Off topic but perhaps fundamental

Why does a spark generate RF?

Bill

" wrote

Why does a spark generate RF?

Because it's a square wave ?

"Brian Gaff" wrote:

No, not really, a spark generates electromagnetic radiation, rf and light
are the same in this respect, of course many things dictate the actual
detection of the various frequencies, like the resonant lengths of anthing
conductive in the circuit.

Well, yes, of course it's electromagnetic radiation, the question concerns
the generation of radio frequencies.

Because it's a square wave ?

By which I was referring to Fourier's theory concerning the equivalence
between a square wave and the summation of a series of sine waves.
OK, a spark is just one cycle of a square wave, but...

"Brian Gaff" wrote in message
...

Its a broad band discharge?

So you're saying that the current flowing in a spark is inherently "noisy",
over a wide range of frequencies. Is that right? In other words the
amplitude of the current varies effectively at random?

That would certainly generate wideband electromagnetic variation, which of
course would come from the entire circuit, not just the spark gap itself.

Presumably this "noisy randomness" is something to do with the way the gas
is ionised - and the way the ions behave - down at the molecular or atomic
level.

I know almost nothing about radio theory, but I vaguely understood that a
circuit can't radiate at longer wavelengths than the physical length of the
circuit itself, but now I've written it down it doesn't seem right. Anyway,
my thought was whether the length of the spark gap (and the rest of the
circuit) affects - or determines - the bandwidth of the radiated RF.

I wonder if it's possible to make a "clean" spark by altering the pressure
or make-up of the gas in the spark gap. Thinking of those glass globes that
have an electrode in the middle and produce smooth, sweeping and curling
paths of glowing light - I guess those glowing tracks are like "smooth" or
"clean" sparks, are they? So having a smooth, continuous flow of current,
they won't radiate significantly?

Fascinating question, Bill.

SteveT

"John Legon" wrote in message
news
OK, a spark is just one cycle of a square wave, but...

Well, I don't think the "square wave" analogy is necessarily the best here,
as it definitely implies a cyclical function. My Fourier theory is a bit
rusty, but I think the closer analogy is that well known and closely related
case of an (effectively) instant rise-time stimulus. Dammit - I can't
remember how I should express it. It's used to analyse systems (mostly as a
theoretical construct) as it generates equal power at all frequencies. It's
a kind of "single shot" stimulus with a vertical leading edge. The
mechanical equivalent is hitting something with a hammer - the recipient
receives energy at all frequencies (in theory, in practice over a wide
frequency range) which is why you can use a hammer blow to make anything
"ring" - very big or very small, very high-pitched or very low-pitched.

Anyway, that, of course, only relates to a single spark. I think Brian's
mental model is a bit more illuminating, as it relates to a continuous spark
discharge, and discusses why such a continuous spark should continue to
generate RF, even after the sharp leading edge has passed.

Really, both must apply - the single-shot stimulus from the leading edge of
the spark (which generates wideband noise) followed by the continuous
wideband noise from the ionisation of the gas molecules.

SteveT

On 12/06/2010 02:24, wrote:
Why does a spark generate RF?


Mobile phone on the utility belt.

--
Adrian C

Steve Thackery wrote:

"Brian Gaff" wrote in message
...

Its a broad band discharge?

So you're saying that the current flowing in a spark is inherently
"noisy", over a wide range of frequencies. Is that right? In other
words the amplitude of the current varies effectively at random?

That would certainly generate wideband electromagnetic variation, which
of course would come from the entire circuit, not just the spark gap
itself.

Presumably this "noisy randomness" is something to do with the way the
gas is ionised - and the way the ions behave - down at the molecular or
atomic level.

I know almost nothing about radio theory, but I vaguely understood that
a circuit can't radiate at longer wavelengths than the physical length
of the circuit itself, but now I've written it down it doesn't seem
right. Anyway, my thought was whether the length of the spark gap (and
the rest of the circuit) affects - or determines - the bandwidth of the
radiated RF.

I wonder if it's possible to make a "clean" spark by altering the
pressure or make-up of the gas in the spark gap. Thinking of those
glass globes that have an electrode in the middle and produce smooth,
sweeping and curling paths of glowing light - I guess those glowing
tracks are like "smooth" or "clean" sparks, are they? So having a
smooth, continuous flow of current, they won't radiate significantly?

Fascinating question, Bill.

SteveT

In the 60s I worked in a busy radio & TV service department and it
occurs to me that discharging large electrolytic capacitors and the EHT
from CRTs never AFAICR caused RF interference on any of the radios -
this, of course, being an environment where a lot of equipment was
operating in much greater proximity than normal.

EHT flashover - anyone else remember the Ekco LOPT shrouds that
decomposed? - produced audible cracks but nothing from adjacent radios.

Mostly these were DC arcs - though the Ekco LOPTs were prone to leak AC
as well. However, it was common practice to check Line Output stage
activity by drawing long arcs off the EHT rectifier anode with an
insulated screwdriver - and I don't recall this causing interference,
either, though we always knew when somebody was ringing the battery
operated back door bell on the floor below us, even though we rarely
heard the bell itself!

--

Terry

On 12 June, 11:53, Terry Casey wrote:
Steve Thackery wrote:

"Brian Gaff" wrote in message
...

