I wonder if I can hear the 50Hz of the mains connecting a subwoofer in series with a 40w incandescent lamp. Will it work or I will just end with a lot of magic smoke?

What incandescent bulb will be the best?

Just get a fixture with a noisy ballast. Much simpler than wiring up a subwoofer to the mains, and less dangerous.

Why I want to use a subwoofer instead of a ballast is because I want to hear only the 50Hz tone, not the harmonics. If I use a 25w incandescent in series the subwoofer should take 25w too or not?

I do not understant the desire, as it would drive me mad after very short time, but that is your business... :-]

Connecting a speaker like you say may make the noise you intended, but it would be extremely unsafe.
Speakers are not designed to handle the mains voltage, here mean a potential difference between the voice coil (connected to one of the mains working line) vs the box, where you may touch it, so it is very high risk of an electric shock.
So for this reason, connecting a standard speaker in such way is a big NO-NO.

The only safe way would be to use a transformer 230/whatever SELV you find (e.g. 6 V is a quite common secondary voltage) and feed the speaker from the insulated secondary via some resistor or a small lightbulb (e.g. 6V/0.3A and 6 V/0.3 A torch bulb or 22 Ohm/3 W resistor will work the same as a 40W bulb in series with the speaker connected to the mains).
You may take the transformer from virtually any old clock radio or similar small mains powered gadget, the voltages there are in the 4.5..9 V ballpark, you just adopt the resistor so you get the 0.2 A through your speaker.

Just get a fixture with a noisy ballast. Much simpler than wiring up a subwoofer to the mains, and less dangerous.

Yep, I have a 400W HPS ballast I can hear anywhere in the flat, it's LOUD 

If this is an "active" sub, it's expecting <1V signal, and this idea is quite a bit different to implement. It would be a lot easier to find a frequency generator application to synthesize 50Hz.

A transformer is the safest way to drive the voice coil directly. 3VAC will drive a 4ohm coil at 2W (loud enough) and 6VAC of course will drive a 4ohm coil at 9W (plenty loud)

"hot" resistance of a 40w lamp is about 1300ohm or ~175mA. In series with a 4ohm coil this is only a drive power of ~0.1W. This will be quite faint at this low frequency. If you put the voice coil on the neutral side of the lamp, it's safer, but still not recommended. Mains voltage is deadly and you really shouldn't play with it.

I believe the tone generator in Audacity would generate a 50Hz tone that would give you the results you want. On the plus side, Audacity is a free downloadable software.

Most speaker cases are made of wood and dont have conductive touchable parts except the terminals, so for an experiment such speaker in series with some ballast on 230V could be sufficient

Besides, there is difference between making an experiment safely vs. making a semi permanent or permanent setup :

 - In experiment, safety is primarily about not touching the connected parts (this includes not letting little brothers and sisters, pets, ...) and not letting anything that fails start a fire (by keeping distance from flammable things, and by being there to respond to any fault situation)

 - In a permanent setup, there are more things to take into account - That wood is not so great isolator when subjected to high(ish) voltages over time, that the speaker terminals are not isolated, possibility of a fire if the ballasting device shorts out, and many more things. They sum up to "doing things properly and not as a hack". In this case, it means isolated supply to the speaker



Clean 50 Hz is quite different from what you would expect. Common hum heard from ballasts etc. is actually 100 Hz (as the ballast attracts the luminaire sheet metal for each half period) and the more "buzz" there is in the sound, the more harmonics (multiples) of 100 Hz are contained within. Also, 50Hz is near the bottom limits of frequencies we can hear (which is not a sharp cut off at 20 Hz), so for the same current will be quieter than 100 Hz

Incandescent lamp and Fluorescent ballast are not linear devices either :

In the lamp, especially lower power lamps with thinner filament, the filament have significant temperature change over the AC cycle. The resistance is highest when the current is highest, so the sine wave will be lowered around the peaks (a sort of very soft clipping)

In the ballast, if you use it alone (at stuck starter current) it starts to saturate when the current is highest, and its impedance drops. This leads to opposite effect from the lamp. The effect can be lowered if you use 2 ballasts in series

If you are interested in sounds of non sinewave 50Hz, you might be interested in connecting the speaker in series with a complete FL or HID lamp setup.... (using caution if it is an experiment, or an isolation transformer if it is a permanent setup)

I'll connect my subwoofer to a 10v transformer, way safer than connecting it directly to the mains, what I was thinking. 

If I connect my 8 ohm speaker to the 10V AC, I should get 13W of discipation in the speaker, I don't know if my 4.5" sub will be overdrived.

Connecting it directly to the mains it will just blow up...

The power calculation is correct, however - Measure the speaker resistance with a multimeter, it might come out lower than the rated resistance (which is rated at specific frequency). We are dealing with 50Hz here, which is bordering on DC in audio terms

I don't have a multimeter for test it, but naturally subwoofers resistances must be tested at low frequencies.

A multimeter is the most important test instrument you can have. If you have nothing else, you need one of these..!

A multimeter is the most important test instrument you can have. If you have nothing else, you need one of these..!

I'm saving some money for buy one, but at the moment I use my homemade continuity test made of 2x AA batteries, two homemade probes with sticks and screws and an (...) diode, so if there is resistance the (...) glows dim. 

This can be used to test continuity or distinguish resistances which are significantly different. In this case the resistances are single Ohms so this circuit won't detect any difference

My first multimeter was DT830 - one of the cheapest multimeter types made by many manufacturers. They are acceptable as a beginner's multimeter, cost anywhere between US $10..20 in retail. However, they are as unreliable as cheap multimeters can get

If you can obtain analog meters from decommissioned electrical equipment (simple one-function meters), that also can be used for experimenting

I used to have an Avo 8 meter. A bit of a beast, but it worked well. It did fail eventually though, I thought I'd struck lucky a week or two ago when I found an almost mint one on eBay, sadly I was outbid at the last minute, which happens to me a lot, I much prefer fixed price sales..!

Currently using a digital meter I got from the now-defunct UK electronics firm Maplins, it was a reasonable price and still works well, but I want to get one with a clamp current meter as the clamp meter I have is basically an industrial electrician's one and is designed for measuring hundreds of amps not small numbers