Chris Pearson:

Dutch of the Elm:

 Where's the high frequency coming from?  A 230V arc might be caused twice every cycle, i.e. once per half cycle, and that's about it (the arc current itself will follow the sine wave, where there is current flowing).  So the crackling or buzzing should have a 50Hz sound to it. 

No, I don't think that is correct. The frequency is much higher than a mains hum, which is two and a bit octaves below middle C. It is more like little shots firing off so you hear the frequency generated by each arc.

What makes you say that?  Concerning the current waveform, mains arcing produces a shouldered 50Hz Sine wave.  Where current stops due to the system voltage being lower than the arc's striking voltage, the current waveform flat-lines.  This occurs at each zero point, hence the shouldered appearance.    


Example as below.

Hi,

The waveform in the pic is as taken by me, no filtering was applied intentionally but maybe some was inherent via my equipment?  (Oscilloscope/leads etc).

Are you saying that you think that harmonics were present? 

What are you referring to as sampling artefacts? 

What makes you expect a fast attack, slow decay characteristic?

Thanks – you may be schooling me soon. 

Dutch of the Elm:

Chris Pearson:

Dutch of the Elm:

 Where's the high frequency coming from?  A 230V arc might be caused twice every cycle, i.e. once per half cycle, and that's about it (the arc current itself will follow the sine wave, where there is current flowing).  So the crackling or buzzing should have a 50Hz sound to it. 

No, I don't think that is correct. The frequency is much higher than a mains hum, which is two and a bit octaves below middle C. It is more like little shots firing off so you hear the frequency generated by each arc.

What makes you say that?  Concerning the current waveform, mains arcing produces a shouldered 50Hz Sine wave.  Where current stops due to the system voltage being lower than the arc's striking voltage, the current waveform flat-lines.  This occurs at each zero point, hence the shouldered appearance.    


Example as below.

That's an electrical waveform, but let's assume that if you connected a microphone to your oscilloscope, you would get something similar. 


Now apply a fourier transform and you will have various harmonics. That is why arcing buzzes rather than hums or whistles.

Also, a series arc will limit the current drawn as it is inline with the load and as far as impedance goes, it is additive. I.e. if the cable doesn't move under load, I don't see why it would be moving (due to magnetic effects) under series arc conditions.

Chris Pearson:

 

That's an electrical waveform, but let's assume that if you connected a microphone to your oscilloscope, you would get something similar. 


Now apply a fourier transform and you will have various harmonics. That is why arcing buzzes rather than hums or whistles.

That's fair to say.  A colleague of mine said similar regards Fourier analysis but I took it that the harmonics it would identify would be only theoretical, i.e. a way to describe the waveform mathematically but not actually altering how the waveform would sound.. Now I need to feed this through a speaker and compare to a pure sine wave, next time I have a chance.


Many thanks.

Dutch of the Elm:

<snip... > 

I took it that the harmonics it would identify would be only theoretical, i.e. a way to describe the waveform mathematically but not actually altering how the waveform would sound.. Now I need to feed this through a speaker and compare to a pure sine wave, next time I have a chance.


Many thanks.

Hmm. the difference between a C on a violin (near triangle wave - the shape of the string under the bow), and the same C on a flute (near pure sine) is only the presence of harmonics with the right levels and phase relationship to the fundamental. To most of us they do indeed sound different ?
Edit
As always someone on youtube already has the same idea   and you can listen to sine, square and triangle waves.

Waveforms with sharp edges can sound much more noticable than smooth edged ones.

Now this is due partly to the ear's response to harmonics, and partly to the fact that small diameter moving objects make poor low frequency speakers,  (as the air has time to go round the moving part from the front to the back, rather than form a decent pressure difference ) but can be quite good treble ones,   If you listen to that you tube video on small 'tinny' laptop speakers the difference is more marked than with larger HIFi ones.






The sampling artefact here is not the traditional aliasing, but a rise time that is so fast we cannot measure. The blocky pixels  in your plot are giving a minimum resolvable time step of about 400us, to my uncalibrated eye - all we can say is that for part of the attack the rise time is faster than that.Now  400us is about the rise time of 1.1KHz ( the D or E  couple of octaves above middle C ? note frequency table ) so there are at least components present at that frequency and may well be quite bit higher.

Attack is fast - you can get impact ionisation of plasma in air at almost the speed of sound.  Decay takes longer, as you have to wait for positive and negative charges to recombine by a random walk process if there is not enough voltage gradient to sweep them out - and once the arc is struck the voltage gradient is quite low compared to striking. A sideeffect iof this is that although 50Hz arcs go out at the zero crossings, higher frequency ones are more like DC, as the plasma does not have time to go dark and is still lit and conductive during the current reversals - an RF arc is more of a scarey hiss than a snackle.