Spark fatness vs current

[Circlotron]

Suppose we have a spark gap and we ramp up the voltage until a spark jumps the gap. Any stray capacitance across the gap will be discharged as the voltage drops from what was needed to jump the gap down to the sustaining voltage. Now suppose we put an external capacitor across the gap. When the spark again jumps the gap the initial current will be huge. What about the thickness of the spark while the capacitor is getting discharged? We could be talking thousands of amps here, so there would surely be a decent magnetic field surrounding the spark that would influence the size and shape of the spark.

[ovnr]

It depends on the energy delivered into the spark, not the current alone.

Sparks are just ionized air, after all. If you have thousands of amps but it only lasts for 1 ns, you're not actually going to get a massive fat spark; there will, however, be a delightful EM pulse - especially due to the required rise/fall time.

[n3rdx]

What is your medium: air, a single gas, partial vacuum or vacuum ? What is the material of the cathode terminal? !Plasma Engineering! Paschen curves!  Breakdown!

Actually, breakdown is the phenomenon you want to look at. You really don't need high voltage to make a spark. Possibly you need to look at the Volts/meter number for the medium you need to cross/to overcome the work function so that electrons can be stripped from the terminals and then a breakdown process is started (which will not finish, it will keep on going until a arc appears).

Try a clamped inductive switch: pull the switch, charge the inductor, force the clamp open and hold on to your safety glasses. Initial step: cross the breakdown phase, then supply current into the anode/cathode junction (like an arc welding process - touch the terminal lightly to create the spark, and then DRAW the arc by separating the terminals slowly while it is melting).

[n3rdx]

I developed electric propulsion systems that went on cubesats some time ago. Here is a partial schematic of one of the workable power processing units (PPU). Flow is a suitable low voltage timed trigger to pre-driver system, essentially a highly charged shot provider, which controls the gate of the IGBT which in turn becomes the "valve" that controls the charge/discharge cycle. Whatever you are going to test (requiring a spark), connect it to the Anode (A) and Cathode (K) terminals at the right edge of the schematic. All component values are non-critical. Take precautions not to let the pre-driver be charged, WITHOUT A LOAD or DISCHARGING it after your cycle of experiments is finished. It will melt and do. bad. things. to other components if the energy is not drained. Like, the IGBT. Lost more than a few that way.

[NiHaoMike]

The gate resistance looks way high. I generally start with 4.7R and then fine tune with the help of a scope.

[n3rdx]

Component values in this schematic are made up. But I have used a variety of values reflecting what I had on hand when I was breadbording various configurations. The value of R  is discussed in IGBT datasheets.

[Circlotron]

It depends on the energy delivered into the spark, not the current alone.
Sparks are just ionized air, after all.

The magnetism is proportional to the current through the spark. I’m interested in whether the ionised channel diameter is affected by the magnetism circulating around it in proportion to the current through it.

[Circlotron]

What is your medium: air, a single gas, partial vacuum or vacuum ? What is the material of the cathode terminal? !Plasma Engineering! Paschen curves!  Breakdown!

Actually, breakdown is the phenomenon you want to look at. You really don't need high voltage to make a spark. Possibly you need to look at the Volts/meter number for the medium you need to cross/to overcome the work function so that electrons can be stripped from the terminals and then a breakdown process is started (which will not finish, it will keep on going until a arc appears).

Try a clamped inductive switch: pull the switch, charge the inductor, force the clamp open and hold on to your safety glasses. Initial step: cross the breakdown phase, then supply current into the anode/cathode junction (like an arc welding process - touch the terminal lightly to create the spark, and then DRAW the arc by separating the terminals slowly while it is melting).

No, I am not interested in breakdown, or making a spark per se. My question is stated in the OP.

[T3sl4co1l]

I don't know of the calculation offhand.

I do know that, regarding magnetic fields, when the current gets high enough to be strongly self-interacting, you get into the realm of Z-pinch, where the current compresses the channel smaller and smaller.  But this does not happen uniformly, as is characteristic of plasmas -- at the same time, it twists and knots up, turbulently.  To the eye (given suitable filtering and magnification), this might happen fast enough to look like a fat blur.

Twisting can be avoided if done fast enough, see Sandia's Z Machine.

Tim

[bson]

With thickness, do you mean cross sectional area or density?  Clearly the density will drop as the flow expands when in air.  In a perfect vacuum it will have no density other than the electron flow.

I would guess with the cap the current can be sustained longer, which means the breakdown plasma will expand further and the cross-section expand further during the spark.  After all, it must have really high pressure and expand rapidly, limited primarily by the duration of the breakdown and the ability of the harness to supply energy, and if the cap can add energy my assumption would be the breakdown conductor will get a little bigger.  But, not being a physicist... 

[iMo]

The magnetic field surrounding the spark will not act such it would squeeze the plasma into a smaller diameter. That would be against the fundamental laws of physics, imho.
If you want to compress the plasma into smaller diameter you must create a special shaped magnetic field with help of an additional energy taken from somewhere else..

[T3sl4co1l]

The magnetic field surrounding the spark will not act such it would squeeze the plasma into a smaller diameter. That would be against the fundamental laws of physics, imho.
If you want to compress the plasma into smaller diameter you must create a special shaped magnetic field with help of an additional energy taken from somewhere else..

FYI:

[iMo]

In order to squeeze the plasma (the same as your spark) ie. in ITER https://en.wikipedia.org/wiki/ITER you need a nuclear power station feeding the superconductive coils inside the chamber..
For any current (better to say "energy") in the spark's plasma you would need a much bigger current (better to say "energy") from elsewhere to form it such it squeezes its diameter down.. 

[Teledog]

Define "fatness" {you talking to me?} 

I get very fine arcs from a 10kV old copier Xformer - intermittent, mind you
Very fine arcs from a VDG
I get "fat" continuous arcs from any neon sign transformer (9-12kV@~30mA)
I do believe any fast "spark" will not be "fat"
Any continuous arcing will ionize the air and become "fat" (ie: TIG/MIG welding/plasma cutting )

Mmm..fat hamburgers...

[duak]

My first experience with high voltage was with an ignition coil from an auto engine.  Without the condenser across the points ie., capacitor across the switching contacts, the secondary spark was thin and orange plus there was a significant spark between the points.   With it, the secondary spark was blue, thicker and definitely louder and there was almost nothing between the points.  I also found it packed a bigger punch if you accidently touched it with the ole' digital voltmeter (finger).  If memory serves, the capacitor decreases the voltage risetime and also increases the lifetime of the points.

Here's an article on the waveforms in an inductive discharge system:  www.motor.com%2Fmagazinepdfs%2F052005_04.pdf&usg=AOvVaw2KYpjBZi0g4f02CaFbFYtZ