I don't mean regular 50Hz/60Hz buzz or normal cracking/clicking sounds during lamp warm up.

One of my HPS lamps makes a single "gong" sound in about 5 seconds after turned off. I've got it recorded in mp3 if anyone is interested.
My LED lamp I'm using on my bedside cabinet makes occasionally strange sounds (sorta similar to phone DMTF tones but in a lower key). I couldn't identify the source of this sound for a few days until I realized these are load management signals in mains (to switch some high energy consuming appliances on/off etc.) making electronics in that lamp "sing". Quite silent but a bit spooky at night.
Another retrofit HPS lamp is yelling in periods lasting seconds/tens of seconds for about 10-15 minutes after turned off. It's an about 15kHz sound (something like an old TV set but stronger), easily audible two rooms away (through the open door). It's pretty spooky. I suspect there is a microscopic crack in the outer envelope. When the lamp is cooling down, the air goes in and the crack works like a whistle.

Do you also have lamps that make strange/spooky/funny/... sounds?

Not really lamps, but ballasts.
And surprisingly not as much the magnetic components, but capacitors, mainly the ceramic ones...
I was long time wondering what is happening until quite recently, when I have discovered, than the piezo beepers (in many electronic devices) and the  higher capacitance ceramic capacitors use essentially the same material for the dielectric...

There is that sound with rising frequency (from few KHz to ultrasound) when charging the flash of a camera

When i was a kid i took a camera apart and put the components apart on long wires, to figure out who makes the noise. It came from the alkaline battery powering the circuit



Perfect Start A ballasts (with thermal start relay) make a click when they reclose the contact, few seconds after switching off the light

Sometimes the ballast is mounted in a way that the entire lantern body (sheet metal) responds to the snap action, and it makes quite musical sound. This happens to higher extent when the lantern itself is mounted in a recessed channel in the ceiling, which is made of sheet metal too

When a big array of lanterns with Perfect Start A ballasts is switched off, the individual relays click off each with a bit different timing (manufacturing and wear variations, ...) and it sounds like a "rain" falling on the sheet metal ceiling from above - starting few seconds after switch off and lasting for a few seconds more



Street lighting here, when installed on concrete or wood poles, is not Earthed. So the lantern may have internal metal body parts, or even the entire lantern body may be made of metal, but is electrically floating

Sometimes cycling HPS lanterns make quite loud SNAP sounds when the ignitor is making its pulses (~1..4 pulses/sec) when the lamp is cycling. I imagine they are results of isolation breakdown somewhere in th lantern (arcing to the unearthed metal body), when the extinguished lamp does not clamp the ignitor voltage and weak isolation somewhere is getting full 5KV



I had a 25W incandescent make some buzz (not at 100 Hz, but at some high harmonic) when powered by square wave from a UPS



I like trance music (among others). Sometimes i imagine sounds in the music as if they were odd sounds made by old lighting installations

... the piezo beepers (in many electronic devices) and the  higher capacitance ceramic capacitors use essentially the same material for the dielectric...

Something like thyristor drives in our old trams (streetcars) or trolley buses? They used to play really nice songs during acceleration or slowing down. I like that a lot as a kid.

I like trance music (among others). Sometimes i imagine sounds in the music as if they were odd sounds made by old lighting installations

LOL - would be great to compose a song from such sound samples. I suppose it could be well accepted provided listeners wouldn't know about the origin of those sounds.

The songs are great as they are. Where i visit in my imagination its not ONLY lighting there, its whole places. The lighting is just part and related only to few of the sounds

Something like thyristor drives in our old trams (streetcars) or trolley buses? They used to play really nice songs during acceleration or slowing down. I like that a lot as a kid.

Well, that comes from a different place (at least when speaking about the Tr14/15 Skoda trolleybusses):
The traction power is DC, usually 600V (when one leg is grounded, so when they are mixed with trams, like in Brno or so) or 750V (when the supply is bipolar, so +375/-375V; when they are alone without any trams or so, e.g. in Budweiss).
And as you know, you may turn the thyristor ON by it's gate, but to turn it OFF, you need to the device. And for that purpose there is a second thyristor, separated by either a capacitor, or with the more modern circuits by a series LC.
In both cases there is an inductor, used to recharge the capacitor back quickly, so the main thyristor could be switched off soon after it was switched ON, so allowing reasonably short pulses, so high operating frequency of the regulator (to not need that heavy filter chokes; they are already contributing about 2 tons per motor, so powered axle).
And the design of the power stage (for both motors in case of the articulated Tr15) there is placed just behind the front left wheel, behind the steel cover door. And it is placed there so, the commutating inductor, exposed to short current lulses, is about 10cm from the steel door. So quite a "nice" electromagnetic loudspeaker...

With the trams I do not know, how exactly the components are arranged, but I don't think it will be too different - just because e.g. the KT8D "music" is nearly the same as from the Tr15...

But that is really a history now, this was used in vehicles designed in the early 80's. In the later ones form the 90's (Skoda Tr21/22, the T3M overhauls,...) were using components completely controlled by their gate (GTO's or even IGBT's), there the noisy commutation LC network is not present anymore. The higher frequency allowed to get rid of the filter choke, but that mean the main motor get exposed by larger ripple current, so it become "singing" by it's own.

And the newest designs (from the 2000 and later) use AC motors (the newest Skoda ForCity even uses single permanent magnet motor per wheel and interestingly no axle between left and right) with pure IGBT inverters operating at kHz range, so they are quiet...

Thanks, that's an interesting piece of history.

(to not need that heavy filter chokes; they are already contributing about 2 tons per motor, so powered axle).

Two tons of chokes per one motor? That's simply scary! But I know ammeters on tram dashboards showed some hundreds of Amps during acceleration. Also it wasn't so rare that fluorescent lighting inside a tram (not sure with trolleybuses) was being restarted in the first second of acceleration.

