

Marc

When I talk to the Ausgrid engineer tomorrow, I am going to try to persuade him to pay a visit.

Given that I can use the RFI to trace the path of the underground 415 volt supply cable from the kiosk on the street to the switch room on the property (i.e. the path that Ausgrid is responsible for) I'm hoping that will be sufficient to get them to meet me onsite to investigate.

Duconbuster: Sad sign of the times! Will have to try & filter it out somehow......

I think avoidance of switch mode technology is the only real answer. Gel batteries or good old linear type power supplies.

Okay, I spoke this morning with an Ausgrid engineer. After hearing the symptoms, he said that he had not experienced a kiosk substation being responsible for such RFI but nonetheless advised me to raise a quality of supply complaint to his attention, and he will arrange for a supply inspector to attend.

So, I've done that and now await further contact from Ausgrid.

(See post below for sound bite.)

May not be able to guess the secret sound, but interested to here it.

I manged to create a YouTube file from a recording I made from a transistor radio sitting on my workshop bench. There is no way I can use an AM valve radio in this environment. I can receive FM on a portable transistor set, so it's affecting mainly (only?) AM.

The interference, as detected by my portable AM radio, is even worse outside my workshop, especially under the building supply cable, and the 3 phase feeders that snake around the building in cable trays.

The frequency and amplitude is unvarying. It is present 24x7, so isn't time of day dependent.

Excuse the sparks graphic in this clip; it's just something I grabbed off the web to fill the screen.

https://youtu.be/e3mA0n8dvyI

I actually don't think its a motor, It may be an inverter, or a fault with a cap / inductor in the circuit, that has set itself up as a relaxation oscillator? I would expect a sawtooth wave for that scenario.

That I would be like the flicking CFL driven by capacitive induction, when the switch is off.

How big is that file & have you tried looking at it with a CRO to see how pure it is, or is not?

All quite fascinating really.

Marc

The MP3 file is ~210Kb.

I can see the waveform in my audio file editing software, here:

...and a frequency analysis, here:



Yes, it could be an accidental relaxation oscillator circuit. Whatever it is, it packs some punch as I can detect it on the street around the kiosk substation. Also, in the switch room it's so strong that I could not get a location fix on it. I would need some DF gear with attenuation capability, even then it might be too strong.

I haven't yet put a CRO on the mains. I'll do that tomorrow via a step down transformer (actually probe proximity will probably be sufficient). A spectrum analyser would be handy.

Image 1&2 not coming up at the time I looked. 210Kb is an attachment size. They do use the power lines to transmit data, that is another wild guess. Do like the idea of a spectrum analyser.

Marc

The images have been sent to Brad. He'll insert them eventually.

210Kb is 20 seconds worth of audio in MP3 format, which is the sample size I happened to grab.

Again, this RFI is relentless and continuous, 24/7.

Back in 1986 I was on holiday in Australia, Cairns in Queensland. I had an AM/FM cassette recorder portable, and the first evening in town I listened around on the AM dial, locals and skip. I did pick up a fair amount of interference, suspect it was coming from the CFL lights the hotel had on its grounds. Those had to be the first generation CFLs back in 1986.

I taped some of the stations I heard, mp3s at http://www.wa2ise.com/aussie_airchecks_apr86/
In http://www.wa2ise.com/aussie_airchecks_apr86/4am%204ca%20apr86.mp3 at 1:12 and at 2:25 you can hear some of the noise under a local station.

I would agree on some of those early CFL's there was one in my previous workshop & there was no such thing as reception when it was on.

This also reminds me of an incident where RFI was stuffing up Telstra's internet locally (made the press): It was traced to a light in a refrigerator? So don't discount the weird & wacky.

My new fridge has a switchmode in it & the transistor radio does not like it and the split system. As I said regulators asleep, you can get away with anything.

Marc

At this stage I'm not discounting anything, but want to try to eliminate possible problems with street supply first. Still waiting to hear back from Ausgrid on arranging an inspection.

If nothing obvious there, then I will arrange to isolate power from common area functions (lighting, ventilation, etc) one by one. If no luck there, I will arrange to isolate power from each business unit one by one.

Common area functions are just about all controlled by timers, so most, if not all, are not 24/7 operation. The RFI is 24/7. And yes, most of the timers are themselves electronic, so haven't discounted a timer as being the source, either.

I actually confirmed the FAX by shutting it off, & in reality, systematically shutting down circuits is basically the only way to go.

Marc