why do spark gaps cause oscillations with a capacitor.

[Capernicus]

In an ideal sense, a spark gap oscillator doesnt even work, if u think about it like the capacitor reaches the spark gap distance and then it would only discharge just enough to drop below the spark gap distance, and it wouldnt cause an oscillation, it would just pass into the capacitor and through the spark gap together, no oscillation.

I dont know why, and I wish you guys could help me with it, but my best guess is somehow when the first bit of electricity comes through then it allows more after it at a lower voltage, because the air has "plasmaized" or something.

Can someone tell me something thats more closer to the practicality of the situation?   Thankyou.

[IanB]

Yes, there is a breakdown or threshold voltage where a spark can be initiated. Until the spark starts the gap is open circuit and has infinite resistance. Once the spark begins there is an ionized path for the current to flow that has a very low resistance. So current will continue to flow through the spark (or arc) until the current source is exhausted. If the current source has a large capacity (e.g. like an arc welder) you will get a very hot, very long lasting arc.

[Capernicus]

Thanks.

So does this mean even reverse travel from another source will also travel though the ionized pathway?

[TimFox]

Don’t forget the inductance in the circuit, known to Marconi.
50 years ago, relaxation oscillators with a 300 V dry battery through an appropriate series resistor to a capacitor in parallel with a small neon bulb were popular, before we had the 555 and LEDs.

[IanB]

So does this mean even reverse travel from another source will also travel though the ionized pathway?

Yes. There are systems where a low power, high voltage source initiates the arc and then a high power, lower voltage source continues the arc once it is initiated.

[Terry Bites]

There is always some parasitic inductance in the system. No matter what you do it will be there lurking in the capacitor and the interconnections. This creates a resonant circuit with the capacitor, so that a spark gap will always generate radiation. The resonance voltage can be enough to refire the spark gap.

When the spark gap reaches its breakdown voltage the air is ionized and a current can flow relatively unimpeded though the plasma channel.
The current pulse stimulates the LC circuit into resonance. Look at the original Hertz experiment- those loops of copper are inductances.

[mawyatt]

Don’t forget the inductance in the circuit, known to Marconi.
50 years ago, relaxation oscillators with a 300 V dry battery through an appropriate series resistor to a capacitor in parallel with a small neon bulb were popular, before we had the 555 and LEDs.

These circuits with ionizing bulbs were also used in very old cheap analog oscilloscopes for the time base sweep. Van Der Pol discovered injection locking utilizing ionizing bulb relaxation oscillators. Later folks made use of ionizing bulbs for strobe use, and still in use today. Very useful highly non-linear devices.

Best,

[T3sl4co1l]

There is always some parasitic inductance in the system. No matter what you do it will be there lurking in the capacitor and the interconnections. This creates a resonant circuit with the capacitor, so that a spark gap will always generate radiation. The resonance voltage can be enough to refire the spark gap.

When the spark gap reaches its breakdown voltage the air is ionized and a current can flow relatively unimpeded though the plasma channel.
The current pulse stimulates the LC circuit into resonance. Look at the original Hertz experiment- those loops of copper are inductances.

Even more specifically: there is stray inductance along wires, and capacitance between wires, from the sheer fact that they are finite size, and the speed of light is also finite.

As for the spark, it has a negative resistance characteristic, which is relevant for long duration arcs; it may also simply be that a spark is impulsive, and the charge delivered by that impulse causes the attached LC network to ring.  In either case, it's down to dynamics of the system, how electricity moves through it over time, especially on very short time scales, as sparks break down in fractional nanoseconds.

Tim

[TimFox]

Poulsen (spell check keeps changing that to “Poulenc”) found that an arc could be stabilized with a magnetic field.  The Federal Radio Co. (predecessor of ITT) bought the rights and improved the invention (“Federal Arc”).  One of the magnetic armatures was rescued from a scrapyard and used in Lawrence’s first cyclotron.
Such an arc presents a negative (incremental) resistance to its terminals:  connecting a resonant circuit there then forms an oscillator.  Since the arc is slow to stabilize, on/off keying was not practical for Morse transmission, but a relay switching a second capacitor into the circuit produced FSK.  The Federal Arc and Alexandersson alternator were the first true CW transmitters (as opposed to the evanescent-wave spark transmitters of Marconi), but were soon made obsolete by vacuum tubes.
A permanent magnet around a small 6D4 thyratron was used by General Radio in their 1390B noise generator.