And the design of the power stage (for both motors in case of the articulated Tr15) there is placed just behind the front left wheel, behind the steel cover door. And it is placed there so, the commutating inductor, exposed to short current lulses, is about 10cm from the steel door. So quite a "nice" electromagnetic loudspeaker...

I thought loudspeakers need a thin (and lightweight) membrane to play - especially higher frequencies. I'd rather expect a small piece of metal to play those songs, not a heavy steel door.

author=merc link=topic=4564.msg32238#msg32238 date=1451753725]
Thanks, that's an interesting piece of history.

(to not need that heavy filter chokes; they are already contributing about 2 tons per motor, so powered axle).

Two tons of chokes per one motor? That's simply scary!

To be exact, this is in the 750V version of the Skoda Tr15 - so the articulated trolleybus made in 80's and early 90's. It has two motors, one driving the middle axle, second the 3'rd axle (in the "trailer" section). But I won't expect the Tr14 ("short" version, so without the "trailer") and the 600V versions to be much different.
If any such filter was used in the trams, I don't know.

But I know ammeters on tram dashboards showed some hundreds of Amps during acceleration. Also it wasn't so rare that fluorescent lighting inside a tram (not sure with trolleybuses) was being restarted in the first second of acceleration.
[/quote]
The T3 had the fluorescents connected directly to the 600V supply (one resistive ballast powering two F25T12 fixtures), so there any significant undervoltage means they extinguish (and In Prague, where was common to have just the first car with the pantographup, I may expect the resistances over all the cabling were higher; In Brno, where all cars in the train use their own pantograph for supply and the cable carried just the control sighnals, the fluorescents never died on acceleration, only when there was an insulated section or so).
All newer trams (include the modernized T3's) have all lighting powered from the 24V battery, so independent on the traction power.

I thought loudspeakers need a thin (and lightweight) membrane to play - especially higher frequencies. I'd rather expect a small piece of metal to play those songs, not a heavy steel door.

Well, the door, like the complete vehicle body is just a steel frame with rather thin steel skin. And that skin then vibrates in the magnetic field from the commutation choke. By the way the choke is just about 8 turns of 10x3mm copper rod on about 30cm diameter, but the peak current there is about 500A, but it is constant, independent of the motor current (it is given by the traction supply voltage, the inductance and the capacitance of the commutation LC; it should always be higher than the motor current for at least recovery time of the thyristor). So the input power into that loudspeaker is not really small, the commutation LC handles 375kVA at nominal 750V, with the maximum supply tolerance, so 900V, it becomes 540kVA. That is of course a reactive power, but even 0.1% of losses there means real power of nearly a kW...

Not relating to motors or lights-but the modulation transformers and reactors in some of our 250Kw transmitters "sing" with the modulation-125Kw of audio being handled there.So you stand outside the transformer vault and can hear the sound.Like a speaker in there!Years ago while riding the Metro subway in Wash DC remember a few paper clips on the car floor.They would stand up and even dance slightly when the train started and stopped.So--don't put anything magnetly sensitive on the floor!Figure the clips were above the motors reactors.

Not relating to motors or lights-but the modulation transformers and reactors in some of our 250Kw transmitters "sing" with the modulation-125Kw of audio being handled there.So you stand outside the transformer vault and can hear the sound.Like a speaker in there!Years ago while riding the Metro subway in Wash DC remember a few paper clips on the car floor.They would stand up and even dance slightly when the train started and stopped.So--don't put anything magnetly sensitive on the floor!Figure the clips were above the motors reactors.

I live near Washington DC. And use the subway to go to college. I'm gonna get a few paper clips and try that on a few Series of cars they have and find which one does it. Of course I'm gonna make a vid of them doing that. LOL!

When I went to a large HF transmitting station you could here the programme modulation on the metal transmission lines like box sections. Àlso a bit off topic but the lights in the station would flicker in sympathy with the mod.Also I had a lightbulb sing at me for a few minutes before it went EOL.

Don't hear the modulation on our line enclosures-probably at that site they were not properly grounded.-Or could be corroded areas in the sections acting as diode "detectors"This also happens with AM stations.The lights flickering-don't have that here since our powerline supply is very "stiff"10Mva-our load right now is less than half the substations rating.Our plant is fed with a 10Mva 115Kv-4160V 3 Ph.Our site used to run more programs.Budget constraints cut back on them and also reduced power on some programs.Our plant used to run two transmitters at 500Kw.Now its 250.And our site now has LED lot and yard lamps.

The HF site I visited had 10 500Kw transmitters the power supply was at 33Kv transformed down to 11Kv then down to 415 volts this was many years ago when the supply wasn't so stiff also often up to 4 tx would carry the same programme so all modulation peaks were at same moment resulting in very hi peak load

Quote from: Ash

I like trance music (among others). Sometimes i imagine sounds in the music as if they were odd sounds made by old lighting installations

LOL - would be great to compose a song from such sound samples. I suppose it could be well accepted provided listeners wouldn't know about the origin of those sounds.

I one time went & recorded sounds from various fluorescents (starters/ballasts/etc)...my intent was to use them & make a 'song' out of them...never did though.
I like trance music too (and a bunch of other types as well)

Unusual to have 500Kw transmitters running from 415V-the wiring to each would be HUGE-usually transmitters over 100Kw are wired to run from voltages like 2500,4160 and on up.The company I work for their overseas  plants the transmitters run from 11Kv 3Ph.Also their overseas plants run from their own large diesel genset plants.The country doesn't have enough power to run the site.We have a couple of transmitters from a closed overseas plant-in storage containers.We don't use them-they used to run from 11Kv.If we were to use them-the main transformers would have to be replaced with 4160V ones.