Its a broad band discharge?

So you're saying that the current flowing in a spark is inherently
"noisy", over a wide range of frequencies. *Is that right? *In other
words the amplitude of the current varies effectively at random?

That would certainly generate wideband electromagnetic variation, which
of course would come from the entire circuit, not just the spark gap
itself.

Presumably this "noisy randomness" is something to do with the way the
gas is ionised - and the way the ions behave - down at the molecular or
atomic level.

I know almost nothing about radio theory, but I vaguely understood that
a circuit can't radiate at longer wavelengths than the physical length
of the circuit itself, but now I've written it down it doesn't seem
right. *Anyway, my thought was whether the length of the spark gap (and
the rest of the circuit) affects - or determines - the bandwidth of the
radiated RF.

I wonder if it's possible to make a "clean" spark by altering the
pressure or make-up of the gas in the spark gap. *Thinking of those
glass globes that have an electrode in the middle and produce smooth,
sweeping and curling paths of glowing light - I guess those glowing
tracks are like "smooth" or "clean" sparks, are they? *So having a
smooth, continuous flow of current, they won't radiate significantly?

Fascinating question, Bill


Can of worms opened springs to mind!!

found this .............

"Spark gap" transmitter circuits were built very much like you would
expect, from their names: an air gap through which a high-voltage
electric spark jumped. Because the pulse durations of the sparks were
so short, the equivalent output frequencies spanned a very wide range,
ultimately rendering this technology impractical due to interference
between multiple transmitters.

Notes:
Anyone who has ever heard "popping" noises on an AM radio produced by
a (pulsed) electric fence of the type used around farms to keep
animals from wandering off will understand how spark-gap transmitters
broadcast across a large range of frequencies.
This question could very well lead into a fascinating discussion on
Fourier transforms, if your students are so inclined. According to
Fourier theory, the shorter the duration of a pulse, the broader its
frequency range. The product of uncertainties for the pulse's location
in time and its frequency is equal to or greater than a certain
constant. Theoretically, a pulse of infinitesimal width would
encompass an infinitely wide (infinitely uncertain) range of
frequencies.
Incidentally, the math behind this is precisely the same as for
Heisenberg's Uncertainty Principle: that quantum physics theory which
states the certainty of a particle's position is inversely
proportional to the certainty of its momentum, and visa-versa.
Contrary to popular belief, this phenomenon is not an artifact induced
by the act of measuring either position or momentum. It is not as
though one could obtain perfectly precise measurements of position and
momentum if only one had access to the perfect measuring device(s).
Rather, this Principle is a fundamental limit on the certainty
possessed by a particle with regard to its position and momentum.
Likewise, an infinitesimal pulse has no definite frequency.

In article
,
wrote:
Why does a spark generate RF?

The simple and basic answer is that any movement of electric charge
produces a variation in the EM fields around the charges.

From which you can go through the behaviour of charges and their associated
fields to discover that the accelleration and/or movements of charges in a
'spark discharge' will generate a corresponding 'pulse' or 'burst' of EM
field flutuations/changes.

Steady movements of charge mainly affect the E-field. Accellerations tend
to affect the H-field. Beyond that the details will depend on the details
of the pattern of charge movements.... and you might have to deal with
Maxwell... :-)

Slainte,

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html

In article , Steve Thackery
wrote:

"Brian Gaff" wrote in message
...

Its a broad band discharge?

So you're saying that the current flowing in a spark is inherently
"noisy", over a wide range of frequencies. Is that right? In other
words the amplitude of the current varies effectively at random?

Yes. It is essentially an 'avalanche' process. So quite noisy in detail.

That would certainly generate wideband electromagnetic variation, which
of course would come from the entire circuit, not just the spark gap
itself.

Presumably this "noisy randomness" is something to do with the way the
gas is ionised - and the way the ions behave - down at the molecular or
atomic level.

Yes.

I know almost nothing about radio theory, but I vaguely understood that
a circuit can't radiate at longer wavelengths than the physical length
of the circuit itself, but now I've written it down it doesn't seem
right.

It can radiate at lower frequencies (longer wavelengths) but the efficiency
may be reduced. It also depends what "the circuit" may include.


Anyway, my thought was whether the length of the spark gap (and
the rest of the circuit) affects - or determines - the bandwidth of the
radiated RF.

The spark gap is the 'device' but the connecting 'wires' and other
arrangements are also part of the circuit or antenna. Witness all the early
experiments that used loops, coils, plates, etc, attached to the spark gap
to change the spectrum and efficiency. Note that this can 'tune' the gap
itself as it causes the potential to vary across the gap, modulating the
current during the spark, or creating a controlled burst of sparks.

I wonder if it's possible to make a "clean" spark by altering the
pressure or make-up of the gas in the spark gap. Thinking of those
glass globes that have an electrode in the middle and produce smooth,
sweeping and curling paths of glowing light - I guess those glowing
tracks are like "smooth" or "clean" sparks, are they? So having a
smooth, continuous flow of current, they won't radiate significantly?

Depends. Quite often teaching labs used an induction coil (like in the
older cars) to generate a series of pulses of high voltage.

Slainte,

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html